00 Front & TOC 4/17/07 8:31 AM Page i

Geologic Applications of
Gravity and Magnetics:
Case Histories
Edited by
Richard I. Gibson
Patrick S. Millegan
SEG Geophysical Reference Series No. 8
AAPG Studies in Geology, No. 43

Published jointly by the Society of Exploration Geophysicists and

the American Association of Petroleum Geologists
Tulsa, OK, USA
Printed in USA
00 Front & TOC 4/17/07 8:31 AM Page ii

ii
Society of Exploration Geophysicists American Association of Petroleum Geologists

SEG Editor: Larry R. Lines Association Editor: Neil F. Hurley

Geophysical Reference Series Editor: Eugene F. Scherrer Science Director: Jack Gallagher
Publications Manager: Ted Bakamjian Publications Manager: Kenneth M. Wolgemuth
Project Editor: Judy Hastings Managing Editor, Publications: Anne H. Thomas
Copy Editor: Rowena Mills
Book design and layout: Martha Dutton, Wakan Grafix

Geologic applications of gravity and magnetics: case histories /

edited by Richard I. Gibson and Patrick S. Millegan.
p. cm. – (Geophysical references; v. 8)
Includes bibliographical references and index.
ISBN 1-56080-078-X
1. Gravity prospecting–Case studies. 2. Magnetic prospecting–Case studies.
I. Gibson, Richard I. II. Millegan, Patrick S. III. Series.
TN269.G412 1998
622’ . 152–dc21 98-28734
CIP

Copyright © 1998 by the Society of Exploration Geophysicists

Society of Exploration Geophysicists

Box 702740 American Association of Petroleum Geologists
Tulsa, OK 74170-2740 P. O. Box 979
USA Tulsa, OK 74101-0979
USA

All rights reserved. No part of this book may be reproduced, stored THE AMERICAN ASSOCIATION OF PETROLEUM GEOLOGISTS (AAPG) DOES NOT ENDORSE OR
in a retrieval system, or transcribed in any form or by any means, RECOMMEND ANY PRODUCTS OR SERVICES THAT MAY BE CITED, USED OR DISCUSSED IN AAPG
electronic or mechanical, including photocopying and recording, PUBLICATIONS OR IN PRESENTATIONS AT EVENTS ASSOCIATED WITH THE AAPG.
without prior written permission of the publisher.
00 Front & TOC 4/17/07 8:31 AM Page iii

SEG and AAPG

wish to thank the following for their generous contributions to this volume

Major Contributors
ACCEL Services, Inc. Geophysical Marathon Marathon Oil
Exploration Oil Company Company
Technology
ARCO ARCO
Northwest
Geophysical
Ark Geophysics Ltd. Geosoft Inc. Associates

Bell Geospace Inc. Oryx Energy

Gibson Consulting
Company
CGG/Geoterrex-
Dighem
LCT, Inc. Shell International
Earthfield Technology Center for Remote
Inc. Sensing Potential
Field Studies
Geophysical
Exploration & Lisle Exploration
Development Company Unocal
Corporation
Contributions are applied against the production costs of publication, thus directly reducing this book’s purchase price and making it available to a greater audience.
iii
Geologic Applications of Gravity and Magnetics: Case Histories
00 Front & TOC 4/17/07 8:31 AM Page iv

iv

Donors The Editors

Carlos Aiken and Holly Hunter Huston, in memory of Gene Greene Richard I. Gibson
Michael Alexander Richard I. Gibson was trained as a geologist at Indiana University. After a
four-year career analyzing kidney stones, he shifted (1975) to gravity and
Austin Exploration Inc. magnetics exploration with AeroService in Houston. He spent eight years
with Gulf’s Technical Services Center and five years as director of gravity
and magnetics for Everest Geotech before founding Gibson Consulting in
Bird Geophysical Golden, Colorado, in 1989.

EDCON

Galileo Geophysics, Inc.

GeoPotential

Integrated Geophysics Corporation

MBL, Inc.

Mobil Corporation

Joyce O’Brien Patrick S. Millegan

Paterson, Grant & Watson Ltd. Pat Millegan, a graduate of Baylor University, has worked exclusively in
potential fields during his 22-year career in geophysical exploration. He
Photo Gravity Corporation has been Marathon Oil’s company-wide potential fields specialist for the
past 15 years. He previously worked with AeroService and Mobil. He has
worked with data from every continent except Antarctica. Millegan cur-
Pearson Technologies, Inc.
rently chairs SEG’s Gravity and Magnetics Committee. He was co-guest
editor of The Leading Edge’s January 1998 issue.
Scintrex Ltd.

SEG Gravity & Magnetics Committee Chairs, past and current (Mike Kowalski,
Tom LaFehr, Betty Johnson, John Bain, and Pat Millegan)

Sidney Schafer and Associates

00 Front & TOC 4/17/07 8:31 AM Page v

Table of Contents

1 Preface/Introduction, Richard I. Gibson and Patrick S. Millegan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Gravity and Magnetics in Oil Exploration: A Historical Perspective, Richard I. Gibson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
De Magnete, Alan Reid and Richard I. Gibson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2 General/Planning/Large Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Integration of Gravity and Magnetic Methods in the Risk-weighted Exploration Decision Process:
Basin Reconnaissance, Robert Pawlowski . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Gravity and Magnetic Analyses Can Address Various Petroleum Issues, Elizabeth A. E. Johnson . . . . . . . . . . . . . . . . . . . . . . . . . 7
Use Higher Resolution Gravity and Magnetic Data as Your Resource Evaluation Progresses, Elizabeth A. E. Johnson . . . . . . . . . . . 8
Gravity’s Role in a Modern Exploration Program, E. F. Greene and C. M. Bresnahan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
General Philosophy 1: Gravity and Magnetic Interpretation, edited by David A. Chapin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Gulf of Mexico Continental Slope—Understanding the Magnetic Response Due to the Salt Intrusion, Corine Prieto . . . . . . . . . . . . . . 14
A Brief History of Micromagnetic Surveying, J. P. Land . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
High-resolution Aeromagnetic Surveying, Patrick S. Millegan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Gravity/Magnetic Signatures of Various Geologic Models—An Exercise in Pattern Recognition, Corine Prieto . . . . . . . . . . . . . . . . . . 20
Satellite Gravity, E. K. Biegert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Gravity Data Define Basin Structure and the Location of Major Oil and Gas Reserves: Examples from
Subandean Basins, Tunisia, and the U. S. Rocky Mountain Region, Johann-Christian Pratsch . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Utility in Exploration of Continent-scale Data Sets, Richard I. Gibson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
The Isostatic Gravity Residual of Onshore South America: Examples of the Utility of Isostatic Gravity
Residuals as a Regional Exploration Tool, David A. Chapin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
One Person’s Regional Is Another Person’s Residual, P. S. Millegan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

v
Geologic Applications of Gravity and Magnetics: Case Histories
00 Front & TOC 4/17/07 8:31 AM Page vi

Table of Contents
vi
The Compilation and Application of Aeromagnetic Data for Hydrocarbon Exploration
in Interior Alaska, John F. Meyer Jr., Louis J. Racic, Richard W. Saltus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
How Basement Lithology Changes Affect Magnetic Interpretation, Patrick S. Millegan, D. E. Bird . . . . . . . . . . . . . . . . . . . . . . . . 40

3 Basin Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Gravity Applications, E. K. Biegert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
The St. George Basin, Alaska, COST #1 Well: An Example of the Need for
Integrated Interpretation, David A. Chapin, SubbaRao V. Yalamanchili, Paul H. Daggett . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Magnetic Reduction-to-the-Pole at Low Latitudes, W. C. Pearson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Interpretation of Magnetic Anomalies at Low Latitudes: Potential Pitfalls, D. E. Bird, S. A. Hall, J, F. Casey, P. S. Millegan . . . . . . . . . . . . 53
General Philosophy 2, edited by David A. Chapin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Aeromagnetic Constraints on the Basement Structure of the Tunghai Shelf and the Okinawa
Trough in the East China Sea, S. Okuma, T. Nakatsuka, M. Makino, R. Morijiri . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Magnetic Frequency-depth Relationship, R. I. Gibson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Aeromagnetic Interpretation of Southwestern Continental Shelf of Korea, Czango Baag, Chang-Eob Baag . . . . . . . . . . . . . . . . . . . . . . . 63
Integrated Gravity Modeling of Salt Features in the Mississippi Salt Basin, Rhonda L. Schenk, John J. Morris, Stuart A. Hall . . . . . . . . . . . 68
Mississippi Salt Basin, R. I. Gibson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Kentucky Basement Note, R. I. Gibson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

4 Plays/Field Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Integration of Gravity and Magnetic Methods in the Risk-weighted Exploration Decision
Process: Prospect Definition Stage, Robert Pawlowski . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
An Integrated 3-D Gravity and Seismic Study of the Ouachita Frontal Thrust Belt, Pittsburg and
Latimer Counties, Oklahoma, Holly Hunter-Huston, E. F. Greene, C. L. V. Aiken . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Gravity Resolution, E. K. Biegert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Interpretation of Rift-stage Faulting in the West Siberian Basin from Magnetic Data, Richard I. Gibson . . . . . . . . . . . . . . . . . . . . . . . . 79
Mag Susceptibility Contrast versus Structure, R. I. Gibson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
00 Front & TOC 4/17/07 8:31 AM Page vii

Case History: Reddell Oil Field, Evangeline Parish, Louisiana, C. T. Austin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

Classic Salt Dome, R. I. Gibson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Nash Dome, R. I. Gibson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
General Philosophy 3: Concerning Modeling, edited by David A. Chapin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Magnetic Anomalies Associated with the North and South Morecambe Fields, U. K., Richard Morgan . . . . . . . . . . . . . . . . . . . . . . . . 85
General Philosophy 4, edited by David A. Chapin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
High-resolution Aeromagnetic Interpretation over Sierra and Yoyo Reefs,
Northeastern British Columbia, John Peirce, Erwin Ebner, Nathalie Marchand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
General Philosophy 5: Concerning Nonuniqueness, edited by David A. Chapin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

5 Mining and Environmental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

General Philosophy 6: A Different Approach (Mining and Environmental), edited by P. S. Millegan . . . . . . . . . . . . . . . . . . . . . . . 104
Gravity and Magnetic Study of the Southern Chocolate Mountains Region, Southeastern California:
Possible Extension of the Mesquite Gold Mine, K. L. Mickus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Acid Mine Drainage in Northern Ontario, T. Jeffrey Gamey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Sulfur Exploration with Core-hole and Surface Gravity, M. Alexander, K. O. Heintz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
An Airborne Multisensor Characterization of an Active
Nuclear Waste Site, T. Jeffrey Gamey, J. Scott Holladay, Jonathan Nyquist, William Doll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

6 State of the Art . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Airborne ALTREX™ Electromagnetic/Magnetic Definition of Hydrocarbon
Alteration Plumes, J. D. Rowe, R. S. Smith, D. D. Beattie, R. K. Warren . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Sources of Magnetic Anomalies over a Sedimentary Basin: Preliminary Results from the Coastal Plain
of the Arctic National Wildlife Refuge, Alaska, Jeffrey D. Phillips, Richard W. Saltus, Richard L. Reynolds . . . . . . . . . . . . . . . . . . . 130
The Rise and Fall of Early Oil-field Technology: The Torsion Balance Gradiometer, Robin E. Bell, R. O. Hansen . . . . . . . . . . . . . . . . . . 135
Results from a High-resolution, 3-D Marine Gravity Gradiometry Survey over a Buried Salt Structure,
Mississippi Canyon Area, Gulf of Mexico, Lincoln F. Pratson, Robin E. Bell, Roger N. Anderson, Dan Dosch, John White,
Clive Affleck, Andrew Grierson, Bryant E. Korn, Ronald L. Phair, E. K. Biegert, Peter E. Gale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

vii
Geologic Applications of Gravity and Magnetics: Case Histories
00 Front & TOC 4/17/07 8:31 AM Page viii

Table of Contents
viii
Gravity Gradiometry, E. K. Biegert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
The Role of Vertical Magnetic Gradiometry in Hydrocarbon Exploration, H. O. Seigel, T. J. McConnell . . . . . . . . . . . . . . . . . . . . . . . . 146
General Philosophy 7: Where Do We Go From Here?, edited by David A. Chapin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148

7 Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

8 Annotated Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

9 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
01 Preface-Intro 4/16/07 3:45 PM Page 1

Richard I. Gibson, Gibson Consulting

Patrick S. Millegan, Marathon Oil Company

1 Preface/
Introduction
01 Preface-Intro 4/16/07 3:45 PM Page 2
Preface/Introduction
This book is the result of a brainstorming session at a meeting in the fall
of 1992 of the American Association of Petroleum Geologists Committee
on Geophysics (now the Geophysical Integration Committee, chaired by
Joel Watkins), then chaired by Ray Thomasson. After some members had
suggested an “atlas” of 3-D seismic case histories, Ray turned to Dick
Gibson, who suggested the idea of a compilation of gravity and magnetic
case histories for nonspecialists. A draft book proposal went through sev-
eral iterations, with significant input from Pat Millegan, Allen Hittelman,
Ray Thomasson, and members of the committee. Dick Gibson and Pat
Millegan ultimately undertook the job of coeditors of Geologic Applications
of Gravity and Magnetics: Case Histories.
The idea for this book came from a perceived lack of recent, instruc-
tive examples of exploration-oriented interpretations of gravity and mag-
netic data. The Society of Exploration Geophysicists’ two volumes,
Geophysical Case Histories, are probably closest in philosophy to this
book. Published in 1948 and 1956, many of the examples in the Case
Histories are relatively dated and specific to particular areas. We hope this
new book provides an update that includes lessons about gravity and
magnetic exploration that can be applied to many parts of the world. The
Utility of Regional Gravity and Magnetic Anomaly Maps (SEG, 1985,
W.J. Hinze, editor) contains some excellent papers dealing with tectonics
that have clear bearing on hydrocarbon exploration, but no paper shows
the relationships among hydrocarbon accumulations, exploration, gravity,
and magnetics. Geophysical texts focusing on gravity and magnetics,
including L.L. Nettleton’s classics, include only a few (albeit often excel-
lent) case histories, and many are dated.
Thus, this book’s target audience is geologists and geophysicists in
operations offices, actively involved in exploration at any level from basin
analysis to prospect generation. Although most of the papers deal with
hydrocarbon exploration, several papers relate to gravity and magnetic
data in mining and environmental applications. A final section is included
on new developments, the state of the art.
The book is not intended for gravity and magnetics specialists
(although we hope they will find it interesting), or for geophysicists inter-
ested in theory, acquisition, and processing, unless those aspects are
important to the geologic and exploration problem and to the decision-
making process.
We believe that the philosophical approach to interpretation is almost
as important as some aspects of a technical interpretation itself. This book
reveals the diversity of philosophies that gravity and magnetic interpreters
embrace, as well as the common threads to which all interpreters aspire.
Preface/Introduction
2
Some aspects of these approaches are given in short statements of philos- Exploration Geophysicists supported this project from its inception. Our
ophy in various sections of the book, based on a free-ranging discussion employers (Gibson Consulting and Marathon Oil Company) allowed us
on interpretation that took place on the Internet in 1996. We offer thanks time and support for work on this task. Lastly, we thank the authors of
to David Chapin for compiling this discussion, and to all the participants the technical papers, short notes, and philosophical statements. Without
for their contributions. This book is a result of the work and enthusiasm them, this book would not exist.
of the entire community. Like Huckleberry Finn, if we had “a knowed what a trouble it was to
Gravity and magnetic data are relatively cheap, so ignoring them is make a book, we wouldn’t a tackled it. . . .” But all in all, it has been
not the best way to reduce risk in today’s competitive economic environ- exciting and interesting to work on Geologic Applications of Gravity and
ment. For regional studies, continent-scale data sets are available for most Magnetics: Case Histories. If there are deficiencies in it, we apologize,
of the world, and they provide a unique look at regional tectonics. More but we plead another Mark Twain observation: “An oyster has hardly any
than one-third of the papers in the book describe non-U.S. work. High- more reasoning power than a scientist has.” Twain undoubtedly would
resolution data for local interpretations make it possible to address ques- apply that statement manyfold to scientist/editors.
tions of fractures in rocks and subtleties of structure and stratigraphy, We hope you find Geologic Applications of Gravity and Magnetics:
expanding plays. Case Histories interesting, useful, and perhaps even controversial. This is
The sections of Geologic Applications of Gravity and Magnetics: a work by a community of intelligent specialists who are alive and active.
Case Histories reflect the broad scope of applications of gravity and mag- Gravity and magnetics have survived the 1980s and are here to help
netic data. Categorization is difficult, so there may be overlap in many remove the nonuniqueness inherent to seismic techniques. Today, even
papers with the approaches described in other sections. In one particular seismic acquisition and processing contractors are actively acquiring and
case, studies about high-resolution aeromagnetics, papers are distributed integrating gravity and magnetic data as parts of their standard geophysi-
throughout the book in appropriate places. cal products. At a minimum, this synergy can reduce exploration risk; at
“General/Planning/Large Scale” includes comparative analyses, its best, it can help you find oil and gas!
strategic projects, and some background examples. “Basin Studies”
focuses on tectonics and specifics within particular basins, and “Plays and
Richard I. Gibson
Field Studies” shows some examples of relatively local exploration appli- Golden, Colorado
cations. Following the “Mining and Environmental Studies” section, sev-
eral papers provide insight into gravity and magnetic techniques which Patrick S. Millegan
are “State of the Art,” developing today. Houston, Texas
This book is not a textbook, although we have tried hard to highlight 1998
the exploration lessons inherent in each technical paper. Additional
instructional aspects of the book are the glossary of gravity and magnetic
terms, provided by Integrated Geophysics Corporation (with assistance
from Richard Hansen of Pearson, DeRidder & Johnson) and an annotated
bibliography, which has pointers to the rich literature of gravity and mag-
netics. Other short “lessons” can be found in stand-alone illustrations or
short features throughout the book.
We thank Ray Thomasson for continual encouragement, suggestions,
and prodding. Reviewers, whose efforts are appreciated greatly, include
Dale Bird, Bill Pearson, Mark Odegard, and several anonymous review-
ers. We appreciate the help of the AAPG, especially Ken Wolgemuth, in
this, the first effort at serious book publication by the coeditors. We thank
all the members of the gravity and magnetics community—as diverse a
group as could be found anywhere—for suggestions, insight, and encour-
agement. The Gravity and Magnetics Committee of the Society of
01 Preface-Intro 4/16/07 3:45 PM Page 3
Gravity and Magnetics in Oil Exploration:
A Historical Perspective
Richard I. Gibson
Gibson Consulting, Golden, Colorado, USA
The first U.S. oil discovery using any geophysical method came in 1924
at Nash Dome, Texas, as a result of a survey with the Eötvös torsion bal-
ance. This gravity-measuring device was invented in 1888, and first was
used for hydrocarbon exploration in Czechoslovakia in 1915–16. It was
very slow to operate and was sensitive to near-surface irregularities, and
these problems provided the impetus for developing a sensitive pendulum
apparatus.
Wyckoff and Eckhardt tested a practical pendulum instrument in
Kansas and Oklahoma in 1925–26. After joining Gulf in 1928, they
refined and developed the tool and it was field-tested in Michigan in April
1930. Gulf and other companies tried other pendulum methods, but by
July 1932 the Gulf (Wyckoff) pendulum was in regular operation.
Cleveland Oil Field (Texas) was found by Gulf with this pendulum;
Conroe Dome, which was invisible in torsion-balance data, also was
defined. The pendulum was a great improvement, but still did not have
the sensitivity and speed of operation required for efficient exploration. At
their peak, pendulum crews could observe a maximum of about 250 sta-
tions per month. The next step was the gravimeter.
Apparently, Humble Oil tested the first gravity meter in the United
States, in 1930, but never placed it in regular use. Gulf developed a
gravimeter in 1932–35 which began routine operation in May 1935. This
instrument was the first practical field gravimeter, with an accuracy of
better than 0.1 milligal, and it was very fast to operate. The Gulf (Hoyt)
gravimeter received general industry acceptance, and thousands of sta-
tions were observed by Gulf crews from 1935 to 1955. An all-time Gulf
high was reached in March 1938, when 7300 stations were obtained; the
overall total for Gulf crews alone was more than 750 000 stations.
Industrywide, torsion-balance crews went from forty-plus crews in 1935
to zero in 1940, while gravimeter crews increased from one in 1935
(Gulf’s) to forty-plus in 1940 (sixteen were Gulf crews). Improvements
over the Gulf meter by Worden and LaCoste & Romberg increased the use
of the gravity method even more. The latter instruments are those in most
common use today.
Gulf had developed a practical water-bottom gravity meter by 1940.
This instrument was simply an encased Gulf gravimeter mounted on a
tripodal stabilizing arrangement. It could be converted to a land-based
meter with a few mechanical adjustments. The water-bottom meter was
used successfully to find salt domes in the Gulf of Mexico before good-
seismic coverage was available. The data also were used (and still are used)
to help plan detailed seismic surveys.
A by-product of seismic and gravity exploration in the swampy coastal
areas of the Gulf of Mexico was the invention in about 1950 of the marsh
buggy, which was used to transport geophysical crews and equipment.
The pendulum apparatus first was used in a moving vehicle (a subma-
rine) in about 1930. Not until 1958 did gravimeters come into use on sur-
face ships. After the LaCoste & Romberg stabilized-platform gravity meter
was introduced in 1965, shipborne gravity surveys became common, usu-
ally in conjunction with seismic and magnetic acquisition as well. This is
the type of shipborne meter in greatest use today. It virtually has sup-
planted more expensive and cumbersome water-bottom operations.
Significant advances in gravity methods took place in the mid- to late
1970s, with development of borehole and airborne gravimetry. Airborne
gravity refers to data acquired from a moving helicopter, in comparison
with helicopter-supported surveys, in which the helicopter lands for a
datum reading. After considerable controversy and development, air-
borne gravity has found application in several exploration situations, such
as remote, unexplored jungle and desert areas, rugged topography, and
the land-ocean interface.
Among the most recent advances in gravity exploration is renewed
use of high-resolution gravity gradiometry. This volume has two papers
that address this technique.
Ground magnetometers were in common use in the 1920s and
1930s. A 1926 Gulf Oil discovery in Garza County, Texas, was based on
interpretation of a magnetic survey. As a reconnaissance tool, however,
ground magnetometry was slow and cumbersome. In an attempt to elim-
inate these drawbacks, the first experimental airborne magnetometer was
tested in the USSR in 1936. This instrument used a rotating induction coil
and had a sensitivity of 1000 gammas, which was inadequate for oil-
exploration work.
The Gulf airborne fluxgate magnetometer was developed by Victor
Vacquier and others in 1939–41. Its 1-gamma sensitivity would make the
meter valuable in exploration, but it was realized immediately that this
device also could detect submerged metallic objects such as submarines.
After successful “detection and disposal” of an enemy submarine in a test
flight in December 1942, the Gulf magnetometer was used throughout
World War II. In April 1945, test flying for hydrocarbon exploration began
in western Pennsylvania. The Gulf airborne magnetometer, the second to be
invented but the first practical instrument, became the industry standard.
Geologic Applications of Gravity and Magnetics: Case Histories
The Gulf instrument was supplanted in about 1955 by the proton-pre-
cession magnetometer, and in the 1960s by optically pumped alkali-vapor
magnetometers. The 0.01-gamma sensitivity of the latter provided the
ability to measure tiny vertical gradients in the Earth’s magnetic field. The
Gulf magnetometer still is used frequently as a ground monitor, and in the
1970s, new fluxgate surveys began appearing on the market as a low-cost
alternative to surveys flown with the higher-sensitivity instruments.
Unlike seismic data, gravity and magnetic data go out of date very
slowly, if ever. In spite of refinements in acquisition and processing, old
data still can provide a wealth of information to the interpreter. Few oil
companies today acquire their own gravity or magnetic data. Specialized
contractors perform an admirable job both in acquisition and processing.
Many oil companies do, however, prefer to do their own interpretations,
integrating proprietary information into the study. Many companies also
have their own processing, refinement, and modeling programs, although
these also are becoming inexpensive and varied on the commercial mar-
ket. Powerful personal-computer-based software packages are available
for everything from modeling to automated depth estimation.
Most analyses of the state of the art in potential fields focus on
advances in computer-based processing techniques, many of which are
merely improved (and very useful) ways of displaying data and processed
data. Another aspect of modern potential-fields interpretation is the direct
application of detailed basement tectonic analysis to prediction of changes
in stratigraphy and structure at a local level, even to the point of prospect
generation. Such application can suggest the distribution of fracture
porosity in rocks, the preferred locations of carbonate buildups, and
answers to other subtle exploration questions.
Improved resolution in terms of acquisition, processing and, most
importantly, geologic information content are making both gravity and
magnetics useful tools for modern exploration programs.
References
Keller, G. R., 1988, The development of gravity and magnetic studies, emphasizing
articles published in the GSA Bulletin: GSA Bull., 100, 469-478.
LaFehr, T. R., 1980, Gravity method: Geophysics, 45, 1634.
Reford, M. S., 1980, Magnetic method: Geophysics, 45, 1640.
Paterson, N. L., and Reeves, C. V., 1985, Applications of gravity and magnetic surveys:
The state-of-the-art in 1985: Geophysics, 50, 2558.
Sweet, George, 1966, The history of geophysical prospecting: Science Press, 3 vol.
3
01 Preface-Intro 4/16/07 3:45 PM Page 4

Preface/Introduction
4
De Magnete, published in 1600 in
England, is commonly recognized as the
first publication of any sort which em-
ploys modern scientific principles. It is
also the first treatise on geophysics. The
author, William Gilbert (1540–1603),
was considered to be England’s most dis-
tinguished scientist during the Elizabethan
era. He was physician to the queen, and
he researched electricity and magnetism.
His descriptions of magnets and of the
Earth’s magnetic field as a centered
dipole are accurate, and the word elec-
tricity derives from his work. Gilbert also
brought the Copernican theory of plane-
tary motion to England. Originally in
Latin, De Magnete was translated into
English by P. Fleury Mottelay (Dover
Publishers, New York, 1958).

Alan Reid and R. I. Gibson

02 General Planning 4/16/07 4:02 PM Page 5

2
Gravity’s Role in a Modern Exploration Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
E. F. Greene and C. M. Bresnahan, Gene Greene & Associates
Gulf of Mexico Continental Slope—Understanding the
Magnetic Response Due to the Salt Intrusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Corine Prieto, Integrated Geophysics Corporation
Gravity/Magnetic Signatures of Various Geologic Models—An Exercise
in Pattern Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Corine Prieto, Integrated Geophysics Corporation
Gravity Data Define Basin Structure and the Location of Major Oil and Gas
20
General/
Reserves: Examples from Subandean Basins, Tunisia, and the U. S. Rocky
Mountain Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Johann-Christian Pratsch, Petroleum geologist
The Isostatic Gravity Residual of Onshore South America: Examples of the Utility
28
Planning/
Large Scale
of Isostatic Gravity Residuals as a Regional Exploration Tool . . . . . . . . . . . . . . . . . . . . 34
David A. Chapin, LaCoste & Romberg
The Compilation and Application of Aeromagnetic Data for Hydrocarbon
Exploration in Interior Alaska . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
John F. Meyer Jr., State of Alaska, Department of Natural Resources,
Division of Oil and Gas
Louis J. Racic, Paterson, Grant & Watson, Ltd.
Richard W. Saltus, U. S. Geological Survey
02 General Planning 4/16/07 4:02 PM Page 6
Integration of Gravity and
Magnetic Methods in the
Risk-weighted Exploration
Decision Process: Basin
Reconnaissance
Robert Pawlowski
ADA Geophysics Inc.
(formerly with Integrated Geophysics
Corporation)
Introduction
The exploration industry is moving
broadly in unison toward the use of
increasingly more rigorous, quantitative,
and consistent techniques for assessing
exploration risk, a movement facilitated
in part by the equally widespread and
parallel effort to compile and maintain
large, globally comprehensive, histori-
cal/technical exploration databases.
Probabilistic models of a prospect’s
chance of succeeding (or failing) are con-
structed routinely now, in part on the
basis of various geologic and geophysical
data interpretation inputs to the explo-
ration-risk model. As this note will dis-
cuss, the geologic integration of gravity
and magnetic data has a definite impact
on the evaluation of certain of the explo-
ration-risk model parameters.
The geologic integration of gravity and
magnetic data can be used to reduce risk
at two key stages of the exploration
process. The first stage, often referred to
as basin reconnaissance, is the role auto-
matically relegated by most explorers to
gravity and magnetic methods. Contrary
to this widespread notion, however, grav-
ity and magnetic methods are equally
effective for reducing risk at the more
local scale typified by the prospect itself.
In cases in which the application of grav-
ity and magnetic analysis negatively
impacts a prospect’s viability, valuable
exploration resources can be redirected
at finding different and, it is hoped, more
economic prospects.
This short philosophical note will sum-
marize some of the chief ways in which
risk is reduced by an exploration philoso-
phy embracing the use of gravity and
magnetic methods during the two main
stages of the exploration process.
Role of Gravity and Magnetics
in the Exploration Process
Basin Reconnaissance Stage
Exploration-risk parameters evaluated
through the geologic integration of grav-
ity and magnetic data during the basin
reconnaissance (or new venture opportu-
nity) stage include:
1) regional hydrocarbon trapping
structures
2) regional hydrocarbon fetch areas
3) regional hydrocarbon source-thick-
ness estimation
4) regional hydrocarbon migration
pathways
5) thermal maturity
6) optimization of seismic survey
placement
The basin reconnaissance stage marks the
time of a company’s initial interest in an
area. This stage is characterized by a mas-
sive effort to capture existing available
data so that the area’s potential for com-
mercial prospects can be evaluated best
using the precious resources available.
During this stage, the integration of grav-
ity and magnetic data is used, among
other things, to delineate basin and sub-
basin boundaries as well as regional
structures proximal to the basinal areas.
The location of regional trapping struc-
General/Planning/Large Scale
tures is, of course, one of the classic
exploration uses of the gravity and mag-
netic methods. In case the basement is
both dense and magnetic, basement
structure and depth can be delineated
and mapped using gravity and magnetic
data. Basement structure is of interest, of
course, because it often is related to
structure in the sedimentary section.
Additional confirmation of regional sedi-
mentary structure is made through use of
gravity data which, unlike magnetic data,
are also sensitive to structures in the sed-
imentary section (although modern,
high-quality, and high-sensitivity aero-
magnetic data are being used to map
some sedimentary structures for which
sufficiently large magnetization contrasts
exist).
Definition of the various basin and sub-
basin geometries enables the areal extent
of the possible hydrocarbon fetch areas
to be evaluated. This is, again, a classic
exploration application of the gravity and
magnetic prospecting methods. The posi-
tion of the fetch areas relative to each
other and to the regional trapping struc-
tures is a useful interpretive result of the
gravity and magnetic analysis. This result
facilitates the high-grading of prospects in
case multiple ones are identified.
Next, calculations of sedimentary thick-
ness are made, later to be related to pos-
sible source thickness (or lack thereof).
These two estimates ultimately are
linked to the volume of hydrocarbons
generated. In the ideal situation (i.e.,
local geology cooperating), magnetic
basement mapping and modeling are
used to define the depth (and thus, thick-
ness of sediments) and configuration of
the basin, and gravity modeling is used
to confirm the thickness and possible
constitution (e.g., clastics as opposed to
6
carbonates, etc.) of the expected sedi- to optimize the seismic program layout
mentary section. (e.g., seismic line orientation, spacing,
and length) by using information from
The location of regional trapping struc- the interpreted gravity and magnetic
tures relative to the fetch areas sets the data. Such an approach helps ensure that
stage for an analysis of the possible seismic moneys are spent most effectively.
hydrocarbon migration pathways, and
the focusing of those hydrocarbons by
regional structure. A nice example illus-
trating and linking these concepts is the
paper by J. C. Pratsch in this volume. It
discusses the influence of regional struc-
ture on the distribution of oil and gas
accumulations in Colombia, and the
interpretation of the regional petroleum-
system dynamics through the qualitative
integration of gravity data.
The one area of basin reconnaissance in
which gravity and magnetic data are not
apparently well exploited is that of ther-
mal maturity modeling. The inference of
present and past distributions of heat
within the sedimentary column is influ-
enced in part by the thermal properties
(thermal conductivities, for instance) of
the crustal rocks. The base of the crust
(or Moho) is definable with regional
gravity data, the basement depth and
configuration with magnetic data, and
the presence of anomalous conductors
(such as salt) often can be determined
with gravity data. Hydrocarbon thermal
maturity modeling currently does not
take good advantage of model parame-
ters estimated using gravity and magnetic
data, namely, the interpreted geometry,
depth, and lithology (and thus thermal
conductivity) of the different crustal
units.
Unfortunately, it is too often the case
that gravity and magnetic analysis is
undertaken after a seismic program has
been designed and shot. It makes more
sense, both technically and economically,
02 General Planning 4/16/07 4:02 PM Page 7
Gravity and Magnetic
Analyses Can Address
Various Petroleum Issues
Elizabeth A. E. Johnson
Unocal Exploration and Production Technology
Sugar Land, Texas, USA
Gravity and magnetic data can be used
in many ways to solve different explo-
ration problems, depending on the geo-
logic setting and rock parameters.
Although most think of gravity and mag-
netics as tools to map structure, these
data can be analyzed to provide insights
to other elements of petroleum explo-
ration and production. Table 1 describes
some of the techniques that can be
applied to address a wide variety of
issues. The impact of all these techniques
increases if the gravity and magnetic
analysis is integrated with other data and
studies.
Table 1.
Issue Gravity and magnetic tasks Integrated with
Source Rock Deposition
• Where were the source rocks deposited? Depth to magnetic basement Seismic data
• How deep are the source rocks? Regional basin enhancements Regional geology
Source Maturation
• Where are the “cooking pots” and fetch areas? Depth to magnetic basement Seismic data
• What is the present-day heat influx into Isostatic residuals Well data
the basin and how much does it vary? Sediment thickness Density and velocity data
• What is the thickness of the crust? Depth versus density modeling Heat-flow data
• What is the overburden? Regional structural modeling
Curie point (regional heat flow)
Delineation of volcanics
Hydrocarbon Migration
• How much relief is there on the basement? Magnetic inversion Well and outcrop data
• What are the “shapes” of the “cooking pots”? Depth to magnetic basement Topography
• Are major vertical conduits near source areas? Vertical fault identification Remote sensing
• Are major lineations present and how do they Gradient analysis Seismic data
relate with more recent geologic features? Regional depocenter and sediment path Sequence stratigraphic analysis
enhancements Seismicity
Reservoir Prediction
• Where are the thickest sediments? Depocenter and sediment path Seismic data
enhancements
• Where is the highest sand probability? Integrated basin modeling Lithologic data (outcrop and well)
• Where was the source of sedimentation? Density inversion Sequence stratigraphic analysis
• What is the influence of tectonics on deposition? Provenance (magnetic lithology) Biostratigraphic data
determination
• Have the sediment depocenters shifted over time? Sedimentary magnetic analysis
• What is the compaction history of the sediments? Paleomagnetic analysis
• Do the sands have lateral continuity and Integrated velocity analysis (2-D and 3-D)
connectivity?
Trap
• Where are the major structures? Residuals and enhancements Seismic data
• What is the structural grain? 2-D/3-D structural/stratigraphic modeling Outcrop information
• Are faults in the sedimentary section? Fault identification—gradient analysis Topography
• Are lateral porosity changes present? Structural inversion Remote sensing
Density inversion Seismicity
Vertical Seal
• Where are salt overhangs? Residuals and enhancements Seismic data
• How thick is tabular salt? Layer stripping Sequence stratigraphic analysis
• How thick are volcanics? Integrated 2-D/3-D modeling
Sedimentary magnetic analysis
Timing
• What are the ages of sedimentary features? Integrated 2.5-D structural/ Density and velocity data
• How do all the petroleum system elements stratigraphic modeling Seismic data
fit together and what is the timing? Layer stripping and enhancements Biostratigraphic data
Tectonostratigraphic analysis Back-stripping
Paleomagnetic analysis Palinspastic reconstructions
7
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:02 PM Page 8

General/Planning/Large Scale
8
Use Higher Resolution
Gravity and Magnetic Data
as Your Resource Evaluation
Progresses
Elizabeth A. E. Johnson
Unocal Exploration and Production Technology
Table 1.
Sugar Land, Texas, USA Tactics Regional reconnaissance Prospect identification Prospect risk reduction Asset delineation and Reservoir performance monitoring
Petroleum system analysis and risk assessment Drill-site decision (less development
Many times, explorationists believe that Play analysis Lease and G&G acquisition complex prospects) Drill-site decision Enhanced recovery
once they have a gravity/magnetic sur- Establishing exploration focus (complex imaging)
vey, they don’t need any more. Does that and G&G expenditure Tectonostratigraphic
hold for seismic data as well? Of course framework
Basin modeling
not. Today, increasing seismic resolution
is becoming the norm. Similarly, higher Gravity Isostatic residual Semiregional structural/ Detailed, integrated 2-D/3-D Integrated 3-D rock properties Time-lapsed precision gravity
resolution gravity and magnetic data Utilization Regional tectonic analysis stratigraphic modeling modeling (with seismic and velocity modeling
Basin and depocenter Target-specific enhancements horizons, density, and Integrated depth migration Integrated reservoir characterization
should be acquired as a play concept
enhancements velocity information) (pre- or poststack)
advances from reconnaissance to delin- Regional modeling Layer stripping for improved Porosity/pressure prediction Borehole gravity—remote Borehole gravity
eation. Even 4-D applications can be Digital data integration delineation of exploration Salt edge/base determination porosity detection
utilized in some situations. Table 1 de- (with remote sensing, etc.) targets Enhanced velocity analysis Detection of shallow hazards
scribes the gravity and magnetic tech- Sensitivity studies tied to
niques that can be used as a resource density and lithology
evaluation program progresses. Each type Gravity 1–5 mGal 0.2–1 mGal 0.1–0.5 mGal 0.1–0.5 mGal .02–.1 mGal
of application requires different data reso- Resolution 2–20-km wavelength 1–5-km wavelength 0.5–2-km wavelength 0.2–1-km wavelength 1–5 years
lution. Survey instruments and software Required* Continental grids, satellite Conventional marine and High-resolution land and .01-.005 mGal (borehole)
tools are available to meet all these speci- gravity, airborne gravity land surveys marine surveys High-resolution land, marine,
and gradiometer surveys
fications, but the cost can vary substan-
tially. Use this table to help plan or Magnetic Regional depth to magnetic Detailed basement Detailed, integrated 2-D/3-D Detailed 2-D/3-D modeling, None published
license the data and technology that fit Utilization basement interpretation modeling—faulting, base- inversion
Regional tectonic analysis Detailed fault and lineament ment structure, volcanics, Integrated depth migration
your need.
Euler deconvolution analysis salt edges, and sediment (pre- or poststack)
Curie point analysis Delineation of volcanics, timing Magnetostratigraphy
salt, and shale “Depth slicing” and lineament
analysis
Sedimentary magnetic
analysis
Magnetic 20-km spacing 2–5-km spacing 0.5–1-km spacing 0.25–0.5-km spacing
Resolution 5–8-km grid 1–2-km grid 0.1–0.5 nT 0.1–0.5 nT
Required* 1–5 nT 0.5–2 nT High-resolution, low-altitude High-resolution, low-altitude
Continental grids, older surveys Modern digital surveys, surveys surveys
marine surveys, digitized Borehole magnetometer
older analog surveys

* Typical required resolution; needs to be tailored to source depth and signal strength
02 General Planning 4/16/07 4:02 PM Page 9
Gravity’s Role in a Modern Exploration
Program
E. F. Greene* and C. M. Bresnahan
Gene Greene & Associates
Houston, Texas, USA
* Deceased
EDITOR’S NOTE: Greene and Bresnahan offer a step-by-step
approach to the use of gravity in the Gulf of Mexico. These
wise words are a fitting tribute to their many years of inter-
pretation experience.
Introduction
In this age of 3-D seismic surveys, seismic inversion, depth migration,
analysis of amplitude variation with offset (AVO), and personal worksta-
tions, can gravity data contribute to a modern exploration program? The
answer is definitely yes!
The following is an attempt to define the role of gravity in oil and gas
exploration, especially in the Gulf Coast, explaining its strengths and lim-
itations.
Much has been written about integrated exploration programs that
incorporate all geologic and geophysical data available, but in practice,
most prospects presented to management or to prospective investors con-
sist only of subsurface geologic information and seismic data. Many
prospect generators do not realize the existence of, or take the time and
effort to use, gravity data already in their files or readily available for pur-
chase. Most prospects can be enhanced and better defined by including
information derived from a gravity survey.
Most geologic features in the sedimentary section associated with the
accumulation of oil and gas are related directly to horizontal density
changes of magnitudes large enough to be mapped by an accurate gravity
survey. A partial list of such features includes anticlines, synclines, reefs,
faults, and horizontal changes in the thickness of salt beds, which, of
course, include salt domes, pillows, and ridges.
The map resulting from a gravity survey is a Bouguer map. A Bouguer
gravity map consists of gravity values (the vertical component of the
Earth’s gravity) which have been corrected for latitude, elevation, and ter-
rain. In the Gulf Coast, these maps are reduced to a sea-level datum. The
Bouguer map is the response of all the horizontal changes in density over
the mapped area, from the surface to the center of the Earth. To derive
the maximum information from these data, they must be “processed” and
interpreted.
The first step in this procedure is to derive a residual map. The resid-
ual maps used in petroleum exploration normally consist of gravity
responses from horizontal changes in density within the sedimentary sec-
tion. These maps can be used in several ways by an explorationist.
Because salt plays a major role in the tectonics of the Gulf Coast and is
the primary source of gravity anomalies of Gulf Coast residual maps, these
maps are very useful exploration tools. As seen in Figure 1, the density of
salt and cap rock (cap-rock density may vary from dome to dome but not
as a function of depth of burial) remains constant at all depths of burial,
while the density of sediments, in the Gulf Coast, increases with depth.
This variation in density results in narrow, intense maximum anomalies
over the shallow salt and cap rock, and wide, broad minimum anomalies
over the deep salt. A study of the minimum anomalies of the residual
gravity map, along with information from subsurface and seismic maps,
can help determine the center of mass of large, deep salt structures and
ridges, and the location and strike of faults. A similar study of maximum
anomalies will help locate shallow salt stocks and cap rock. This method,
although useful, is subject to many pitfalls such as anomaly overlap,
choice of the proper type of residual map, and lack of sufficient gravity sta- Figure 1.
tions to define the anomalies adequately. A more productive and precise
use of gravity residual maps is in conjunction with 3-D modeling.
(.25 gravity units), the elevation accuracy must be less than
The Anatomy of a Good Gravity Survey ± 6 inches because an error of one foot of elevation will create
The following steps must be taken to design and implement a gravity an erroneous elevation correction of .5 to .8 units. To maintain this
degree of accuracy, the error in location of the station must be held
survey that will produce an accurate Bouguer gravity map:
to less than 15.25 m (50 ft).
1) Determine the shortest wavelength anomaly (smallest diameter) that
will be encountered in the area to be surveyed. This must include Three-dimensional Modeling
both the desired sedimentary anomalies and the unwanted “noise” The most powerful tool available to the gravity interpreter is a 3-D grav-
anomalies, such as minimum readings from near-surface limestone ity modeling program designed around an accurate and versatile calcula-
sinks present in areas with karst topography. This wavelength will tion algorithm such as the Talwani and Ewing (1960). A gravity in-
govern the station spacing of the entire survey. At least two samples
terpreter, using constraints from subsurface and seismic data and sedi-
per wavelength are needed to prevent aliasing and adequately define
mentary density data from density logs, can construct detailed models of
an anomaly. Once the station spacing has been determined, the
entire survey should be metered on a square grid of this spacing. salt domes. Employing these models, one can detect the presence of and
can map unusual shapes and overhang of a dome. The modeling program
2) The accuracy of the survey is dependent on three primary factors: can be used not only to model single structures but also many structures
the proper location of the station, the measurement of the elevation over an extended area of deep-seated salt pillows and ridges, including
of the station, and the accuracy of the gravity meter. If we assume both the top and base of the salt. Reefs, faults, anticlines, synclines, and
a case where the accuracy of the gravity meter is ± .025 mGal igneous plugs also can be modeled by similar techniques.
9
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:02 PM Page 10
The following is our method of using 3-D gravity modeling to obtain
an accurate map of the salt.
1) A contour map of the top and base of the salt over the mapped area
is drawn, using all constraints from seismic and geologic data.
2) Each contour of the salt structure map is digitized.
3) The gravity contrast between the sediments and salt obtained from
density logs is used to calculate the gravity response of the salt at
each contour level.
4) The gravity response from all contour levels is integrated over the
entire modeled structure and a calculated gravity map is produced
which has grid points identical to the grid points of the residual
map.
5) The calculated map is subtracted from the residual map, producing a
difference map which shows the location and amplitude of the mis-
fit between the gravity response of the models and the gravity
response of the structures.
6) The modeled structures are altered and recalculated to obtain a bet-
ter fit between the calculated and the residual map.
7) This procedure is repeated through several iterations until the best
fit is found, while remaining within the constraints of the seismic
and subsurface data.
This procedure results in a structural interpretation that is the best fit
of the seismic, gravity, and subsurface geological data.
The Use of 3-D Modeling in the Gulf Coast
Natural features and man-made obstacles, along with permit problems,
prevent the explorationist from obtaining an ideal gravity map.
Nevertheless, in a mature area such as the Gulf Coast, we have the nec-
essary information for the employment of gravity modeling as a produc-
tive exploration and exploitation tool.
Gravity maps of sufficient accuracy and station spacing for 3-D mod-
eling of Gulf Coast salt domes and other structures are available for pur-
chase. If, in a particular case, the station spacing is inadequate, additional
gravity stations can be run at a reasonable cost. The well control is gen-
erally dense enough to provide adequate information about the depth to
salt, top thickness and density of the cap rock, approximate location and
depth to the domal flanks, and local variations of sedimentary densities.
General/Planning/Large Scale
Seismic lines across the domes also can be used to establish constraints of
their size and depth.
Salt domes
Over the last few years, we have modeled many of the piercement domes
of the Gulf Coast. A large percentage of the shallow piercement domes
have extensive overhang and very small, deep stocks. In some cases, the
narrow, deep stocks flare out again with greater depth to form large, deep
diapirs. Richton Dome of Perry County, Mississippi, is an example of such
a dome (Figures 2, 2a, and 2b). In some cases, the domes may be
detached completely from the deep mother salt layer or deep diapirs.
Seismic data processing over salt structures
Most seismic data-processing companies now have very sophisticated
2- and 3-D stacking and migration programs that need or require a start-
ing model prior to stack. Because of the extreme difference between the
low-velocity sediments and high-velocity salt, it is imperative that the
starting model of a salt dome or pillow be as accurate as possible. Gravity
models of the salt have been employed with success, as this starting
model.
Regional modeling
We have been involved in the construction of 3-D models over large areas
of the Gulf Coast, in conjunction with regional seismic lines, to map both
the top and base of the Louann Salt formation. These studies revealed the
existence of domes, extensive salt ridges, areas where the salt thins, areas
where the salt is absent, and areas where the base of the salt is far from
being flat, suggesting the possibility of structures below the salt.
This method of 3-D modeling also has been used to model the large
reef trends of the Edwards and Sligo formations, large regional faults,
igneous plugs, and basement features.
Residual Maps—Their Construction, Use,
and Limitations
A residual gravity map is computed by subtracting a regional map from
the Bouguer map or by filtering the Bouguer map with a wavelength fil-
ter. The result is a map in which some of the original anomalies of the
Bouguer map either have been removed or suppressed. One of the pri-
mary reasons for creating a residual map is to remove the anomalies orig-
inating in or below the basement and to preserve the sedimentary
anomalies undistorted. Through the years, many techniques have been
10
developed with varying degrees of success. The following is a brief dis-
cussion of three residual methods.
Modeled residuals
A model residual map is produced by creating a 3-D model of the base-
ment anomalies, calculating the map of their responses, and subtracting
this map from the Bouguer map. If the size, shape, depth, and density of
all the basement anomalies were known, a map of their gravity responses
could be calculated and subtracted from the Bouguer map, thus creating
a residual map which contains all the sedimentary anomalies with no dis-
tortion. Unfortunately, in the Gulf Coast we are a long way from having
adequate information concerning the basement structures that produce
gravity anomalies. Therefore, we use other methods, of lesser accuracy, to
create residual maps.
Profile residuals
Our profile residual map is produced by plotting profiles at equal intervals
across the Bouguer map in north-south and east-west directions. A curve
that matches the interpreted regional anomalies is fitted to each profile
and all the curves from the intersecting profiles are fitted, adjusted, and
combined to create a regional map. Values from the regional map then are
subtracted at each gravity station from the Bouguer map to create the
residual map. The pitfall of this method is in determining the exact shape
of the regional anomalies. In the Gulf Coast, we are fortunate in that the
basement over the majority of the area is deep and its resulting anomalies
are smooth, broad, and of very long wavelengths. Even in the areas where
the basement is shallow, such as the Jackson Dome and the Sabine Uplift,
the basement anomalies appear to be broad and smooth.
Even though there is some overlap between sedimentary and base-
ment anomalies, a profile residual constructed by an interpreter who is
knowledgeable in the regional geology of an area can produce a reliable
residual map. This map will consist of sedimentary anomalies having a
minimum distortion of shape and amplitude. We have found these resid-
ual maps ideal for modeling salt and other structures.
Wavelength filtering and ring-residual maps
Prior to the use of digital computers, gravity interpreters calculated resid-
ual maps by the ring-residual method. The process was slow and labor-
intensive. The Bouguer map was gridded on a square grid and a value was
recorded at each grid point. The interpreter then placed a template of
rings around each grid point and read the gravity values around each ring.
The values of each ring were averaged and the difference between this
02 General Planning 4/16/07 4:02 PM Page 11

average value and the value at the center grid point was recorded as the
residual value for the grid point. This method is sensitive to the dimen-
sions of the circles used. For example, a poor choice of grid spacing will
distort the residual map by either filtering out small anomalies or by
smoothing several small anomalies into a single large anomaly.
With the advent of the computer, many gravity interpreters adopted
the ring-residual method and renamed it wavelength filtering. Many
times, grid points are created by computer gridding programs in areas
where there are very few or no gravity stations. In some cases, the result-
ing residual maps are produced without the original gravity-station loca-
tions shown. Such maps contain false or badly distorted anomalies subject
to misinterpretation.
By using several rings and weighing factors, interpreters can apply the
ring-residual (wavelength-filter) method to more complex calculations
such as second vertical derivative and upward continuation techniques.
Swartz (1954) used 2-D Fourier transforms to study the filtering effect of
ring-residual operators. He illustrated how this or any other wavelength
filter which uses circular operators distorts the map by amplifying round
anomalies and suppressing elongated ones (Figure 3). In other words,
although these types of residual maps sort the anomalies according to
Figure 2a.
wavelength, they distort the shape and amplitude of the anomalies
according to shape. As a result, many explorationists have misinterpreted

Figure 2b.

Figure 2.

11
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:02 PM Page 12

General/Planning/Large Scale
12
General Philosophy 1
Gravity and Magnetic
Interpretation
edited by David A. Chapin
LaCoste & Romberg
Austin, Texas, USA

The following represents an informal on-

line discussion held on the gravity and
magnetic Internet mailing list (grvmag-l).
I have taken the liberty of reorganizing
and categorizing the different postings in
a slightly different order than they
appeared in the discussion. I also have
made an attempt to highlight summary
statements that distill a particular con-
cept or theme.

I would like to thank all the participants

in the discussion for sharing interesting
insights into the modern interpretation
process as applied to gravity and magnet-
ics data.

Figure 3. From Swartz (1954). General Philosophy

One philosophy I use for interpretation is
the process-of-elimination approach. I
these ring-residual and wavelength-filtered maps and have become disen- References like to use this approach not just for
chanted with the use of gravity because the distorted anomalies do not Dean, W., 1958, Frequency analysis for gravity and magnetic interpretation: interpretation but for processing as well.
relate properly to the geologic facts. We are concerned that this problem Geophysics, 23, 97–127. The idea is very simple: Remove every-
still exists for explorationists who use computer-derived wavelength-fil- Swartz, C., 1954, Some geometrical properties of residual maps: Geophysics, 19, thing you know about the observed data.
tered maps of the type described above. For additional information of fre- 46–70. What is left is what you don’t know,
quency analysis of potential-field data, see Dean (1958). Talwani, M., and Ewing, M., 1960, Rapid computation of gravitational attraction of which is often what you are searching
three-dimensional bodies of arbitrary shape: Geophysics, 25, 203. for.
Conclusions
Additional References It goes something like this for gravity
Although most of the gravity data in company files and for sale is more
Greene, E., and Bresnahan, C., 1987, Three-dimensional gravity modeling of Gulf data:
than twenty-five years old, the maps of these data can play a very impor- Coast salt domes: Oil and Gas J., October 12, 64–69.
tant role in exploration. In conjunction with subsurface and seismic data 1) you acquire data
Nettleton, L., 1971, Elementary gravity and magnetics for geologists and geophysicists:
and through the use of 3-D modeling, gravity can make a significant con- 2) you remove the global earth field
Soc. Expl. Geophys.
tribution to the mapping of both the top and base of the Louann Salt, its from the data (something you know
many complex salt domes, ridges, and other structures. The proper use of about)
gravity data will lower the cost and improve the success ratio of explo-
ration for oil and gas.
02 General Planning 4/16/07 4:02 PM Page 13
3) it is not correct at your elevation, so
you correct for that (something you
know)
4) you are left with Bouguer gravity
5) it contains crustal anomalies from
deeper than your target depth
6) you correct it for isostatic effects or
some other long-wavelength approach
7) you are left with a “residual”
anomaly
8) you model the anomaly for all the
geology you know; typically, these
are the shallower layers
9) you subtract that from the residual
10) what is left over is what you don’t
know about, presumably at depths
of exploration interest
Very often, it is precisely these signals
that identify undiscovered geologic struc-
tures of exploration interest.
I’m not saying that I always do the
above, but it is one approach I often
take. Notice that various parts of the
process go under different names: refer-
ence field correction, isostatic correction,
stripping, etc. But to me, it is all part of
the same process, the same overall phi-
losophy. We are basically subtracting
known components and examining what
is left over.
David Chapin
LaCoste & Romberg
Austin, Texas, USA
The layman may need to take one step
back, so to speak, to understand that the
gravity field (curve) is a summation of
gravity effects (anomalies) derived from
everything from the Earth’s core to the
gravimeter (and in the case of terrain,
around the gravimeter—a concern for
processing).
As an interpreter, our task becomes one
of picking this curve apart into anomalies
that we “understand” (control) and
anomalies that we want to learn about.
Often, that process is not as difficult as it
sounds, as long as the interpreter is care-
ful and uses all the available information
(well control, seismic, magnetics, out-
crop geology, etc.).
A step-by-step process is an excellent
approach. It is also helpful to keep
reminding yourself of what geologic
question you want to answer with your
analysis. Why am I doing this? What is
the throw on a fault? What is the struc-
tural closure? What is the mass of salt? Is
there a basin? Is there a reef?
Then ask the questions: What do I
know? What do I think I know? Then by
modeling, filtering, direct correlation,
etc., start attributing anomalies to those
things that you “know.” If you are care-
ful, you soon will get to a set of residual
effects that can be “tested” (modeled) so
you can develop your conclusions.
I have found that interpreters can
become so involved with the computer
tasks (i.e., building the model, running
the suite of filters, applying interesting
new algorithms) that they may lose sight
of the original geologic question. The
stated problem was not to build a model
or achieve a certain degree of “fit”
between observed and calculated data.
The original question was geologic, so
reminding ourselves of that question is
necessary. The computer tasks are a way
of developing products to help us con-
clude answers to the geologic question
with a certain degree of confidence.
Geologic Applications of Gravity and Magnetics: Case Histories
The objective is to interpret the geology,
not the gravity/magnetics. The potential
fields are only an expression of some-
thing geologic.
Pat Millegan
Worldwide Exploration Services
Marathon Oil Company
Houston, Texas, USA
Meters, not mGals, are my working
approach. I would rather tell people that
there is an interpreted structure of X
meters wide at a depth of Z meters deep
then tell people that there is a 20-mGal
anomaly at a certain place. Of course,
you need to hedge a bit by putting error
bars on that estimate.
David Chapin
It has been stated without rebuttal that
Bouguer and isostatic “corrections” are
really interpretations. I agree completely,
and would argue that in many cases the
interpreters (clients?) should be involved
early on when important interpretation
decisions are made.
Once you have “interpreted” and re-
moved what you think you know, we
are left with a “residual” anomaly.
Then I commonly use two complemen-
tary methodologies:
1) interpretation by inspection
2) interpretation by modeling
We always should keep the geologic
question in mind during our processing
and interpretation, choosing the appro-
priate techniques to resolve the geologic
issue. For example, it may be that
detailed 3-D modeling is not required to
distinguish between end-member geo-
logic styles.
By “interpretation by inspection,” I mean
looking at the profiles and maps to dis-
cern anomalies and features of interest
and then noting correlations with other
geologic and geophysical observations.
Where are the sediment fairways? Where
are major basement structures? How do
faults correlate between regional seismic
lines? Sometimes this type of interpreta-
tion is sufficient to answer the explo-
ration questions being raised. It may
require some modeling to be able to rec-
ognize the potential-field signatures of
various structural styles. Filtering and
display techniques are important tools
to highlight the features of interest.
In summary: Attempt to draw empirical
relationships between the potential-fields
data and all other geologic or geophysical
data.
“Interpretation by modeling” can be per-
formed at various levels of detail, but to
me the key issue is to use a “minimalis-
tic” method. I try to choose the simplest
modeling technique that will lead me to
distinguish between the possible geologic
scenarios under consideration. I also
would include depth-estimation methods
in this category. In short:
Make some quantitative measures through
modeling of the geologic problem.
Bayes’s theorem is an important and use-
ful tool when applied to interpretations
from either category. It is especially use-
ful because it helps to quantify risk and
the uncertainties in the interpretation. As
an interpreter, I can’t recall a single time
when I have been able to say with cer-
tainty that the Earth is as modeled.
However, it is usually possible to assign
some probabilities to different scenarios.
Those probabilities generally are trans-
lated into potential volumes or are used
in exploration decision trees. Knowing
that different interpretations are equally
probable is important information.
That brings up the importance of under-
standing our error budget, not just for
data acquisition but also for interpreta-
tion. It is very important to understand
the distribution of errors in the observa-
tions we are modeling, but it is just as
important to understand the variation of
geologic parameters we think we know.
If you are stripping off gravity signatures
of the geology between the observation
surface and the target of interest, you
had better know the range of variation
and distribution patterns of those inter-
vening densities (and possibly deeper
ones as well). Parameter variation, model
resolving power, and ideal body analysis
are important tools, not to be over-
looked.
In the context of the observational and
interpretational error budgets, uncer-
tainty of the solution can be assigned
probabilities.
Understand how errors can affect the
interpretation.
E. K. Biegert
International Center for Remote Sensing
and Potential Field Studies
Shell
Houston, Texas, USA
13
02 General Planning 4/16/07 4:02 PM Page 14
Gulf of Mexico Continental Slope—
Understanding the Magnetic Response
Due to the Salt Intrusion
Corine Prieto
Integrated Geophysics Corporation
Houston, Texas, USA
EDITOR’S NOTE: This paper provides a perspective on the resolv-
ing power of high-sensitivity magnetic data in prospective
areas of complex salt tectonics. Prieto attacks the geologic
problem with simple preliminary models. In this way, the
interpreter can predict and decipher the natural anomalies
with greater confidence. Try this in the Gulf of Mexico—it
works!
Introduction
It has been estimated conservatively that 15% of the U. S. domestic oil
production and 17% of its natural gas production come from fields off-
shore Louisiana along the continental shelf margin. It is this past produc-
tivity that has led the industry to enter the deeper waters of the con-
tinental slope to explore for future reserves. Exploration is a challenge
because of the extent of the untested area and its geologic and geophysi-
cal complexities.
The objective of this footnote is to illustrate the enhancement to
exploration efforts by the adjunct of magnetic data. Understanding the
cause and character of the magnetic response due to the salt/sediment
interface will lead to the integration of the magnetic data with other geo-
physical methods. The integration process should lead to a more accurate
geologic solution of an exploration problem.
Magnetic Susceptibility
The magnetic responses of salt structures in the Gulf typically are charac-
terized by small, low-amplitude, negative magnetic anomalies. Although
the full understanding of the physics of the signature is being debated
within the industry, these responses are identifiable in modern high-sen-
sitivity total-magnetic-intensity field data. Figure 1 is a sample of a 1986
aeromagnetic data set over a small area along the flexure trend. The high
accuracy of the data permits contouring at an interval of 0.5 gammas.
General/Planning/Large Scale
14
The lithology of the Tertiary and Quaternary sedimentary section of
the Gulf of Mexico consists of sands, shales, and salt (halite). The mag-
netic signature of these rock types is dependent on their relative content
of magnetic minerals, principally magnetite. The volume of magnetite
present in any given rock determines its level of magnetic susceptibility.
The susceptibility gives a measure of a rock type’s potential to produce a
magnetic response. The salt (halite) is known to have diamagnetic prop-
erties (see Nettleton, 1976) having a magnetic susceptibility (k) average
equal to –0.8 × 10–6 cgs units. An average magnetic susceptibility (k) of
shale is 51 × 10–6 cgs units. Sandstone has exhibited an average suscepti-
bility of 35 × 10–6 cgs. The following table summarizes these average
magnetic susceptibilities and lists the number of samples used to derive
the average. The number of samples is considered to be extremely small.
Magnetic Susceptibility Table
Rock Type No. of Samples Average k × 10–6 cgs
Salt 6 –0.8
Shale 142 51
Sandstone 240 35
Magnetic Anomaly
The observed high-frequency magnetic response is more than likely a
combination of physical effects, for example, the negative response due to
the salt and the positive responses due to the shales and sandstones. The
main contributor is the shale sequence because of its known higher mag-
netite content. The significant magnetic anomalies observed in the Gulf
are negative. It is, therefore, the opinion of the author that we primarily
are observing in the data a “lack of magnetite” effect, plus a small nega-
tive contribution from the diamagnetism when the salt replaces the shale
or sand. The lack of shales resulting from a salt intrusion produces the
magnetic susceptibility contrast.
Magnetic Signature
A set of characteristic curves has been computed to illustrate the type of
magnetic response generated by a few salt structures commonly seen
along the continental slope. The magnetic susceptibilities used in the
computations are k = 80 × 10–6 cgs for the sediments, and k = 0.0 × 10–6
cgs for the salt. All the calculations were done using a magnetic inclina- Figure 1. Sample of a 1986 aeromagnetic data set over a small area along the flexure
tion of 90°N. The characteristic curves are intended to lay the foundation trend.
for understanding the geologic reasons and/or geometrical configurations
02 General Planning 4/16/07 4:02 PM Page 15

of the structures that produce the magnetic response generated by the

salt/sediment interface.
The first series is a salt dome shown on Figures 2a and 2b. The top of
the dome is at a depth of 1219 m (4000 ft) below the surface of the water.
It represents a dome with a diameter of 1524 m (5000 ft). At a depth of
3658 m (12 000 ft), the salt mass flares to a diameter of 3048 m (10 000
ft) at a depth of 6096 m (20 000 ft). The Figure 2a computation produces
a 6-gamma response when the dome is attached to the mother salt. The
detached salt dome (Figure 2b) exhibits a 30% reduction in the ampli-
tude, a 4-gamma response. The question of detachment is a recurring
exploration problem in the continental slope. The top of a salt dome can
be identified from seismic data. The integration of the magnetic data can
resolve the possibility of detachment and, with structural modeling, can
determine a depth envelope of the detachment.
A second series is presented in Figures 3a through 3c, representing a
salt ridge. The depth to the top of the salt mass is shown at various
depths—914 m (3000 ft) (Figure 3a), 2438 m (8000 ft) (Figure 3b), and
3962 m (13 000 ft) (Figure 3c). The salt mass is shown as having a con-
stant width of 3048 m (10 000 ft) to a depth of 4267 m (14 000 ft). The
width then begins to flare out to 6096 m (20 000 ft) at a depth of 6096 m
(20 000 ft). The length of the ridge (distance in and out of the plane of Figure 2a. Magnetic response of an attached salt dome. Figure 2b. Magnetic response of a detached salt dome.

Figure 3a. Magnetic response of a salt ridge at a depth of 3 kilofeet. Figure 3b. Magnetic response of a salt ridge at a depth of 8 kilofeet. Figure 3c. Magnetic response of a salt ridge at a depth of 13 kilofeet.

15
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:02 PM Page 16
Figures 3a through 3c) is 15 090 m (50 000 ft). The positions of the
inflection points of the anomaly correspond to the edges of the ridge. The
amplitude of the anomaly is attenuated, with depth varying from 13.0
gammas at 914 m (3000 ft) to 5.5 gammas at 2438 m (8000 ft) and to
2.0 gammas at 3962 m (13 000 ft).
There are significant salt masses in the area that exhibit three-dimen-
sionality (i.e., width = length). The depth, lateral extent, and geometry of
these salt masses are in constant debate among explorationists. The third
series, Figures 4a through 4c, exhibits a salt mass at a depth of 1829 m
(6000 ft). Its diameter is equal to 6096 m (20 000 ft). Figure 4a shows a
symmetrical salt dome that produces a symmetrical anomaly with an
amplitude of 10 gammas. The second example, Figure 4b, shows a
“rolled-over” salt dome. The areal extent at 1829-m (6000-ft) depth is the
same as that shown on Figure 4a. The salt dome is attached at a depth of
6096 m (20 000 ft) but with a width of only 1219 m (4000 ft). The geom-
etry of the salt mass is asymmetrical. The anomaly produced is asymmet-
rical with a significant reduction in amplitude (6.5 gammas) from the
previous example. The third of the series, Figure 4c, shows the salt dome
with a variation of its topography. An 1829-m (6000-ft) graben has been
cut across the top of the dome. Resolving the high-frequency positive
Figure 4a. Magnetic response due to a salt structure 6000 feet below the water surface.
General/Planning/Large Scale
anomaly superimposed on the broad negative anomaly is important
because of its repetitive observed occurrence in the Gulf.
It should be noted that these calculations have used a susceptibility
value based on (textbook) preliminary data. Since 1986, several oil and gas
companies have been collecting magnetic susceptibility measurements
across the Gulf. Structural modeling for prospect support should include
these measured values for magnetic susceptibility and its regional variations.
The model calculations shown in Figures 2 through 4 are intended to
illustrate the resolving capabilities of the magnetic data. These character-
istic curves are not the total extent of the contributing information that
can be obtained by investigating the anomaly signatures of geologic struc-
tures and integrating them with other geophysical and geologic data sets.
Conclusion
This brief discussion is intended as an introduction to the resolving capa-
bilities of aeromagnetic data along the continental slope. The quality of the
aeromagnetic data presently available is sufficiently high to allow the
magnetic response caused by the salt/sediment interface to be interpreted.
The interpretation of the salt tectonics of the area can be enhanced with
integration of the magnetic data.
Figure 4b. Magnetic response of a rolled-over salt dome at a depth of 6 kilofeet.
16
The characteristic curves calculated to illustrate the origins of the
magnetic response are only a few from an extensive library. The inclusion
of such response curves or structural models in our exploration scheme
should assist in resolving some of the geologic problems that seismic data
indicate are occurring along the slope. For example, the configuration
of shallow diapiric structures can be enhanced with the addition of the
magnetic data. The rock-composition question also should be resolvable.
The reader must take note that not all aeromagnetic data flown to
date will have the resolution to determine a salt/sediment interface.
Special acquisition criteria must be met. Yet acquiring and interpreting the
aeromagnetic data are still very cost-efficient.
Acknowledgments
Special thanks to Phil L. Lawrence for his critique and to Geoterrex, Ltd.,
and Jeff Rowe for the aeromagnetic data example.
Reference
Nettleton, L. L., 1976, Gravity and magnetics in oil prospecting: McGraw-Hill Book Co.
Figure 4c. Magnetic response of a salt dome with varing topography.
02 General Planning 4/16/07 4:02 PM Page 17
Additional References A Brief History of
Editorial staff, interview, 1993, International Offshore, the Energy Group of PennWell Micromagnetic Surveying
Publishing Co.
J. P. Land, consultant
Lafayette Geological Society, 1973, Offshore Louisiana oil and gas fields, October. Houston, Texas, USA
Paine, W. R., and Meyerhoff, A. A., 1976, Gulf of Mexico Basin: Interactions among Micromagnetics is the petroleum-
tectonics, sedimentation, and hydrocarbon accumulation: Gulf Coast Associated
Geological Societies, 20, 4–44.
exploration form of magnetic surveying
for recording and interpreting all fre-
Talbot, C. J., and Jackson, M. P. A., 1987, Internal kinematics of salt diapirs: AAPG quencies of the magnetic field, from the
Bull., 71, 1068–1093.
shortest wavelength anomalies sourced
Woodbury, H. O., et al., 1973, Pliocene and Pleistocene depocenters, outer continen- in the shallowest sedimentary formations
tal shelf, Louisiana and Texas: AAPG Bull., 57, 2428–2439. to the longest wavelength anomalies
sourced at or beneath the basement sur-
face. This is accomplished by using hand-
held magnetometers on the ground or by
using airborne magnetometers flown at a
terrain clearance in the order of 150 m
(500 ft). The objective is to map all mag-
netic indications of structure from the
top of the sedimentary column to crys-
talline basement.
The earliest magnetometers allowed
measurement accuracies of 10 to 25 nT
(gammas). In 1936, W. P. Jenny wrote
that the method had been “revolution-
ized” by improvements in the instru-
ments and in field techniques that
allowed accuracies of ± 2 nT. Magne-
tometers involved in the early mapping
of sedimentary responses were mechani-
cal, tripod-mounted, magnetic balances
that measured either the vertical or hori-
zontal component of the Earth’s field.
These and later mechanical devices,
eventually improved to allow accuracies
of ± 1 nT, were the mainstays for surface
work until the coming of the proton
magnetometer in the 1960s, with its
1-nT sensitivity that eventually was
improved to 0.1 nT.
In 1948, the Gulf Airborne Fluxgate
Magnetometer, developed during World
War II for antisubmarine warfare,
became available for commercial use.
Geologic Applications of Gravity and Magnetics: Case Histories
This tool, which allowed the rapid inven-
tory of mineral resources regardless of
terrain and accessibility, caused a revolu-
tion in reconnaissance surveying world-
wide. The Mark II Gulf instrument, with
its analog data and noise envelope in the
order of 0.25 nT, was the primary tool
until the digitally recorded, optically
pumped, cesium-vapor magnetometer
arrived in 1962. Its sensitivity and sam-
pling rate have been improved over the
years to .001 nT and .01 second. In air-
borne work, an important factor in
improving data quality has been the
improvement in flight-path identification,
from the snapshot photography tech-
nique patented in 1952 by Homer Jensen
of Aero Service Corp. to the current digi-
tal recording of each sample point’s coor-
dinate using the Global Positioning
System (GPS) satellite network, in time
synchronization with video flight-path
recovery.
Perhaps the first printed mention of the
term micromagnetics was in the W.P.
Jenny article (1936) in which he stated:
The basic assumptions for the interpre-
tation of magnetic anomalies have been
changed completely. In former years we
thought that all magnetic anomalies
were caused by the basement or shallow
igneous rocks. Today we know that
there exists sufficient difference in the
magnetic susceptibility within the sedi-
mentary columns to produce noticeable
magnetic anomalies above structures of
commercial interest.
In the years that followed, Jenny carried
out many ground surveys and published
many articles on the subject of micro-
magnetics. When the airborne magne-
tometer came on the scene, he became
perhaps airborne micromagnetics’ most
ardent advocate and practitioner. In
1953, he began an airborne micromag-
netic survey of the Frio-Miocene Trend of
the Texas and Louisiana Gulf Coast, a
program of 45 000 square miles that pro-
duced a unique perspective of the region’s
shallow and deep structural grain. Most
importantly, it led to local structures that
then were investigated systematically,
using subsurface geology and appropriate
surface methods in a very successful
development of drilling prospects.
The next major point in the history of
micromagnetics came in 1979 with T. J.
Donovan’s observations of sedimentary
magnetic responses coincident with geo-
chemical alteration of surface formations
and production at the Cement Field in
Oklahoma. A shallow subsurface zone
was found by drill-cutting analysis to
have a higher magnetic susceptibility
than the other remaining sediments. His
basic assumption, that the upward
microseepage of hydrocarbons from a
reservoir causes geochemical alteration
in the overlying sediments, has been doc-
umented subsequently. We now know
that depending on whether the magnetic
minerals thus created are magnetic or
nonmagnetic, positive or negative mag-
netic responses will be the measurable
end product.
At this point in the history of micromag-
netics, we know that when near-surface
sediments are displaced structurally, a
measurable magnetic anomaly will be
created. Such an anomaly may be, in
part, the result of geochemical alteration,
but magnetics alone will not tell us
whether hydrocarbons are trapped in the
underlying sediments. Only the drill can
establish that. Micromagnetics is a recon-
naissance tool that can resolve local
anomalies having the potential of being
hydrocarbon related.
17
02 General Planning 4/16/07 4:02 PM Page 18
So that the interested individual can
investigate the subject beyond this brief
history, I have selected a bibliography
that follows.
References
Donovan, T. J., Forgey, R. L., Roberts, A. A.,
1979, Aeromagnetic detection of diagenetic
magnetite over oil fields: AAPG Bull., 63,
245–248.
Jenny, W. P., 1936, Micromagnetic surveys: Gulf
Coast structures may be outlined by this new
method: The Oil Weekly, 81, 23–32.
Jensen, H., 1961, The airborne magnetometer:
Scientific American, 204, 151–152.
Additional References
Andrew, J. A., Edwards, D. M., Graf, R. J., and
Wold, R. J., 1991, Empirical observations
relating near-surface magnetic anomalies to
high-frequency seismic data and LandSat data
in eastern Sheridan County, Montana:
Geophysics, 56, 1553–1570.
Donovan, T. J., 1974, Petroleum microseepage at
Cement, Oklahoma—evidence and mecha-
nism: AAPG Bull., 58, 429–446.
Flanagan, G., Davis, S. G., Campbell, C. L., and
Doughtie, J. K., 1988, Integration of high-sen-
sitivity magnetics and seismic data to define
salt and sediment structures in the Gulf of
Mexico: 58th Ann. Internat. Mtg., Soc. Expl.
Geophys., Expanded Abstracts.
Foote, R. S., 1987, Correlations of borehole rock
magnetic properties with oil and gas produc-
ing areas: Assoc. of Petroleum Geochemical
Explorationists Bulletin, 3, 114–134.
Foote, R. S., and Long, G. J., 1988, Correlations
of oil and gas producing areas with magnetic
properties of the upper rock column, eastern
Colorado: Assoc. of Petroleum Geochemical
Explorationists Bulletin, 4, 47–61.
Friedberg, J. L., 1976, The two-gamma fault:
46th Ann. Internat. Mtg., Soc. Expl. Geophys.,
Expanded Abstract.
Gay, S. P., and Hawley, B. W., 1991, Syngenetic
magnetic anomaly sources, three examples:
Geophysics, 56, 902–913.
Henry, W. E., 1989, Magnetic detection of hydro-
carbon microseepage in a frontier exploration
region: Assoc. of Petroleum Geochemical
Explorationists Bulletin, 5, 18–29.
Jenny, W. P., 1932, Geophysical prospecting in
Gulf Coast district had to overcome many dif-
ficulties: Oil and Gas J., 31, 14–16.
——— 1935, Some practical examples of mag-
netic prospecting: Oil and Gas J., 33, 49,
33–35.
——— 1941, Geological and geophysical profiles
through the Eola Field, Louisiana: Oil and Gas
J., 40, 18, 42–44.
——— 1942, Micromagnetic surveys in the
Sparta-Wilcox trend of Texas and Louisiana:
The Oil Weekly, 106, 2, 20–23.
——— 1952, Structural correlation of micro-
magnetic and reflection surveys: World Oil,
134, 4, 67–72.
——— 1961, High aeromagnetic accuracy pro-
vides detailed coverage: World Oil, 153, 4,
90–92.
——— 1961, Regional magnetic data show
prospective trends: World Oil, 153, 5,
148–149.
——— 1961, Many old, updated magnetic
prospects prove to be valid: World Oil, 153,
6, 124–130.
——— 1961, Aeromagnetics develop new
prospects and techniques: World Oil, 153, 7,
105–106.
——— 1984, Aerial micromagnetic profiles pin-
point local structure: Oil and Gas J., 82, 30,
79–81.
Land, J. P., 1989, The micromagnetic method: in
Advanced surface exploration, a short course:
Soc. Indep. Prof. Earth Sci.
——— 1991, A comparison of micromagnetic
and surface geochemical survey results:
Association of Petroleum Geochemical
Explorationists, 1, 12–35.
——— 1994, Val Verde basin study integrates
gravity, magnetic, geologic data: Oil and Gas
J., 92, 43, 72–79.
Schumacher, D., Magnetic anomalies associated
with oil and gas fields: Unpublished.
General/Planning/Large Scale
Wold, R. J., 1984, Aeromagnetic detection of
anomalies produced by chemical alteration
over oil-gas deposits: TerraSense, Inc.
High-resolution
Aeromagnetic Surveying
Patrick S. Millegan,
Marathon Oil Company
Houston, Texas, USA
I would like to make a distinction
between the application of state-of-the-art
aeromagnetics for geologic interpretation
and micromagnetics as a hydrocarbon
direct-detection method. Very few people
promote micromagnetics as a direct-
detection method anymore. High-resolu-
tion aeromagnetics (HRAM) for struc-
tural interpretation, on the other hand, is
enjoying something of a renaissance. Not
only did magnetic methods survive last
decade’s industry slump, but they actu-
ally advanced through the period.
HRAM is descriptive of an exploration
methodology targeted at very detailed
high-resolution interpretation of the full
geologic section. HRAM surveys are
standard today because of an increased
interest during the 1980s in sampling
high-frequency, low-amplitude magnetic
anomalies that many people thought
were direct hydrocarbon indicators.
The magnetic field was sampled at a
1
⁄10-second increment, from an altitude
of 80–150 m, along lines spaced 250–
500 m apart. The striking new data
shook conventional wisdom, and caused
a reevaluation of rules of thumb. Today,
aeromagnetic contractors are busier
than ever.
The conventional use of aeromagnetics,
to map basement structure, has had to
make room for new concepts. An explo-
rationist now thinks of using high-
resolution aeromagnetics when:
1) the target is subtle (salt)
2) the target is intrasedimentary
(faulting, alteration minerals), or
3) there is a need to detail a geologic
structure (prospect analysis, produc-
tion geophysics)
As early as the 1930s, Dr. W. P. Jenny
proposed the use of land magnetic sur-
veys to interpret sedimentary structure.
Sediments contain enough magnetic min-
erals to cause subtle magnetic anomalies.
Contrasts (structural offsets, usually
faults) shallow in the section cause
higher frequency, lower amplitude anom-
alies than the deeper, more intense base-
ment effects. The key to the concept is
the acquisition of the “right” data.
Terry Donovan’s controversial papers
from the 1980s on Cement Field and the
Arctic National Wildlife Refuge (ANWR)
fanned the flames of interest in direct
hydrocarbon indicators. But nothing
comes easily. Later work by Rich
Reynolds, Brooks Ellwood, Stuart Hall,
Doug Elmore, and others details the
complexity of the geologic problem.
Anomalies from near-surface detrital
magnetite (abandoned stream channels
or glacial till) can be mistaken for hydro-
carbon alteration effects. Magnetic min-
erals can be the result of geochemical
oxidation, completely unrelated to hydro-
carbons. Alteration minerals are short-
lived. Hydrocarbons may even destroy
magnetic minerals. And these do not
even address the issues of migration
paths and seals.
18
But all is not lost. The remains of hydro-
carbon-eating magnetotactic bacteria
exist in the sedimentary section.
Authigenic magnetic minerals are docu-
mented, and in sufficient quantity to
cause measurable anomalies. Can these
anomalies be “captured” and interpreted
with confidence?
The ability to interpret sedimentary
structures, salt, or possible hydrocarbon
indicators is highly dependent on:
1) an understanding of the near-surface
geology and mineralogy
2) controlling or editing cultural noise
3) a strong follow-through integration
with the deeper subsurface geology
Confidence is gained if these more subtle
low-amplitude, high-frequency anomalies
coincide with the more “classic” signa-
tures of deeper geologic structures as
determined from gravity, magnetics, or
seismic interpretations. It is better to
think of micromagnetics in terms of geo-
logic interpretation rather than depend-
ing on them as direct hydrocarbon
indicators.
The exploration community has dis-
tanced itself from the direct-detection
concepts of micromagnetics in favor of
the concept of HRAM’s high-resolution
geologic interpretation. It is clear that
the densely spaced lines and broad fre-
quency spectrum offer more possibilities
for the interpreter. In many ways, the job
of interpretation is harder. Previously
interested only in crystalline basement,
the interpreter now evaluates anomalies
that come from the sedimentary section.
Have no fear, deep, basement-related
structures are imaged very well in
02 General Planning 4/16/07 4:02 PM Page 19
HRAM surveys targeted at the sedimen-
tary section. The reverse may not be
true, however. Conventional high-alti-
tude, wide-line-spacing surveys have a
difficult time confidently imaging subtle
effects. Hence, HRAM surveys are now
the norm.
Whatever the motivation for surveying,
the exploration industry has a new gen-
eration of exciting magnetic maps. The
challenge for this new world of aeromag-
netic interpretation is the careful and
sometimes tedious analysis of quality, tar-
geted data. Valid geologic conclusions are
necessary. The story about a contractor’s
magnetic residual (band-pass-filtered)
map being labeled as “the Miocene” is a
true one. It was the result of the correla-
tion of one fault (active during the
Miocene) from seismic with one linear
trend in the filtered magnetic map. This
is not good practice on any level!
Interpretation must now stay up with, or
catch up with, these exciting new data.
HRAM Case Histories
in this Volume
Prieto’s first paper in this first section of
case histories deals with the magnetic
response of salt bodies in the Gulf of
Mexico. Being diamagnetic, salt causes
subtle magnetic minima that aid struc-
tural evaluation. HRAM surveys designed
to detect these anomalies are especially
helpful since the gravity “nil zone” (zero
density contrast between salt and sedi-
ments) complicates the use of gravity and
seismic alone.
The Morecambe Bay case history by
Richard Morgan documents correlations
between known reservoir characteristics
(structural and mineralogical) and ter-
rifically interesting, subtle magnetic
anomalies isolated from an HRAM sur-
vey. It is followed by a case study of the
Sierra and Yoyo Reefs in northeastern
British Columbia, Canada. Peirce, Ebner,
and Marchand offer extremely valuable
advice on problems inherent to areas
with cultural noise. Following careful
editing, processing, and analysis, each
study confidently interprets sedimentary
sources of magnetic anomalies.
In the final section of this volume, Rowe,
Smith, Beattie, and Warren provide a dis-
cussion of a combined acquisition tech-
nique using electromagnetics and
HRAM. This paper’s strength is its inte-
grated approach for interpreting hydro-
carbon alteration effects, showing
provocative correlations for two produc-
ing fields. T. J. Donovan’s 1984 paper on
ANWR fueled much of the modern inter-
est in micromagnetics. Now in 1998,
Phillips, Reynolds, and Saltus offer pre-
liminary results from a magnetostrati-
graphic study of the Arctic National
Wildlife Refuge (ANWR).
References
Donovan, T. J., and Forgey, R. L., 1982,
Remanent magnetization and magnetic sus-
ceptibility of two oriented cores, Cement oil
field, Oklahoma—A reference data set: USGS
Open-file Report 82–848.
Donovan, T. J., Hendricks, J. D., Roberts, A. A.,
and Eliason, P. T., 1984, Low-altitude aero-
magnetic reconnaissance for petroleum in the
Arctic National Wildlife Refuge, Alaska:
Geophysics, 49, 1338–1353.
Additional References
Bazylinski, D. A., Frankel, R. B., and Jannasch,
H. W., 1988, Anaerobic magnetite production
by a marine, magnetotactic bacterium: Nature,
334, 518–519.
Cisowski, S. M., and Fuller, M., 1985, The asso-
ciation of strongly magnetized rocks and
hydrocarbon production in reduced facies sed-
iments prone to combustion metamorphism
[abs.]: EOS, 66, 867.
Geologic Applications of Gravity and Magnetics: Case Histories
Cisowski, S. M., and Fuller, M., 1987, The gen- Reynolds, R. L., Webring, M., Grauch, V. J. S.,
eration of magnetic anomalies by combustion and Tuttle, M., 1990, Magnetic forward mod-
metamorphism of sedimentary rock, and its els of Cement oil field, Oklahoma, based on
significance to hydrocarbon exploration: Geol. rock magnetic, geochemical, and petrologic
Soc. Am. Bull., 99, 21–29. constraints: Geophysics, 55, 344–353.
Eliason, P. T., Donovan, T. J., and Chavez Jr., P. S., Sager, W. W., 1986, Magnetic-susceptibility mea-
1983, Integration of geologic, geochemical, surements of metal contaminants in ODP Leg
and geophysical data of the Cement oil field, 101 cores: Proceedings, Initial Reports (Part
Oklahoma, using spatial array processing: A), of the Ocean Drilling Program, 101,
Geophysics, 48, 1305–1317. 39–45.
Jenny, W. P., 1936, Micromagnetic surveys: Gulf
Coast structures may be outlined by this new
method: The Oil Weekly, April 27.
Kilgore, B., and Elmore, R. D., 1989, A study of
the relationship between hydrocarbon migra-
tion and the precipitation of authigenic mag-
netic minerals in the Triassic Chugwater
Formation, southern Montana: Geol. Soc.
Am. Bull., 101, 1280–1288.
Lovely, D. R., Stolz, J. F., Nord Jr., G. L., and
Phillips, E. J. P., 1987, Anaerobic production
of magnetite by a dissimilatory iron-reducing
microorganism: Nature, 330, 252–254.
Machel, H. G., and Burton, E. A., 1991, Causes
and spatial distribution of anomalous magneti-
zation in hydrocarbon seepage environments:
AAPG Bull., 75, 1864–1876.
McCabe, C., and Elmore, R. D., 1989, The
occurrence and origin of late Paleozoic remag-
netization in the sedimentary rocks of North
America: Rev. of Geophysics, 27, 471–494.
Oehler, D. Z., and Sternberg, B. K., 1984,
Seepage-induced anomalies, “false” anomalies,
and implications for electrical prospecting:
AAPG Bull., 68, 1121–1145.
Reynolds, R. L., Fishman, N. S., Hudson, M. R.,
Karachewski, J. A., and Goldhaber, M. B.,
1985, Magnetic minerals and hydrocarbon
migration: Evidence from Cement (Oklahoma),
North Slope (Alaska), and the Wyoming-Idaho-
Utah thrust belt [abs.]: EOS, 66, 867.
Reynolds, R. L., Fishman, N. S., Wanty, R. B.,
and Goldhaber, M. B., 1987, Iron sulfide
minerals at Cement oil field, Oklahoma—
Implications for the magnetic detection of
oil fields: Geol. Soc. Am. Bull., 102,
368–380.
19
02 General Planning 4/16/07 4:02 PM Page 20
Gravity/Magnetic Signatures of Various
Geologic Models—An Exercise in
Pattern Recognition
Corine Prieto
Integrated Geophysics Corporation, Houston, Texas, USA
EDITOR’S NOTE: A gravity/magnetics interpreter often is asked for
a catalog of geophysical responses to geologic configura-
tions. The importance of understanding what anomaly to
“expect” from your geology cannot be stressed enough.
Introduction
In every geologic province, the physical parameters are unique. The
prospective anomaly signature associated with each individual area will
vary accordingly. The pattern recognition of an anomaly signature is of
vital importance to all interpretation sciences. The following discussion is
intended for geophysicists and geologists interested in obtaining a work-
ing knowledge of what type of signal pattern should be observed to inte-
grate gravity and magnetic data with seismic and geologic data for
development of exploration targets. The incorporation of the integration
process in exploration has been known to reduce risk and therefore
reduce the total cost of exploration.
The gravity and magnetic anomaly signature characteristics are results
of one or more physical parameters such as the configuration of the anom-
alous zone, density, velocity, and porosity contrasts, magnetic susceptibil-
ity contrasts, and the depth to the anomalous body. An excellent tool to
aid in pattern recognition is structural modeling. A simple collection of
geologic structures has been modeled and their calculated gravity and
magnetic responses are discussed to identify anomaly characteristics. The
collection is not intended to be a library of curves, but rather an exercise
in pattern recognition. The 2-D modeling scheme used an interactive
graphic system. The well-known Talwani algorithms (see references)
were used as the basis for the computations.
The density and magnetic susceptibility contrasts are based on data
from various geographic areas from the Superior Oil Company explo-
ration cases and from published literature. The structures of the models
have been kept simple to emphasize how depth, density, and susceptibil-
ity contrasts vary the anomaly characteristics.
General/Planning/Large Scale
20
The following discussion refers to the residual anomaly as the
observed gravity and magnetic field. This distinction is made early to
avoid confusion. The gravity and magnetic interpreter is aware of the
importance of a good geologically based first-order residual field. Another
clarification is that the magnetic calculations use the magnetic inclination
of 60°N, which is the average magnetic inclination for the United States.
It is important to note that the magnetic signal of a structure is highly
dependent on its geographic location on the earth. The interpreter ini-
tially should identify the magnetic signal caused by the structure in ques-
tion, i.e., the geographic location of the structure and effect of the
magnetic inclination on the magnetic anomaly due to the structure.
Faults
The standard recognized fault signature is a steep gradient, whether it is
on a gravity, magnetic, structure, or seismic time map. In the gravity case,
the gradient steepens as the fault becomes shallower, as illustrated in
Figure 1a. But note that the inflection point of the gradient is at the cen-
ter of the fault at all depths. This signature is valid for a sedimentary fault 1a
in which homogeneous beds are not substantially offset. A magnetic
anomaly would not exist in this case, since the basement was not
involved in the faulting and the sedimentary section was also assumed to
be nonmagnetic. When the basement or a magnetic sedimentary layer is
faulted, a magnetic anomaly does occur, as shown in Figure 1b. The
anomaly amplitude depends on both depth and magnetic susceptibility. In
certain exploration environments, the question of basement involvement
often is answered through the perusal of a gravity and magnetic map.
If the gravity and magnetic anomalies are positive over a questionable
seismic structure, the basement probably is involved.
A second case for a normal fault is shown in Figure 1c. A high-density
layer has been introduced into the geometry. It produces a gravity gradi-
ent as in the first case, but the gradient has increased because of the
increase in the density contrast of the layer which is faulted. The high-
density layer could represent a carbonate formation. The lower density
layers could represent a clastic sequence. In this case, the basement is
involved in the faulting; therefore, a magnetic signature is generated. The
model calculation shown is for a 60° magnetic inclination. Note that the
magnetic anomaly calculated appears as a negative anomaly. This results
from various geophysical parameters: (1) the geometry of the structure, 1b
the fault, (2) magnetic inclination, and (3) direction of the model traverse,
here 0° azimuth. Figures 1a and b.
The same fault model as shown in Figure 1c, with a low-density layer
substituted for the high-density layer, is shown in Figure 1d. A positive
gravity anomaly occurs when the low-density layer is offset completely by
02 General Planning 4/16/07 4:02 PM Page 21
the fault and is displaced by the higher density layers. The magnetic sig-
nature would remain as shown in the previous example (Figure 1c), since
the basement structure has not changed.
The three geologic cases described have distinct anomaly characteris-
tics that can be correlated to rock composition or structure or both. The
amplitude of the anomaly will vary as contrasts and depth vary, but the
signature characteristics for each case should remain the same.
Dipping Beds
A dipping bed produces a gravity anomaly similar to that of a fault. The
main difference is that as the bed nears the surface, the gradient of the
anomaly becomes steeper and the magnitude of the anomaly tends to be
greater. The structural model calculation of the gravity effect as shown in
Figure 2 uses density contrasts which could represent the thrusting of
older beds over younger beds, or a dike intruding into a sedimentary
sequence. The edges of the denser bed can be associated with the inflec-
tions on the gravity curve. The highest amplitude on the curve occurs in
1c the direction of denser beds. The closer the dip of the beds to vertical, the
greater the symmetry in the gravity anomaly that will be observed. The
amplitude of the anomaly is dependent on the density contrast and depth.
If gravity data over an overthrusted bed and a constraint from seismic data
for the top of the overthrusted bed were available, structural modeling
of the gravity data would assist in resolving the thickness of the over-
thrusted section.
1d
Figures 1c and d.
Figure 2.
Geologic Applications of Gravity and Magnetics: Case Histories
Synclines
A syncline produces a minimum closure on most geophysical maps. The
amplitude and characteristics of the gravity and magnetic anomaly asso-
ciated with a syncline are dependent on (1) depth, (2) the type of sedi-
mentary fill, (3) the amount of sedimentary warping, and/or (4) the
involvement of the basement rocks. The magnetic inclination also plays
an important part in the character of the magnetic anomaly. In these dis-
cussions, the inclination has been kept constant at 60°N.
A folded sedimentary syncline with uniformly increasing density with
depth as shown in Figure 3a will produce a minimum gravity anomaly
(due to sedimentary warping) centered over the syncline; this model does
not produce a magnetic anomaly, since the sedimentary section is shown
as nonmagnetic. The basin model shown in Figure 3b illustrates the
gravity and magnetic anomaly characteristics that will identify a basin’s
extent and depth. The basement involvement increases the amplitude
and sharpness of the gravity signature, and it also generates a magnetic
anomaly. The areal or lateral extent of the basin or graben can be delin-
eated by identification of the gravity and/or magnetic basin-bounding
fault signatures. Again, remember that the magnetic signal is dependent
on its location on the globe. The broad minimum magnetic anomaly
shared by the two maxima situated at the edges of the basin would be
characteristic of the 60°N inclination. The basin would be expressed as a
minimum magnetic closure surrounded by higher frequency maximum
closures.
Figure 3a.
21
02 General Planning 4/16/07 4:02 PM Page 22

3b
3c
Figures 3b and c.
The faulted-folded syncline shown in Figure 3c is an example of a
more complicated geology. The gravity and magnetic signatures are dis-
torted, yet the signature characteristics remain. The gravity minimum is
still present, although it is offset over the upthrown side of the fault. The
General/Planning/Large Scale
magnetic signature is more complicated, but the basement faults are iden-
tifiable. The basin edges can be noted and the magnetic anomaly in the
center of the syncline indicates the central fault structure.
A syncline can often be masked on the seismic section by diffractions
or velocity problems. Assuming increased density with depth, the syncline
gravity signature will be a minimum anomaly; the magnetic signature will
occur only if basement rocks (assuming nonmagnetic sedimentary section)
are involved in the structure. It is rare that gravity and magnetic data can-
not verify a synclinal structure.
Anticlines
A simple folded symmetrical anticline produces a symmetrical positive
gravity anomaly. The amplitude and characteristics of the gravity and
magnetic anomaly associated with an anticline are dependent on (1)
depth, (2) the amount of sedimentary warping, and/or (3) the involve-
ment of the basement rocks. The magnetic inclination also plays an
important part in the character of the magnetic anomaly. In these discus-
sions, the inclination has been kept constant at 60°N.
As the geology becomes more complicated, so do the gravity and mag-
netic responses. So we depart from the simplest of structures and start the
discussion with a faulted anticline.
A normal-faulted anticline, as shown in Figure 4a, consists of a sedi-
mentary sequence of density values that increase with depth and a faulted
basement uplift. This structure produces a broad maximum gravity anom-
aly indicating the areal extent of the entire uplifted section. Two higher
frequency positive anomalies are superimposed on the broad anomaly;
one anomaly is associated with the surface outcrop of the fault and the
second anomaly is situated over the apex of the rollover. The minimum
between the two high-frequency closures is caused by the wedge of rela-
tively low-density material between the fault and the rollover. The mag-
nitude of this minimum anomaly is governed by the density contrasts and
the thickness of the wedge, but is an anomaly characteristic generally
observed on faulted anticlines.
The magnetic anomaly occurs only when the basement is involved in
the structure. The magnetic anomaly calculation shown in Figure 4a is for
an anticline with a basement structure. For comparison, the magnetic
anomaly calculation shown in Figure 4b is for an anticline without base-
ment structure. Note the individual fault signatures on the model in Figure
4b; the first one occurs at the inception of the upwarping of the anticline
on the left side of the model, the second fault is at the apex of the anticline,
and a third fault is at the downwarp of the anticline on the right side of the
model. The faults all vary in amplitude and wavelength, indicating diverse
22
structural characteristics. If the question is “What is the extent of basement
involvement in an exploration prospect structure?,” magnetic data gener-
ally can confirm basement and determine any secondary faulting.
The normal-faulted anticline previously discussed has been compli-
cated with the introduction of a high-density layer, as shown in Figure 4c.
The anomaly produced is similar to the previous model (Figure 4a). The
increase in density contrast sharpens the frequency of the anomalies
previously identified. The higher-density layer could represent a tight
limestone, limestone/anhydrite, or a volcanic sequence. The magnetic
anomaly would remain the same, as shown in Figure 4a, unless the high-
density layer was magnetic volcanic rocks.
The case model shown in Figure 4d illustrates the complexity in the
observed signatures caused by the introduction of a shallow (or surface)
high-density layer. This model could represent surface volcanics or a tight
limestone. If the local geology is not well known, magnetic data would dis-
tinguish clearly between the volcanics and the carbonates. The case shown
in Figure 4d has magnetic susceptibility assigned to the high-density layer,
which would suggest it to be volcanics. The magnetic anomaly caused by
the basement alone (Figure 4a) has been complicated with the introduc-
tion of the shallow magnetic sequence. If the shallow layer was dense car-
bonate, the magnetic signature would be as shown in Figure 4a. This
model is shown to illustrate that not all gravity- and magnetic-observed
anomalies are interpreted easily without structural modeling to assist the
interpreter in separating the signatures of the causative structures.
The final example of an anticline signature illustrates perhaps one of
the most frequently forgotten and most important aspects of gravity inter-
pretation. The gravity anomaly is a composite signal of all the density con-
trasts due to (1) the basement structure and (2) the sedimentary warping
of all the beds above the basement structure. To determine the total thick-
ness of the sedimentary section in a basin or to verify a structure for a
prospect, the total gravity contribution due to all the density contrasts
throughout the section to the observed gravity anomaly must be kept in
mind. When computing the gravity effect of a basement horst, it is impor-
tant to include the more subtle effects caused by the resultant sedimen-
tary warping above the horst. As shown in Figure 4e, when the warping
extends upward through several formations, the combined gravity effect
of each sedimentary layer to the total anomaly easily can exceed the base-
ment contribution.
The salt pillow (Figure 4f) is a common anticlinal model. It has a neg-
ative density contrast resulting from the sediment/salt interface, and
therefore a resultant minimum anomaly signature. The amplitude is
dependent on depth to the top of the pillow and thickness of the pillow.
02 General Planning 4/16/07 4:02 PM Page 23

4a 4b 4c

4d 4e 4f

Figures 4a, b, c, d, e, and f.

23
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:02 PM Page 24
4g
Figures 4g and h.
A point to note is that the positive gravity contribution due to sedimen-
tary warping over a salt pillow could cancel the negative signature. But
that fact in itself would assist in the interpretation of the salt structure.
Many seismic prospects are dependent on the existence of a salt structure
or a salt withdrawal. An integration of the seismic and gravity data could
provide an answer through the gravity calculation of the structural con-
figuration of the seismic geometry. For example, if the observed gravity
data have no anomaly where the seismic section indicates sedimentary
warping (or a potential structure), the introduction of a salt pillow would
be necessary to eliminate the positive anomaly resulting from the draped
beds; therefore, the gravity substantiates the seismic structure.
A case for a carbonate buildup is shown in Figure 4g. This example
represents a limestone reef embedded in a clastic sequence. The reef
would produce a positive gravity anomaly. The lateral extent of the reef
can be identified easily with the extent of the positive gravity anomaly. If
the top of the reef could be constrained, the thickness of the reef could
be modeled using gravity data.
The following discussion of characteristics that can be observed in
gravity anomalies is very dependent on the quality and coverage of the
gravity data available over the reef. The amplitude and shape of a gravity
anomaly vary noticeably because of porosity variations. The reef modeled
in Figure 4g is a symmetrical reef. It is shown with no porosity (blue
curve), which produces a symmetrical gravity anomaly. The gravity anom-
General/Planning/Large Scale
4h
aly due to a reef with 10% porosity filled with either oil or gas has been
reduced in amplitude, but a more important characteristic is that the
anomaly has been flattened, compared to the anomaly resulting from no
porosity. Many reefs produce from the fore-reef zones rather than uni-
formly from across the top. The character of the gravity anomaly will vary
depending on the position of the porous zone, as illustrated in Figure 4h.
The capability of distinguishing oil- from gas-filled porosity is difficult,
mainly because of acquisition problems. Strict field specifications must be
adhered to, and in many cases, the access to an area restricts the capabil-
ity of acquiring data for the repeatable 0.10-mGal data quality required.
Intrusives
Intrusives can be of many sizes and compositions and can occur in many
geologic provinces. Due to well-known diffraction problems, seismic sec-
tions reveal intrusives by the no-reflector zone they generate. Often, no use-
ful velocity information can be extracted to aid in the determination of the
rock composition of the intrusive. A solution to this problem can be deter-
mined by interpreting the gravity and magnetic signature of the intrusive.
Three typical intrusive compositions are salt (Figure 5a), shale (Figure
5b), and basalt or mafic igneous (Figure 5c). In our examples, all assumed
variables are constant except the density value due to the intrusive. It is
shown intruded into a typical sedimentary sequence where density
increases with depth, and where the sedimentary section has no signifi-
24
cant magnetic susceptibility value. The respective gravity signatures, due
to the intrusives, are significantly different. Salt produces a substantial
negative gravity anomaly and minimal magnetic anomaly (not shown).
Shale produces a minimum gravity anomaly, an order of magnitude less
than that produced by the salt, and no magnetic anomaly would be gen-
erated. The basalt or mafic intrusive would produce a significant positive
gravity and magnetic anomaly. The sharpness (frequency) of the anom-
alies due to the salt and the basalt will increase as the depth to the top of
the intrusive decreases or as the circumference of the intrusive decreases.
The model in Figure 5d illustrates the gravity anomaly produced by a
salt dome at various depths to the top of salt. The deepest salt column pro-
duces the broadest negative anomaly, shown here at a depth of 3.66 km
(12 000 ft). As the salt dome shallows, the anomaly increases in ampli-
tude and sharpness (frequency). The gravity effect due to the salt in the
crossover zone contributes nothing to the anomaly. Crossover or nil zone
is the area where the salt and sediment densities are equal (p = 2.2 g/cc),
therefore generating a zero density contrast. The crossover zone illus-
trated in Figure 5d is shown as 762 m (2500 ft) thick for emphasis. In
general, this zone is a few hundred feet thick. As the salt moves above
the crossover zone, a positive gravity anomaly is generated and superim-
posed on the broader negative anomaly because of the deeper density
contrasts. The addition of cap rock adds to the positive anomaly situated
on the broad negative anomaly. This is a characteristic anomaly signature
for many shallow Gulf Coast salt domes. The sediment densities shown
are typical of the U.S. Gulf Coast region.
Igneous intrusives produce a variety of signatures with characteristics
dependent on magnetic inclination, magnetic susceptibility, and configu-
ration. The first two figures show models for a very thin dike and sill com-
puted at two locations; Figure 5e is at 90°N inclination or at magnetic
north, Figure 5f is for 60°N or at an average inclination for the United
States. The skewness or distortion of a magnetic anomaly due to the mag-
netic inclination is identifiable and easily calculated. Note that the anom-
alies generated at 90°N are symmetrical in character. The interpreter
must identify the signal pattern for his geographic area before attempting
to integrate the magnetic information. An anomaly will exhibit total
reversals in signature when observed in the Southern Hemisphere.
The dip of a dike can be interpreted from the anomaly curve. The dike
model shown in Figure 5e is rotated through 90° to become a horizontal
sill. The resulting anomaly shown in Figure 5e changes from a steep sym-
metrical anomaly (90° dip at 90° inclination) to an asymmetrical anom-
aly (45° dip at 90° inclination) and back to a symmetrical anomaly with
a maximum amplitude for the sill at 90° inclination.
02 General Planning 4/16/07 4:03 PM Page 25

5a 5b 5c

5d 5e 5f

Figures 5a, b, c, d, e, and f.

The dike/sill model is repeated in Figure 5f, except this time the mag-
netic inclination has changed to 60°N. The anomalies generated are iden-
tified now as asymmetric, and they vary as demonstrated.
The magnitude of all the anomalies is dependent on the magnetic sus-
ceptibility and will vary according to the magnitude of the magnetic sus-
ceptibility of the type of rocks involved. The effect is linear: The higher
the magnitude of susceptibility, the higher the amplitude of the magnetic
anomaly. For example, a mafic intrusive at the same depth, the same
geometry, and intruding the same sedimentary section will have two
times or more the amplitude of a granitic plug.
As a sill extends laterally, there is a distance where the anomaly grad-
ually will separate into two distinct responses. In the Northern Hem-
isphere at 60°N inclination, the southern (left) end exhibits the positive
half of the anomaly, and the northern (right) end exhibits the negative
half. The addition of a feeder root to the sill generates another complete
anomaly (positive/negative pair) to the magnetic signature, as shown in
Figure 5g. The edges of a sill and the width, dip, and depth of the feeder
root can be interpreted from the magnetic signature. If the source of the

25
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:03 PM Page 26
Figure 5g.
anomalies was caused by a shallow salt sill and intrusive, the anomalies
would be reversed and would be an order of magnitude less in amplitude
than those from anomalies shown in Figure 5g.
Salt has a unique magnetic characteristic that interests the exploration
community. The magnetic anomalies are low amplitude but are nonethe-
less identifiable with state-of-the-art data-acquisition techniques. Due to
the diamagnetic character of salt, its magnetic signature is reversed, or
negative. Therefore, the dike/sill anomalies shown in Figures 5e and 5f,
if they had been generated by salt, would be the reverse of what is shown,
and the magnitude of the anomalies would be minimal. In general, the
maximum depth for identifiable, repeatable magnetic anomalies due to
salt is 2.44 km (8000 ft) below recording source.
Basement Models
Magnetic anomalies from crystalline or magnetic basement need not be
associated with structure on the basement. The broad magnetic closures
seen on total-magnetic-intensity anomaly maps are often due to changes
in the rock composition within the basement. When tectonic stresses
were applied, these zones of weakness were more likely to fracture and
fault. Therefore, the variation of basement composition often is correlated
to basement structure. In the following examples, the horizontal scale
was expanded to include more than one lithologic contact in a model.
Areas of higher susceptibility were assigned a positive density contrast.
A simplified basement transition zone has been modeled in Figure 6a
as one vertical boundary. Depth to basement is at 6.1 km (20 000 ft).
General/Planning/Large Scale
Both the magnetic and gravity anomaly responses for the intrabasement
contrast have been generated. The magnetic response for two magnetic
inclinations (60°N and 90°N) also are shown for comparison. The con-
trasts shown are considered to be conservative. The important character-
istic here is the anomaly wavelength (approximately 15.2 km, or 50 000
ft) produced by the transition zone at this depth. Intrabasement anomalies
generally have broad-wavelength anomalies.
Additional examples of the magnetic anomaly responses for two mag-
netic inclinations for intrabasement variations are shown in Figure 6b.
This example illustrates multiple boundaries, each having an assigned
magnetic susceptibility. The two magnetic anomalies generated are
caused by the two blocks with contrasting geophysical parameters. Note
that the narrower block produced a lower amplitude anomaly. A corre-
sponding broad-gravity anomaly is generated and shown in Figure 6c.
These models have anomalies generated by only the lithology variations
within the basement.
Suprabasement effects (basement topography) have been generated
using a uniform basement lithology. Structure (1.52 km, or 5000 ft) has
been added to two blocks and assigned a density contrast with the sedi-
ments equal to 0.25 g/cc. Both anomaly amplitudes exhibit an increase
in frequency content; Figure 6d shows the magnetic response and Figure
6e shows the gravity response. When comparing the magnetic anomalies
generated by basement structure versus basement lithology variation
(Figures 6d and 6b respectively), the anomaly wavelength and amplitudes
must be examined. The anomalies due to structure always will exhibit
higher frequencies. The gravity responses calculated for the same struc-
tures are shown in Figures 6c and 6e. The amplitude and the frequency
content of the gravity anomaly due to variation in basement lithology
have been increased by the addition of anomalies due to the basement
structures (compare Figures 6c and 6e).
The principal difference between intrabasement (lithology-variation)
and suprabasement (basement-topography) anomalies is that the former
have lower frequencies and higher amplitudes than the latter. In many
cases, when trying to match an intrabasement anomaly with a suprabase-
ment model, the structure computed is generally geologically unsound and
usually the solution violates other available geophysical data. The differences
in the characteristics of magnetic anomalies due to intrabasement (lithology)
and suprabasement (topography) causes always can be distinguished.
Conclusion
In conclusion, the geologic models shown have been limited to basic
structures for the express intention of illustrating the variation of anom-
aly signatures caused by variations in rock properties. By no means is this
26
the extent of gravity and magnetic interpretation or modeling capabilities.
Complex structures can be interpreted and modeled with confidence.
The quality of the solution is governed by the variety of constraints. For
example, if seismic data can input a minimum and maximum depth to an
anomalous zone, gravity and magnetics can assist in determining the
lithology and the configuration of the anomalous zone. In certain geologic
provinces, gravity and magnetic data also can be used as a good quality-
control monitor on seismic data. For example, the question “Is it velocity
or structure?” generally can be resolved with the integration of potential-
fields data.
Pattern recognition is the basis of effective interpretation. One of the
keys to accurate pattern recognition is the modeling of the prospective
structure, either to identify the characteristic signal or to support the
interpreter’s concepts.
Acknowledgment
Most of the models shown here are simplified examples from many actual
exploration cases. This paper started at the Superior Oil Company, where
initially real-world numbers were released for the model calculations long
before it was common to do so. Thanks go to it and to its managers, wher-
ever they may be today, for their encouragement, interest, and permission
to present these results. Many thanks also to my proofreader,
R. S. Pawlowski, and to my graphic illustrator, S. N. Stephens.
References
Talwani, Manik, and Heirtzler, James R., 1964, Computation of magnetic anomalies
caused by two-dimensional structures of arbitrary shape: Computers in mineral
industries, part 1: Stanford University Publications, Geological Sciences, 9,
464–480.
Talwani, Manik, Worzel, J. Lamar, and Landisman, Mark, 1959, Rapid gravity compu-
tations for two-dimensional bodies with application to the Mendocino submarine
fracture zone, J. of Geophys. Res., 64, 49–59.
Additional References
Bolt, Bruce, Adler, Bernie, Rotenburg, Manuel, and Ferbach, Sydney, 1973, Methods
in computational physics: Academic Press, Inc.
Flavelle, A. J., and Anfiloff, W., Non-standard gravity anomalies over sedimentary
structures: Austral. Petr. Expl. Assn. J., 16.
Grant, F. S., and West, G. F., 1965, Interpretation theory in applied geophysics:
McGraw-Hill Book Co.
Shuey, R. T., and Pasquale, A. S., 1973, End corrections in magnetic profile interpreta-
tion: Geophysics, 28, 507–512.
02 General Planning 4/16/07 4:03 PM Page 27

6a 6b 6c

6d 6e

Figures 6a, b, c, d, and e.

27
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:03 PM Page 28
Satellite Gravity
E. K. Biegert
Shell Exploration and Technology Company
Houston, Texas, USA
Gravity derived by observing orbital vari-
ations of satellites as they are perturbed
by the Earth’s gravity field provides valu-
able information about the deep internal
structure of the Earth. The coverage is
global, encompassing both land and
oceans. These data are used in studies of
plate tectonics, subduction zones, core-
mantle and mantle-crust anomalies, and
isostatic compensation beneath mountain
ranges.
Gravity derived from satellite altimetry
measurements and analysis of satellite
orbits is an inexpensive way to obtain
good spatial coverage over large areas of
ocean. It has been used to map fracture
zones, seamounts, hot-spot chains, mid-
oceanic ridges, subduction zones, and
many previously undiscovered features.
In offshore areas where reliable bathyme-
try is available, the effects of sea-bottom
topography can be removed to produce
gravity maps suitable for mapping conti-
nental-margin structure and detecting
sedimentary basins. The long-wavelength
components of satellite gravity maps can
be used to tie and level smaller marine
gravity surveys, providing a common
mesh in which local high-resolution sur-
veys are imbedded. In frontier areas,
satellite gravity can be processed with
bathymetry data to detect submarine
basins.
General/Planning/Large Scale
28
Gravity Data Define Basin Structure 1) an upper geologic sedimentary sequence in the basin should consist
of low-density rocks, overlying deeper high-density rocks
and the Location of Major Oil and Gas
2) topographic differentiations at the basin surface should be minor or
Reserves: Examples from Subandean absent, or the gravity database should be adjusted correctly for topo-
Basins, Tunisia, and the U. S. Rocky graphic surface effects
Mountain Region 3) there should be no major influence from intrabasement density vari-
ations, or such variations must be corrected for using complemen-
Johann-Christian Pratsch tary magnetic data to map depth to basement
Petroleum geologist
Houston, Texas, USA Where these conditions are met, Bouguer gravity maps indicate
EDITOR’S NOTE: Frontier exploration starts with correlations with 1) thickness of an upper (younger) low-density unit (for example, total
the known. Beyond that, the interpreter draws conclusions. thickness of Tertiary and Late Cretaceous sediments in the Llanos
Pratsch’s paper documents provocative correlations in some Basin, Colombia), and
very prospective areas. 2) both regional and semiregional structure at the level of the lower
(older) high-density unit (for example, structure at top Cretaceous,
top Paleozoics, or top basement in the Llanos Basin, Colombia)
Abstract
Bouguer gravity maps can be quite valuable in oil and gas exploration to Regional structure maps are used in the analysis of preferred hydro-
obtain regional deep-structure form-line maps. Such maps critically help carbon migration pathway position and directions, because (Figure 1)
to assess regional oil and gas migration pathways. Quick data turnaround
and low cost make this approach especially valuable for early exploration 1) hydrocarbons migrate under the influence of subsurface pressures,
phases in large or remote basins on- and offshore. and
2) subsurface isobars are parallel to regional structure, so that
Introduction
All oil and gas accumulations are the result of lateral and vertical hydro-
carbon migration from the generating depocenter into structural and/or
stratigraphic traps. The position and direction of such main or preferred
hydrocarbon migration pathways can be predicted: They are controlled by
sediment permeability and by basin geometry (i.e., regional structure) at
the level of main carrier beds.
Permeability cannot be predicted and is assumed in this approach to
be omnipresent in fractured and porous rocks. Permeability restrictions
will cause shortening of lateral hydrocarbon migration pathways; they
will not alter regional migration directions. Basin geometry is obtained
from regional structure maps. Ideally, regional seismic depth maps tied to
reliable wells should be used. However, such maps are rarely available. In
their place, regional Bouguer gravity maps are used successfully as
regional deep-structure form-line maps if three main conditions are Figure 1. Basin geometry, preferred migration directions, and prospectivity differences
fulfilled: (from Pratsch, 1986).
02 General Planning 4/16/07 4:03 PM Page 29

3) hydrocarbon pathway directions tend to lie perpendicular to regional

isobars, to regional structure contour lines, and to regional Bouguer
gravity contours (Figure 1) (Pratsch, 1986)

Ideal are kidney-shaped depocenters (Figure 1); here, regional hydro-

carbon migration pathways are focused toward positive structural anom-
alies plunging basinward. Such structurally positive areas also may offer
preferred conditions for shallow-water reservoir development of both clas-
tics and/or carbonates. It is here that the largest oil and gas fields in a
basin will occur.
The importance of these observations is well illustrated by the expe-
rience that in any producing basin in the world, 75% or more of its
reserves are located on only 25% or less of its acreage (Pratsch, 1986).
There must be a critical minimum density of gravity stations or gravity-
survey lines. A repeatable accuracy of about 5 mGal is required. This
method, therefore, is well suited as interpretation of “old” surveys and for
airborne surveys. Where the quantity/density of gravity stations permits
the calculation of residual gravity anomalies, even local structures can be
identified; this may become an important factor in preinvestment acreage
evaluations.

Examples
Most basins with a (low-density) Tertiary/Mesozoic sedimentary basin fill
overlying (higher density) Paleozoic and basement and located over uni-
form basement are ideal candidates for the gravity-based hydrocarbon
exploration approach outlined above. Such basins occur in most produc-
ing regions. One such region is the Subandean Basin trend in South
America, including the Lower Magdalena and Llanos Basins in Colombia
(Koch, 1995).
Lower Magdalena Basin, Colombia
In the Lower Magdalena Basin, Colombia, a published Bouguer gravity
map (Kellogg et al., 1991) shows two gravity lows interpreted as struc-
tural lows and potential depocenters; they are separated by a positive
anomaly interpreted as a structurally high ridge (Figure 2). There is also a
closed gravity high on the eastern side of the northern low. The basin con-
tains a thick Tertiary clastic and carbonate section (to about 3657 m, or
12 000 ft), possibly overlying Mesozoics and basement (Pratsch, 1995).
On the central high (“Cicuco High”), a Tertiary sedimentary sequence of
about 1829-m (6000-ft) thickness lies on Jurassic igneous rocks, part of
the Central Cordilleran Paleozoic/Mesozoic Volcanic Arc complex.
Producing fields of the Cicuco Field trend exist on the basement ridge,
Figure 2. Colombia, Lower Magdalena Basin; regional Bouguer gravity and major oil
and gas fields (gravity from Kellogg et al., 1991) (from Pratsch, 1995).
another field (El Difícil) on the structural high to the east. Main reservoirs
in Cicuco are Tertiary clastics, carbonates, and fractured basement, prob-
ably Cretaceous clastics in El Difícil. Source rocks are present in
Cretaceous and Tertiary shale units. Thermal maturity of these potential
source rocks is sufficient only in the depocenter lows. Several additional
Bouguer gravity anomalies exist in the basin; they may indicate the pres-
ence of several additional undrilled structural and structural-stratigraphic
prospects.
Llanos Basin, Colombia
Tertiary, Mesozoic, and Paleozoic sediments thicken westward across the
Llanos Basin toward the eastward-thrusting Eastern Cordillera (Koch,
Figure 3. Colombia, Llanos Basin; regional Bouguer gravity and major oil fields; dotted:
East Cordillera thrust front (gravity from Kellogg et al., 1991) (from Pratsch, 1995).
1995). A published Bouguer gravity map (Kellogg et al., 1991) (Figure 3)
indicates a regional westward-dipping monocline interrupted by several
prominent positive anomalies: the Arauca High, Macarenas High, and
Rubiales High. Each of these regional highs is a basement-rooted regional
uplift. More than 85% of the basin’s present oil and gas reserves occur on
these basement highs in local traps (Pratsch, 1994a). Reservoirs are sand-
stones of Paleogene and Late Cretaceous ages; source beds are in Late
Cretaceous and Paleogene shales. Thermal maturation is present in
Tertiary and Upper Cretaceous source rocks in the area of the Eastern
Cordillera and in a narrow band to the east parallel to its eastern flank;
both areas are contained in the Bouguer gravity low area.

29
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:03 PM Page 30

General/Planning/Large Scale
30
Central Onshore Tunisia
Tunisia’s geology reaches from the North African Paleozoic Platform or
Craton in the south to the Atlas Mountain mobile belt as part of the
Mediterranean active margin in the north. All oil and gas fields in the
country clearly are located inside or near generating local depocenters
(Pratsch, 1994b). Future oil and gas exploration requires a definition of
additional depocenters.
In the past, public-domain regional gravity data in Tunisia (Ferjani et
al., 1990) (Figure 4) have been interpreted as indicating (only?) Moho
structure (Pratsch, 1994b). However, the same data also indicate the pos-
sible presence of a large (150 km × 300 km) Paleozoic basin in the area
north of the intra-Paleozoic Telemzane High and west of the well-known
north-south structural axis. It is part of the North African Paleozoic cover Figure 6. Tunisia; section across Telemzane High (from ETAP/Schlumberger, 1990).
of the North African Craton (Figure 5).
Powder River Basin, Wyoming, USA
The Powder River Basin of Wyoming, USA, is one of numerous Rocky
Mountain foreland basins (Pratsch, 1986). Its sedimentary fill is of
Paleozoic to recent ages; several potential source-bed, reservoir, and seal-
rock units are present. The basin is asymmetrical, with a deeper and
wider western portion and a shallower and narrower eastern portion.
Thrusting occurs on both flanks against the bordering structural highs.
Paleozoic and Cretaceous oil and gas source rocks are mature in the deep-
est western portion of the basin. The distribution of oil and gas fields is
difficult to assess on the basis of published regional structure maps.
However, a public-domain regional Bouguer gravity map (Figure 7) (SEG,
1982) shows a northwest trend of individual closed negative lows, a
Figure 5. Tunisia; distribution of Paleozoic sediments (from Ferjani et al., 1990). major closed positive high, and several basinward- (westward-) plunging
positive gravity “noses.”
The twenty-two largest fields in the basin (each one with more than
Supporting stratigraphic data exist in few wells across the Telemzane
15.0 million barrels oil cumulative production by 1984) had produced a
High (Figure 6) (ETAP/Schlumberger, 1990); they indicate that this
combined total of 1.5 billion barrels oil, or 78.5% of the basin total
Paleozoic basin contains a thick clastic Paleozoic sedimentary sequence
(Pratsch, 1986). All but two of these fields (Hartzog Draw and House
(more than about 1524 m, or 5000 ft) of Cambrian to Permian ages.
Creek) are shown on our map, located on positive gravity anomalies.
Triassic red beds and evaporites are thought to form a regional seal.
These two fields, however, are located on similar positive gravity fea-
Reservoirs and potential source beds are expected in several Paleozoic
tures expressed on unpublished oil-industry gravity maps, containing
units, similar to their stratigraphic equivalents in producing areas of north-
more detailed data than the SEG map. After all, our gravity map is just
east Algeria and south Tunisia. Major lateral migration pathways for
an enlargement of the published (SEG, 1982) regional map, scale
Paleozoic-generated hydrocarbons will be from the regional low toward
1:2.5 million.
the basin flanks and into intrabasinal highs. No exploration for these tar-
These results are somewhat astonishing because stratigraphic factors
gets is yet under way, to our knowledge. Reserves may lie in the multi-
are important contributors in all fields in this basin, and the effective
100.0 million barrels oil equivalent (BOE) class; both oil and gas are
structural-stratigraphic traps cause numerically smaller gravity expres-
Figure 4. Tunisia; regional Bouguer gravity and structural lows (gravity from Ferjani et
expected.
sions than those indicated by the Bouguer gravity map.
al., 1990) (from Pratsch, 1994b).
02 General Planning 4/16/07 4:03 PM Page 31
Figure 7. Wyoming, Powder River Basin; regional Bouguer gravity and major oil fields
(from Pratsch, 1986); BH—Black Hills Uplift; BM—Bighorn Mountains Uplift;
HV—Hartville Uplift; WR—Wind River Basin.
The explanation is complex but logical (Figure 8): Bouguer gravity
anomalies in this basin seem to have been caused by both basin structure
and by internal basement density variations. Basement structure most
likely was subdued but sufficient to cause topographic differentiations in
Cretaceous seas; sandstone maxima (delta lobes, bars, channels, river sys-
tems) were located depending on basement-rooted topography. As a
result, facies changes from downdip shale to updip sand occur over base-
ment anomalies indicated as positive gravity highs. A more in-depth
analysis will require detailed gravity residuals, magnetics, well, and seis-
mic data. The fact remains that nearly 80% of the Powder Basin oil pro-
duction is in fields located on Bouguer gravity positive anomalies shown
on a published 2.5 million-scale regional map. This observation appears
quite important for future oil and gas exploration strategies in many other
basins, in the Rocky Mountain region and elsewhere.
References
Entreprise Tunisienne d’Activites Petrolieres (ETAP)/Schlumberger, 1990, Tunisian
exploration review.
Ferjani, A. B., Burollet, P. F., and F. Mejri, 1990, Petroleum geology of Tunisia: ETAP.
Kellogg, J. N., Ghodley, V. M., Ropain, U. C., Bermudez, A., and Aiken, C. L. V.,
1991, Gravity field of Colombia, eastern Panama, and adjacent marine areas: Geol.
Soc. Amer., map and chart series, MCHO7O.
Koch, E., 1995, Kolumbien/Colombia, in H. Kulke, Ed., Regional petroleum geology
of the world: Gebr. Borntraeger, II, 18–524.
Pratsch, J. C., 1986, The distribution of major oil and gas reserves in regional basin
structures—an example from the Powder River Basin, Wyoming, USA: J. of Petr.
Geol., 9, 393–412.
——— 1994a, The location of major oil- and gasfields: Examples from the Andean
Foreland: J. of Petr. Geol., 17, 327–338.
——— 1994b, Tunisia’s oil and gas potential and future plays: 4th Tunisian Petroleum
Exploration Conference, proceedings, 1–24.
——— 1995, Nordwest Kolumbien, in H. Kulke, Ed., Regional petroleum geology of
the world: Gebr. Borntraeger, II, 518–520.
Soc. Expl. Geophys., 1982, Gravity anomaly map of the United States exclusive of
Alaska and Hawaii, scale 1:2.5 million.
Figure 8. Wyoming, Powder River Basin; reservoir development and location of major
traps (from Pratsch, 1986).
Summary
Our experience shows that 75% or more of a basin’s oil production is con-
tained in only 25% or less of the basinal area. A major oil-concentration
process is indicated where lateral and vertical migration from the gener-
ating basin lows (“depocenter”) onto structural highs lying in the “right”
migration pathways is an important parameter. Critical hydrocarbon
migration pathways are related to basin structure in existence at the time
of migration of oil and gas. Of all the data that can be used to define basin
structure, gravity is a fully acceptable low-cost tool. Where gravity data
are integrated with other data (magnetics, seismic, well, surface), the final
results will become even more reliable. For difficult terranes such as rain
forests, deserts, or offshore areas, modern airborne and satellite-derived
gravity surveying techniques will be especially valuable.
31
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:03 PM Page 32
Utility in Exploration of
Continent-scale Data Sets
Richard I. Gibson
Gibson Consulting
Golden, Colorado, USA
Small-scale assemblies of gravity and
magnetic data sets and their interpreta-
tion are valuable for hydrocarbon explo-
ration. Since 1982 in the United States
and 1988 worldwide, such data have
become increasingly available and used
in oil companies, some of whose person-
nel never may have looked at a gravity
map before. This philosophy paper has a
list of applications of these data sets and
their interpretations, followed by some
pitfalls and a list of some existing conti-
nent-scale data sets.
Applications
1) to provide a regional context for
exploration strategy, concession and
farmout evaluations, and medium to
long-term planning
2) as a means of high-grading and refin-
ing geologic parameters that may have
impact on exploration
3) to identify hitherto unknown (or little
studied) features whose effects on the
prospective section should be investi-
gated further
4) as a means of defining the way geo-
logic and tectonic features are
expressed in gravity and magnetic
data, and the mapping of such features
in an academic sense (at least)
5) to suggest locations where the
prospective section may be unexpect-
edly thick or may contain large struc-
tures (classical reconnaissance use of
gravity and magnetics)
6) to identify locations where the disci-
pline (gravity or magnetics) is good
at or has difficulties in the resolution
of geologic problems
7) in areas of good control, to interpret
the data in detail in light of local
geology and tectonics (especially
appropriate in frontier areas or areas
where a company or individual has
little experience or knowledge)
8) in areas of good control, to identify
analogs to existing hydrocarbon pro-
duction, and sometimes to derive and
define new play concepts, even in
mature basins
9) to permit the identification and inter-
pretation of large features that cannot
be “seen” in large-scale displays of
data or in individual surveys, no mat-
ter how high the quality
10) to serve as a starting point for more
detailed work
11) to provide a first-pass guide to plan-
ning seismic and other work
12) to serve as an index to data and data
quality so that informed decisions can
be made regarding further gravity- or
magnetic-data acquisition
13) to provide a base of experience and
interpretation that can be applied
worldwide, whether by individuals or
through use of in-house reports, a
resource for first-pass, “quick-answer”
interpretations
General/Planning/Large Scale
In general, “comprehensive” interpreta-
tions of such data sets are preliminary,
first-pass efforts limited in terms of inte-
grating other local information into the
data set. By all means, such interpreta-
tion should be done in as much detail as
feasible in a reasonable time frame.
Interpretation of all or much of a data set
will be more informative than piecemeal
local interpretation that is not set into a
regional context. In my opinion, such
work requires, at a minimum, enough
geologic and tectonic knowledge in any
area to constrain the interpretation of the
gravity and magnetic data, especially in
terms of structural style and orientation
of inferred features such as faults.
Otherwise, literally any inference is pos-
sible. It is also my opinion that such
analysis is done best by a geologist who
understands the limits and shortcomings
of gravity and magnetic data in such dis-
plays, rather than a physics-oriented geo-
physicist who does not know enough
geology. No interpretation of gravity or
magnetic data is unique.
After a first-pass, interpretive context has
been established, others with more
knowledge about specific areas should
refine these regional interpretations in
the light of other geologic, geochemical,
or geophysical constraints. In many
areas, these small-scale displays contain
enough information to guide and con-
strain even very local work. However,
such guiding only rarely can be provided
through a cursory glance at the data
without familiarity with the data set (its
spacing, origin, etc.) and at least general
knowledge of the geologic and tectonic
meaning of the data at local levels (even
though that knowledge may be interpre-
tive rather than confirmed).
Pitfalls
Data and Grid Spacings
The single biggest problem in examining
small-scale data sets, which invariably
are created from diverse original sources,
is failure to be aware of the spacing of
the original data or grids derived from
them, or both. Digital compilations per-
mit the redisplay of data at any scale,
using various contouring packages. Such
displays may be inappropriate to the orig-
inal data set or the derived grid. For
example, the University of Leeds gravity-
data assemblies (see below) actually con-
sist of a grid of points that is 5 minutes ×
5 minutes in size, irrespective of the
original data spacing. Thus, if you are
looking at a map that shows the original
data spacing to be so dense as to be
black (at 1:2,000,000 scale, a spacing of
about 1 km or less), you might expect to
see great detail in the maps. You cannot,
because the grid, created by weighting
all the original data points within a
5-×-5-minute “cell,” is too broad to re-
solve features smaller than about
2.5 minutes, or about 5 km across.
Equally misleading are areas of broad
spacing or no data at all. Today’s con-
touring programs can handle such situa-
tions, and I have seen beautiful color
Versatec contour maps of regions that
don’t have a single original control point
within 100 km. Usually, such areas will
be fairly obvious-—extremely broad
anomalies, unlike those in adjacent areas.
But many contour programs can create
nonexistent closures or gradients. The
upshot of this is that these kinds of data
sets should not be examined (in my opin-
ion, at all, by anyone) and interpreted
without a firm awareness of the spacing
of original data and the size of the grid.
32
Usually, this means having the station
locations actually plotted on the map.
What about displays of these data at dif-
ferent scales? In general, the scales cho-
sen by the compilers of such data sets
are appropriate to their spacings. Scales
such as Leeds (1:2,000,000), DNAG
(1:5,000,000), and USA and USSR
(1:2,500,000) are small enough to obvi-
ate many (but by no means all) of the
problems of original data spacing and
variability, and grid size. At the same
time, they are large enough to provide
useful detail for hydrocarbon explo-
ration—far, far more detail than simply
the major tectonic features. Specific
faults and intrabasement lithologic con-
tacts are quite reliably interpretable, as
examples. In my opinion, it is not wrong
to redisplay such data sets at larger
scales, as long as the viewer is aware of
the original grid size, original data distri-
bution, and the nature of the contouring
program and the artifacts it may create.
There are some limits to this, of course.
It is not appropriate to map DNAG data
at 1 inch = 4000 ft, even if the original
data would warrant such a display. Go to
the original data.
Variations Among Surveys
In addition to the problems of data spac-
ing, other survey parameters should be
considered when examining maps that
are amalgamations of diverse original sur-
veys. The Leeds gravity maps have
undergone exhaustive reprocessing to
make data from different surveys con-
form to a single Bouguer correction den-
sity and other factors. This is a highly
desirable thing to do, and it reduces the
need to worry about many of the original
survey parameters. Some other assem-
blies may not be reprocessed so thor-
oughly, so apparent breaks between
02 General Planning 4/16/07 4:03 PM Page 33
surveys with different parameters may
occur. This problem is more common in
assemblies of magnetic data. In particu-
lar, variations in aircraft flight altitude
may make for spurious depth differences
across survey boundaries. The argument
could be made that such problems could
be eliminated by upward or downward
continuing of the assembled surveys to a
single datum. This is true, but the
amount of information lost by such pro-
cessing negates the advantages, in my
opinion. I much rather would look at an
assembly of original data, even if the con-
tours have been forced into one another
along survey boundaries (as they have
been in the magnetic map of the United
States), as long as I know where the sur-
vey boundaries are and what the survey
parameters are. Thus I can look at the
dramatic difference in magnetic character
between Iowa and Illinois and, knowing
that the difference in aircraft altitude is
about 5500 ft, refrain from interpreting
a major basement fault along the Missis-
sippi River. The existing magnetic map of
the former Soviet Union suffers from a
lack of knowledge of such parameters; it
is possible to identify suspected artifacts
in the data when anomalous data breaks
appear in cardinal directions (e.g., north-
south or east-west) that do not make
sense tectonically. There are remarkably
few such locations in this particular data
set.
Bathymetry
A problem with gravity data offshore is
the influence of ocean-bottom topogra-
phy. Until recently, many offshore gravity
data were free-air data, because inade-
quate knowledge of bathymetry pre-
cluded the preparation of Bouguer
gravity maps. Free-air gravity maps can
mirror the contours of bathymetry. The
intelligent way to resolve the problem is
to interpret the data with bathymetry
directly overlaid so the interpreter can
take it into account. Increasingly, good
regional bathymetric information pro-
vides a means of computing the Bouguer
effect so “bathymetric anomalies” can be
eliminated. Be aware that such computa-
tions are only as good as the bathymetric
grid (whose spacing is subject to the
caveats listed above), and some variabil-
ity does exist in the quality of bathymet-
ric data today. Bathymetric effects,
essentially subsea-terrain effects, can pro-
duce error in the best Bouguer gravity
map unless the bathymetry is well known
enough that terrain corrections can be
made. An important specific feature in
gravity data to be careful of in interpreta-
tion is the continent-ocean transition
anomaly, an intense linear high found
along many rifted margins. It is nearly
coincident in places (but not everywhere)
with the shelf break, so there is a bathy-
metric effect to consider.
Geologic Pitfalls
One of the greatest difficulties in looking
at a continent-scale gravity map is differ-
entiating between a broad area of thick
crust and a broad basin. In fact, without
some constraints, it is not possible to
make this distinction. Thick crust pro-
duces a broad gravity low because the
relatively low-density crust sticks down
into relatively high-density mantle.
Precambrian shields and more local
“pods” of crust can produce this effect,
although variations in the upper mantle
and asthenosphere, without thickening
the crust, can do so as well. More impor-
tant from the hydrocarbon-exploration
point of view is the fact that a shallow,
broad area of low-density material (i.e., a
basin) can create a similar gravity effect.
One would expect narrower, near-surface
(versus lower-crust) anomalies to be pres-
Geologic Applications of Gravity and Magnetics: Case Histories
ent that would indicate geologic varia-
tions within the basin, but if the data
spacing is broad, such features may not
be resolved. What remains is a broad
gravity low. One might think surface geo-
logic mapping of the world is good
enough to make this distinction, but it is
not. In many parts of Africa and some
places in South America, it is impossible
to tell if certain gravity lows are thick
sediments or thick crust, and the existing
geologic maps are either contradictory or
uninformative.
In some places, intrabasement sutures
create density contrasts that result in lin-
ear lows similar to elongate grabens filled
with sedimentary rocks. This problem
also exists in places where the surface
may be mapped accurately as Quaternary
or Tertiary deposits, leaving the subsur-
face interpretation unconstrained. In
many, but not all, locations where this
problem exists, magnetic data would per-
mit the interpretation of shallow base-
ment in a thick pod of crust versus
nonmagnetic sediments and deep mag-
netic basement.
Exceptions exist, of course—sedimentary
basins with abundant magnetic vol-
canics, and crystalline basement that is
not magnetic. It is a fact that some well-
known basins coincide with gravity
highs because they contain thick, dense
material—typically inferred to be basalt,
especially if the basin is an extensional
rift near a contemporary sea-floor spread-
ing axis. Sometimes such gravity highs
are interpreted more reasonably as man-
tle highs or bona fide structures (base-
ment level or shallower) or even sedi-
mentary variations (1000 m of anhydrite,
density 2.9, affect the gravity anomaly!).
Let the tectonic setting be your guide.
A similar geologic problem with mag-
netic data is the differentiation of shallow
basement from volcanics. Essentially, it
cannot be done without at least some
knowledge of geology and/or tectonics.
However, a good clue lies in the fact that
if shallow volcanics are found in an area
of deep basement, two distinct magnetic
horizons will exist, and this is often dis-
cernible in the magnetic data set. If the
volcanics and basement are near each
other vertically, this distinction becomes
difficult or impossible without further
processing and analysis.
The outline above lists some of the prob-
lems that are more or less specific to
small-scale data sets. Other pitfalls of
gravity and magnetic interpretation also
apply, although some (e.g., cultural
effects, shallow sources such as glacial
drift, etc.) effectively are “filtered out” by
the gridding and display scales typical of
these compilations.
Some Existing Continent-scale
Data Sets
North America. The DNAG (Decade of
North American Geology) data set is
based on a grid derived from a huge
number of original data points, and
includes both gravity and magnetics. The
published map scale is 1:5,000,000.
Many sources were used, but the U.S.
and Canada portions were based on
existing assemblies. U.S. (magnetics):
mapped at 1:2,500,000, highly diverse
original data, survey boundaries and
parameters provided. Alaska and Hawaii
mapped separately at the same scale.
Many states mapped at 1:1,000,000 or
1:500,000. U.S. (gravity): 1:2,500,000.
Canada (magnetics): 1:5,000,000, from
highly consistent, high-quality original
surveys (usually published at 1:50,000).
The Canadian government surveys are
some of the highest quality published
magnetic maps in the world. The North
American data are available on tape and
on CD-ROM (with topography, seismic-
ity, and other data) from the National
Geophysical Data Center in Boulder,
Colorado.
Gravity Maps of Africa and South
America. These two projects are propri-
etary compilations prepared by the
University of Leeds, England, for consor-
tia of oil companies. The maps are avail-
able for purchase. These compilations
differ from government-published data in
that they include confidential data pro-
vided to Leeds by their sponsoring oil
companies. However, many of the data
are published or held by major institu-
tions such as the Defense Mapping
Agency (now NIMA, National Imagery
and Mapping Agency)(USA), ORSTOM,
and Bureau Gravimétrique Internationale
(BGI) (France), and others, including
published data for various countries.
Leeds has done the immense job of care-
fully evaluating the data sets and rere-
ducing them to consistent datums with
the same correction factors. Effectively,
this makes these data into two new data
sets, to be thought of in their own rights.
The problems discussed above still apply,
but these assemblies are much better
than some with more cursory corrections,
or none at all. In spite of some large
“holes” in the data sets, these maps are
by far the best gravity maps of these
continents. Leeds’ display scale is
1:2,000,000 and 1:5,000,000. The
5-minute grids result in almost 1 million
grid nodes for each area, based on about
500 000 original data on land on each
continent, and 1 to 2 million data points
in each offshore area.
33
02 General Planning 4/16/07 4:03 PM Page 34
Soviet Union. The former USSR is cov-
ered (on land and non-Arctic marginal
seas) by magnetic data of quality gener-
ally between good USGS data and
Canadian government data. Published
display scale was 1:2,500,000, and these
contour maps were digitized separately
by Gibson Consulting (free enterprise at
work) and by the U.S. Naval Oceano-
graphic Office (your U.S. tax dollars at
work). The few areas of questionable
data quality usually are indicated on the
original published maps (ground surveys,
for example). The digitized data are
available from either of the sources men-
tioned above. GETECH (University of
Leeds) now is brokering point- and line-
based gravity and magnetic data for
Russia (minimum purchases apply).
Australasia and China. The continent
of Australia is covered nearly completely
by high-quality gravity and magnetic sur-
veys, comparable to those of the Cana-
dian government. Published scales of the
assemblies are 1:2,500,000 (magnetics)
and 1:5,000,000 (gravity). The data have
been incorporated into Leeds’ (GETECH)
southeast Asia gravity and magnetic pro-
jects, similar to the Africa and South
America projects described above.
Gravity and/or magnetics surveys for
China are also available through
GETECH (Leeds) and Paterson, Grant &
Watson Ltd. (Toronto).
Europe. GETECH has completed a com-
pilation of gravity data for west/east
Europe, finished in 1994. Contact
GETECH for more information. Many
country-scale or regional (e.g., Scandi-
navia) assemblies of gravity or magnetic
data have been made previously. Gravity
and magnetic data, together with other
data sets and pretty colored maps, are
available on CD-ROM as part of the
European Geotraverse Database, which
accompanies A Continent Revealed, a
book published by Cambridge.
World. Haxby (1988) published a grav-
ity map (1:40,000,000 scale) of the
world’s oceans based on SEASAT data,
whereby radar altimeter measurements
of sea height are converted to free-air
gravity data. These data are available on
tape from the National Geophysical Data
Center in Boulder, Colorado. A more
complete data set using more and newer
satellite passes is available through Dr.
David Sandwell at Scripps Observatory
in California, via the Internet. I under-
stand you can download it free! In 1991,
the Geological Survey of Canada pub-
lished the Bouguer Gravity Anomaly
Map of the World (1:50,000,000), based
on gravity holdings of the BGI (France)
(1.8 million data points on land), satellite
data, and the world bathymetric data
base. Although these maps are very small
scale, the information content is excel-
lent and they are useful for comparing
the expressions of regional and subre-
gional features from continent to conti-
nent—especially if you squint. As of
1992, Russian scientists were working
on a magnetic map of the world at
1:10,000,000 scale. A free-air gravity
map of the world was published in
1982 ( Bowin et al., GSA MC-45,
1:22,000,000 scale).
Time
Based on my experience, it takes an
interpreter (reasonably well versed but
not necessarily expert in the geology, tec-
tonics, hydrocarbon plays, and geography
of the world), supported by geologists’
guidance and assistance, five to eight
man-months to perform a thorough
review interpretation (emphasizing geo-
logic interpretation and hydrocarbon
General/Planning/Large Scale
prospectivity) of one of these data sets.
More detailed integration, say, of gravity
with magnetics should be expected to
take longer, and considerable information
could be derived from two to three addi-
tional months of interpretation.
34
The Isostatic Gravity Residual of Onshore
South America: Examples of the Utility
of Isostatic Gravity Residuals as a Regional
Exploration Tool
David A. Chapin
LaCoste & Romberg
Austin, Texas, USA
EDITOR’S NOTE: What a great map! This study tackles the age-old
problem of gravity regional and residual separation in a truly
rigorous and innovative way. The result is a clearer picture of
geologic features.
Abstract
With the recent availability of continental-scale gravity data sets, a useful
method to make regional comparisons becomes important. Isostatic grav-
ity residuals remove unwanted crust-mantle effects from the Bouguer
data and allow comparisons of large-scale geologic features, regardless of
their elevation. Features such as the Andes Foreland and the Altiplano
Basin can be imaged better for exploration purposes.
Introduction
The past fifteen years have been an exciting time for those involved in
regional geophysical interpretation. This is primarily because of the intro-
duction of large, continental-scale gravity and magnetic data sets, such as
those compiled in North America by the DNAG committee (Geol. Soc.
Am., 1987), and the numerous compilations by the University of Leeds
(now GETECH), which includes South America, Africa, Europe, the for-
mer Soviet Union, and Asia (Fairhead, et al., unpubl.). These data sets
offer a new synoptic view of continental areas for oil explorationists.
Continental-scale gravity data sets pose a special problem for regional
interpretation. This is because of the well-known relationship between
the Bouguer anomaly and elevation: In a regional sense, Bouguer gravity
is more negative over higher elevations than at lower elevations. As an
example, in South America, the regional gravity field ranges from a min-
imum of about –400 mGal over the high Andes to near 0 mGal at sea
level. For almost 150 years, this has been understood to be the expression
of the gravity effect at the crust-mantle interface (Pratt, 1855; Airy, 1855;
02 General Planning 4/16/07 4:03 PM Page 35
Woollard, 1959; and Tsuboi, 1979). Dutton (1889) termed the concept
of crust-mantle support for topography isostasy.
Most oil explorationists are not concerned about these large, deep-
seated effects. We are more interested in the shallower features. But iso-
static gravity effects in continental-scale Bouguer gravity can confuse our
ability to compare similar-appearing features which exist at different ele-
vations.
For a synoptic view of a continental gravity data set, it is important
that the regional isostatic field be removed in some intelligent fashion.
Chapin (1996) provides details about a better method to remove the iso-
static regional. Figure 1 shows the output of this methodology.
Significance of Isostatic Residuals
Both Heiskanen and Vening Meinesz (1958) and Simpson et al. (1986)
discussed in detail the significance of isostatic residuals. If our Earth
model completely explains the gravity of the Earth, then there would be
no residual. But since, as explorationists, we are searching for local
unknown inhomogeneities in the crust (i.e., the next billion bbl struc-
ture), failure of our Earth model is quite important. Indeed, once the
imperfect Earth model (which is considered the isostatic regional ) is
removed, we have a better chance of identifying local anomalies.
There are three major causes for these isostatic residual anomalies:
1) failure of main assumptions in the isostatic and other Earth models
used for computation. Some of this is minimized through a more rig-
orous approach towards the isostatic model (Chapin, 1996).
2) large-scale lithospheric loading and/or flexure. These are lateral
effects which are manifested as “local” vertical imbalances. In other
words, if the vertical column of rock were decoupled from the adja-
cent crust, it would tend to seek a different bouyant level.
3) isostatic imbalances. These are the anomalies of most interest for
exploration purposes. There are three types:
a) The cause of the imbalance is still present. Examples: uplift of
mountains, subsidence of basins.
b) The cause of the imbalance was recently present but is no longer
present; therefore, the Earth still is readjusting. Example: glacial
rebound in Canada and Fennoscandia.
c) The cause of the imbalance is gone, but rebound stresses do not
exceed elastic limit of lithosphere. Example: many sedimentary
basins and structures within these basins.
From an exploration perspective, it is clearly group No. 3—isostatic
imbalances—which is the most important. There are two excellent exam-
ples in South America of how isostatic residuals can aid in regional explo-
ration.
Example A: Andes Foreland
Along the eastern foreland of the Andes is a long, continuous residual-
gravity low that runs from Venezuela to Argentina (Figure 1, locations A).
This feature cannot be imaged successfully in the original Bouguer grav-
ity. The foreland gravity low corresponds to a continuous series of basins
which contain the bulk of the South American hydrocarbon reserves. This
includes such fields as the Cusiana (Tachet des Combes and Mondy,
1995) and Caño Limón in Colombia (Salazar, 1990; Gabela, 1990), the
prolific Oriente Basin in Ecuador (Rosania S., 1990), and the Nuequen
Basin in Argentina (Stabler, 1990), among others. This particular gravity
feature marks basins that are near sea level in Venezuela (Krause and
James, 1990) as well as those that are more than 3000 m in elevation in
Bolivia (Stabler, 1990). Yet the anomaly has a consistent appearance
throughout the continent. There is probably an element of lithospheric
flexure associated with the loading of the Andes on the Brazilian Craton
overprinting the anomaly (Lyon-Caen et al., 1985); the isostatic model
used for Figure 1 did not account for flexure. This anomaly seems to mark
the same or similar geologic structure along its entire length—that of a
continuous series of foreland basins.
Example B: Altiplano
Isostatic residuals can help one understand the regional structure within
the Altiplano area of western Bolivia. Here, at elevations above 3500 m,
the Bouguer gravity cannot resolve a basin within the very low gravity
values associated with the regional isostatic effects of the high Andes
(Omarini and Götze, 1991). However, with the successful application of
isostatic residuals, the image of a consistent basin can be seen clearly. The
well-delimited green area located on Figure 1, location B, is an isostatic Figure 1. Isostatic gravity residual map for the University of Leeds South American
residual low which identifies the large Altiplano basin. This basin was data set modified from Chapin (1996). Blank areas correspond to offshore areas as
masked by the overwelming isostatic gravity effects of the Andes and well as areas which lack adequate station control. Cooler colors indicate residual-
gravity lows and hotter colors are gravity highs. The zero contour is between the
never has been seen this clearly in the gravity before. green and yellow color bands. A and B are the locations of examples discussed
in the text.
Conclusion
If the explorationist needs to look at the regional geology over a wide vari-
ety of elevations, it becomes important to analyze isostatic gravity residu-
als. They can provide a wealth of information to the interpreter. One of
35
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:03 PM Page 36

General/Planning/Large Scale
36
the most important benefits is the ability to compare similar structures in Pratt, J. H., 1855, On the attraction of the Himalaya Mountains, and of the elevated
One Person’s Regional Is
areas of different elevations. It can provide a more consistent view of an regions beyond them upon the plumb-line in India: Philos. Trans. R. Soc. London,
145, 53–100. Another Person’s Residual
entire continent of gravity data for regional exploration.
Rosanía S., G., 1990, Petroleum prospects of the sedimentary basins in Ecuador, in A “regional” is what you remove from
Acknowledgments Ericksen, G. E., Cañas Pinochet, M. T., and Reinemund, J. A., Eds., Geology of the the data to better image what you want
Andes and its relation to hydrocarbon and mineral resources: Circum-Pacific to “see” (residual). The best regional/
I would like to thank ARCO International Oil and Gas Co. and ARCO Council for Energy and Mineral Resources, Earth Science Series, 11, 415–430. residual combination is achieved by
Exploration and Production Technology for allowing me to publish this applying geologic thought to the prob-
paper. Thanks to J. D. Fairhead, GETECH and University of Leeds, for per- Salazar, R. B., 1990, Hydrocarbon resources in the sub-Andean basins of Colombia, in lem. What is the geologic target? The
mission to display the data for South America. U.S. Patent No. 5,390,114 Ericksen, G. E., Cañas Pinochet, M. T., and Reinemund, J. A., Eds., Geology of the hypothetical curve represented below
Andes and its relation to hydrocarbon and mineral resources: Circum-Pacific
has been issued on the methodology used to display the isostatic residual. Council for Energy and Mineral Resources, Earth Science Series, 11, 345–362.
illustrates the residualization concept,
the idea of zeroing in on a particular
anomaly or wavelength, simplifying the
References Simpson, R. W., Jachens, R. C., Blakely, R. J., and Saltus, R. W., 1986, A new isostatic
residual gravity map of the conterminous United States with a discussion on the sig- profile and, hopefully, the interpretation.
Airy, G. B., 1855, On the computation of the effect of the attraction of mountain- If your goal is a better understanding of
nificance of isostatic residual anomalies: J. Geophys. Res., 91, 8348–8372.
masses, as disturbing the apparent astronomical latitude of stations in geodetic sur-
the asthenosphere, you may want to cap-
veys: Philos. Trans. R. Soc. London, 145, 101–104.
Stabler, C. L., 1990, Andean hydrocarbon resources-—An overview, in Ericksen, ture the very long wavelength blue
G. E., Cañas Pinochet, M. T., and Reinemund, J. A., Eds., Geology of the Andes anomaly. The red anomaly may shed
Chapin, D. A., 1996, A deterministic approach toward isostatic gravity residuals: A
and its relation to hydrocarbon and mineral resources: Circum-Pacific Council for light on crustal thinning, while the green
case study from South America: Geophysics, 61, 1022–1033.
Energy and Mineral Resources, Earth Science Series, 11, 431–438. anomaly relates to basin depth. The
Dutton, C. E., 1889, On some of the greater problems of physical geology: Bull. short-wavelength orange anomaly may
Tachet des Combes, J., and Mondy, J. F., 1995, An evaluation of the uncertainties in be caused by the hydrocarbon trapping
Philos. Soc. Wash., 11, 51–64.
the structural definition applied to the delineation of the Cusiana Field (Colombia):
Presented at the St. Petersburg ’95 International Geophysical Conference and
structure.
Gabela, V. H., 1990, Exploration and geologic framework of the Caño Limón oil field, Exposition.
Llanos Orientales de Colombia, in Ericksen, G. E., Cañas Pinochet, M. T., and P. S. Millegan
Reinemund, J. A., Eds., Geology of the Andes and its relation to hydrocarbon and
Tsuboi, C., 1979, Gravity: George Allen & Unwin (Publishers) Ltd., 1983, English
mineral resources: Circum-Pacific Council for Energy and Mineral Resources, Earth
translation.
Science Series, 11, 431–438.

Woollard, G. P., 1959, Crustal structure from gravity and seismic measurements: J.
Geol. Soc. Am., 1987, Gravity anomaly map of North America—1:5,000,000: com-
Geophys. Res., 64, 1521–1544.
piled by the U.S. Geol. Surv. for the Decade of North American Geology.

Heiskanen, W. A., and Vening Meinesz, F. A., 1958, The earth and its gravity field:
McGraw-Hill Book Co.

Krause, H. H., and James, K. H., 1990, Hydrocarbon resources of Venezuela—their

source rocks and structural habitat, in Ericksen, G. E., Cañas Pinochet, M. T., and
Reinemund, J. A., Eds., Geology of the Andes and its relation to hydrocarbon and
mineral resources: Circum-Pacific Council for Energy and Mineral Resources, Earth
Science Series, 11, 405–414.

Lyon-Caen, H., Molnar, P., and Suárez, G., 1985, Gravity anomalies and flexure of the
Brazilian Shield beneath the Bolivian Andes: Earth and Planetary Science Letters,
75, 81–92.

Omarini, R. H., and Götze, H. J., 1991, Global Geoscience Transect 6: Central
Andean Transect, Nazca Plate to Chaco Plains Southwestern Pacific Ocean,
Northern Chile and Northern Argentina: Am. Geophys. Union.
02 General Planning 4/16/07 4:03 PM Page 37
The Compilation and Application of
Aeromagnetic Data for Hydrocarbon
Exploration in Interior Alaska
John F. Meyer Jr.
Department of Natural Resources, Division of Oil and Gas
Anchorage, Alaska, USA
Louis J. Racic, Paterson, Grant & Watson Ltd.
Toronto, Ontario, Canada
Richard W. Saltus, U. S. Geological Survey
Denver, Colorado, USA
EDITOR’S NOTE: Within the past decade, huge compilations of grav-
ity and magnetics data have become available to the interpreter.
The authors document the importance of creating a map of
properly merged data for tectonic interpretation in Alaska.
Introduction
To stimulate interest and provide background information for future
petroleum and mineral exploration in interior Alaska, the state of Alaska,
Division of Oil and Gas, and the U.S. Geological Survey joined in a coop-
erative effort to compile and merge all of the publicly available magnetic
data throughout interior Alaska. Interior Alaska extends from the Brooks
Range in the north to the Alaska Range in the south and from the
Peninsula on the west to the Alaska-Canadian border. Budget and time
constraints restricted the initial compilation to the area of state-controlled
lands within the Alaskan interior, extending from 61°N to 66°N and
144°W to 159°W (Figure 1). The primary sedimentary basins within this
area consist of the Copper River Basin, Susitna Basin, Minchumina Basin,
Holitna Basin, and Middle Tanana Basin (Kirschner, 1994).
The data processing was performed by Paterson, Grant & Watson Ltd.
of Toronto, Ontario, under contract to the state of Alaska. Aeromagnetic
maps at a scale of 1:500,000 have been published jointly by the U.S.
Geological Survey and the state of Alaska, Division of Oil and Gas (Meyer
and Saltus, 1995). In this study, twenty-three magnetic surveys, compris-
ing 280 000 line-kilometers, flown between 1954 and 1982, were
merged to produce two digital data grids. The first grid, referred to as the
composite grid (Figure 2), was produced to retain the closest resolution
to the original data for each survey and was
designed to be used for detailed modeling and
local depth-to-basement determinations. The
second grid, referred to as the merged grid
(Figure 3), was produced to simulate a survey at
a constant 305 m (1000 ft) aboveground. This
grid was designed to be used for regional analy-
sis within the area. The digital data grids are pub-
licly available (contact the National Geophysical
Data Center in Boulder, Colorado, by telephone,
303-497-6084, or email, info@ngdc.noaa.gov).
A Bouguer gravity map of the same region also
has been produced (Meyer et al., 1996).
The merged aeromagnetic data provide a
uniform and objective base map for tectonic
interpretation and resource exploration. It is the
best regional data set for detailed modeling of
3-D crustal structure. The regional magnetic
anomaly patterns reflect variations in lithology
and structure of the crystalline basement, the
nature and extent of magmatic components of Figure 1. Location of the magnetic data compilation, showing major basin areas.
the crust, and the location and geometry of
many tectonic boundaries.
the maps were digitized by hand (by digitizing the intersections of flight
The Compilation lines with contour lines). In one case, the data were available only in a
gridded format. These data sets were reprojected to the Albers Conic pro-
Compilation of a single, uniform magnetic data set from a series of indi-
jection. Each survey was edited to eliminate any obvious errors. Many of
vidual aeromagnetic surveys requires corrections and adjustments to
the surveys had pronounced line-to-line misleveling, which was corrected
account for differences between data sets collected at different times or
using decorrugation techniques (Urquhart, 1989). Documented datum
under different conditions. For example, the measured values from an
shifts or trends previously removed from the data were restored prior to
aeromagnetic survey collected at 305 m (1000 ft) aboveground will be
the Definitive Geomagnetic Reference Field removal. The data then were
quite different from those collected at 3657 m (12 000 ft) aboveground.
gridded at a grid-cell size corresponding to 1⁄5 of the nominal line spacing
Also, since the Earth’s primary magnetic field changes with time, surveys
for each survey using a minimum curvature algorithm. The cell size was
measured ten or twenty years apart can differ significantly. Mathematical
200 m for high-resolution surveys and varied to 2 km for the Department
models can adjust these flight-level contrasts and correct for time varia-
of Energy National Uranium Resource Evaluation data sets.
tions. This section is a description of the specific procedures required for
The composite grid (Figure 2) was created by resampling the grids to
the compilation of the interior Alaska aeromagnetic data.
a common cell size of 1 km and trimming the grids to allow a space
The twenty-three individual surveys that make up this compilation
between adjacent surveys. Simple datum shifts (i.e., addition or subtrac-
were flown at widely varying flight heights and line spacings and in dif-
tion of a constant value) were used to align the background levels; how-
ferent modes (i.e., a constant elevation aboveground, or “draped”; or a
ever, the original flight altitude of each survey was maintained.
constant barometric elevation). The data were available in one of three
To create the merged grid (Figure 3), each draped survey that was not
formats: digital profile archives, digitized map contours, or gridded data.
flown at an elevation of 305 m was continued upward to that elevation
Profile archives produce the highest quality grids, but many of the origi-
using standard 2-D fast Fourier transform (FFT) filtering techniques. The
nal archive data were not available. Where only contour maps existed,
37
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:03 PM Page 38

General/Planning/Large Scale
38

Figure 2. Composite magnetic map—All gridded survey data were reinterpolated to a 1-km cell size. The original altitude of each survey Figure 3. Merged magnetic map—All gridded data were continued to 305 m above the ground. The twenty-three surveys were
was maintained. leveled and merged to produce a seamless regional aeromagnetic grid.

survey flown at a constant barometric height was draped to the same
305-m elevation using the Compudrape technique (Paterson et al., 1990;
based on the “Chessboard” technique of Cordell, 1985). In this algo-
rithm, the magnetic grid is continued upward and downward to ten par-
allel survey datums that bracket the topographic relief. The final draped
grid is computed using these ten parallel grids to interpolate the magnetic
value at each grid cell to the 305-m terrain clearance level. Higher-order
correction surfaces were applied to obtain seamless joins between sur-
veys, required because of the high variability of the flight-line spacing and
consequent resolution of the grids. An automatic algorithm was used to
extract a coincident profile from regions of overlap between adjacent data
sets to calculate the difference. A correction surface was computed using
the difference profile, which then was added to one of the grids. A dis-
tance-based weighted-averaging technique was used to link the surveys
into the final, seamless merged grid (Black et al., 1995).
Utility to Exploration
Regional aeromagnetic data sets provide critical information on the tec-
tonic framework of the upper crust. The patterns and amplitudes of
anomalies reflect the depth and magnetic character of crystalline base-
ment, the distribution and volume of intrusive and extrusive volcanic
rocks, and the nature of boundaries between magnetic terranes. In addi-
tion, analysis of regional aeromagnetic data sets can yield estimates of the
depth to bottom of magnetic sources, which puts constraints on the ther-
mal structure of the upper crust (Blakely and Connard, 1989).
Interior Alaska is a collage of more than a dozen amalgamated litho-
tectonic terranes, most having distinctive aeromagnetic signatures. These
distinctive signatures reflect differences in the amount, polarity, and dis-
tribution of magnetic minerals (primarily magnetite and its solid solutions
with ulvöspinel) in the upper crust. The newly merged aeromagnetic data
set (Figure 3), which minimizes contrasts at survey boundaries, allows us
to take a uniform synoptic view of these terranes. A “terraced” map
(Figure 4) accentuates the contrasts between these magnetic terranes and
eases interpretation of their boundaries; the terraced map was con-
structed from a reduced-to-pole version of the aeromagnetic compilation.
As explained in the glossary of this volume, reduction-to-the-pole trans-
forms the observed magnetic anomalies to the anomalies that would be
observed if the Earth’s magnetic field were vertical (Blakely and Connard,
02 General Planning 4/16/07 4:03 PM Page 39
Figure 4. This map displays the result of the terracing operator (Cordell and McCafferty, 1989) applied to the merged aeromagnetic data set of Figure Copper River, Susitna,
3 after reduction-to-the-pole (see text for more details). The solid white lines trace the primary magnetic terrain boundaries, defined by the terrace Minchumina, Holitna,
operator as discontinuities between terrace levels. In Alaska, many of these primary boundaries coincide with mapped lithotectonic terrain boundaries. and Middle Tanana ba-
The dashed white lines represent secondary magnetic boundaries, defined more subjectively as delineating regions with contrasting magnetic anom- sins in the eastern part
aly patterns. These secondary boundaries represent both lithotectonic terrane boundaries and other geologic boundaries within lithotectonic terranes.
1989; Blakely, 1995). The asymmetric effect of the inclination of the
Earth’s magnetic field is relatively minor in Alaska, but reduction-to-the-
pole still aids in clear interpretation of boundaries of causative magnetic
source rocks. The terracing operator (Cordell and McCafferty, 1989) fur-
ther aids in identification of magnetic source boundaries by steepening hor-
izontal magnetic gradients. In other words, a mathematical procedure was
used to emphasize boundaries between rocks with contrasting magnetic
properties. Of course, the magnetic properties themselves are not
of primary interest here; instead, they give us valuable clues about impor-
tant lithologic or tec- Conclusion
tonic boundaries. Strictly
The digitally merged aeromagnetic map of interior Alaska is a valuable
speaking, the terracing
tool for tectonic interpretation and resource exploration. It allows mathe-
method should be ap-
matical manipulation of the data, such as the terracing operator, to
plied to magnetic field
enhance features of tectonic interest. The boundaries on the terraced
data only after the
aeromagnetic map (Figure 4) primarily reflect boundaries between the
pseudogravity trans-
magnetic basement terranes of interior Alaska. These boundaries are
form has been applied
important to both mineral and petroleum exploration as zones of con-
(Cordell and McCaf-
centrated deformation, fluid flow, magmatism, and/or heat flow. Where
ferty, 1989). However,
these boundaries can be traced beneath the postaccretionary cover rocks,
in practice, it provides
they are of particular importance to exploration. In addition, the merged
more useful delineation
compilation allows for a quick reconnaissance view of all basins and for
of shallow magnetic
the extraction of data for coarse 2-D or 3-D modeling.
sources on a regional
scale when directly
applied to magnetic field References
data. Black, P. A., Green, C. M., Reford, and S. W., 1995, A pragmatic approach to conti-
The interior Alaska nental magnetic compilations: 65th Ann. Internat. Mtg., Soc. Expl. Geophys.,
Expanded Abstracts, 773–774.
basement terranes are
blanketed for much of Blakely, R. J., 1995, Potential theory in gravity and magnetic applications: Cambridge
Univ. Press.
the region by postac-
cretionary cover rocks, Blakely, R. J., and Connard, G. G., 1989, Crustal studies using magnetic data, in
Pakiser, L. C., and Mooney, W. D., Eds., Geophysical Framework of the Continental
including the Upper United States: Geol. Soc. of Am. Memoir, 172, 45–60.
Cretaceous Kuskok-
Cordell, L., 1985. Techniques, applications and problems of analytic continuation of
wim and Yukon-Koy- New Mexico aeromagnetic data between arbitrary surfaces of very high relief:
ukuk-Kobuk basins in Institut de Géophysique, Université de Lausanne, Bull. No. 7, 96–99.
the western part of the Cordell, L., and McCafferty, A. E., 1989, A terracing operator for physical property
region and the Cenozoic mapping with potential field data: Geophysics, 54, 621–634.
Kirschner, C. E., 1994, Interior basins of Alaska, in Plafker, G., and Berg, H. C., Eds.,
The geology of Alaska, The geology of North America, G-1, 469–493.
Magoon, L. B., 1993, Alaska Interior, Kandik, Interior Lowlands, and Copper Basin
Provinces, in Powers, R. B., Ed., Petroleum exploration plays and resource esti-
(Kirschner, 1994). The mates, 1989, onshore United States—region 1, Alaska; region 2, Pacific Coast: U.S.
postaccretionary basins Geol. Surv. Bull. 2034-A, A42–45.
of interior Alaska are considered to have minor petroleum potential, mostly con- Meyer, J. F. Jr., and Saltus, R. W., 1995, Merged aeromagnetic map of interior Alaska:
sisting of limited small-scale gas prospects (Kirschner, 1994; Magoon, 1993), U.S. Geol. Surv. Map GP-1014, scale 1:500,000.
but little exploration has been done. For many basins, thickness of basin fill is Meyer, J. F., Jr., Saltus, R. W., Barnes, D. F., and Morin, R. L., 1996, Bouguer Gravity
estimated from gravity models with no drilling constraints. Few exploratory Maps of Interior Alaska: U.S. Geol. Surv. Map GP-1016, scale 1:500,000.
wells have been drilled. Minor oil and gas shows have been found in the Paterson, N. R., Reford, S. W., and Kwan, K. C. H., 1990, Continuation of magnetic
Copper River Basin and unconfirmed oil seeps have been reported in the data between arbitrary surfaces: Advances and applications. 60th Ann. Intl. Mtg.,
Soc. Expl. Geophys., Expanded Abstracts, 666–669.
Middle Tanana Basin (Kirschner, 1994). Detailed analysis of the individual
flight-line profiles in this data compilation can be used to model the geometries Urquhart, T., 1989, Decorrugation of enhanced magnetic field maps: 59th Ann. Intl.
Mtg., Soc. Expl. Geophys., Expanded Abstracts with Biographies, 371–372.
of these interior basins.
39
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:03 PM Page 40

How Basement Lithology
Changes Affect Magnetic
Interpretation
Patrick S. Millegan, Marathon Oil Company
Houston, Texas, USA
D. E. Bird, Bird Geophysical
Houston, Texas, USA
Many people are intimidated by the mag-
netic method, but they are quite willing
to attack gravity interpretation. This is
usually a result of the inclination/decli-
nation issues with magnetics, and the
fact that most people view a gravity map
as a structural map. Magnetic maps are
not structural maps; they are a contoured
representation of magnetization changes
in the geology. This is harder for people
to “visualize” in terms of real rocks and
structure. Gravity interpretation seems
easier because a gravity high usually cor-
General/Planning/Large Scale
40
liminary models prior to undertaking any ity. You will get a magnetic anomaly at up with knowledge of the expected basement is changing in depth. This is
gravity or magnetic interpretation. each edge of the basin. Since there is a structural style (a rift, for example), the quick analysis rather than a quick glance.
direct, mathematical correlation between geologist gets a very simple picture of the
There is a common misconception about anomaly wavelength and depth, you can basin. In fact, the basement rock change Scenario
magnetics: If there are rock-type changes calculate the depth to the contrast. That has helped rather than hindered the
Well-intentioned explorationist (E):
in basement, then magnetics cannot tell would be to the top edge on either side interpretation.
“But the magnetic map showed a low in
anything about basement structures. This of the basin.
our area of interest, so that means there’s
statement is false. Visual inspection of maps must be ac-
sedimentary section rather than base-
This is not very satisfying to the explo- companied by some visual and quantita-
ment, right? Why did we hit basement?”
Magnetic anomalies are a result of two rationist, however, because he is inter- tive analysis ! Linear trends or breaks in
things: ested primarily in the thickness of the trends will correlate either with lithology
Gravity/magnetics specialist (GMS):
1) a lateral contrast in rock structure or sedimentary section or whether there is changes or structure (faults). After a few
“It means that your area of interest is
a structural trap within the basin. But if correlations with known (and expected)
2) a lateral contrast in rock composition lower susceptibility than the surrounding
basement changed susceptibility (Figure geology, the interpreter can make depth
area, not necessarily sedimentary. Does
2), an anomaly would exist with a wave- calculations. Anomalies of similar wave-
If there is no lateral contrast, the mag- that linearity to the low make any sense
length that is a function of the depth to length in an area give depth information
netic field is flat. Consider a very simple, with the regional tectonics?”
the top of the contrast (top basement). for that area. If there are strong anom-
straight-sided graben filled with nonmag- Now the interpreter has three depth alies, probably related to basement lithol-
netic sediments (Figure 1). All of the E: “Not really . . . but you know magnet-
points, two at the edges and one on the ogy changes, and their areal distribution
basement rocks are the same susceptibil- ics . . . it could be a basement rock
basin floor. If these depths are contoured shows changes in wavelength, then the
change.”

200

200
200
200

relates with a structural high (except in

CALCULATED MAGNETICS (nT)

CALCULATED MAGNETICS (nT)

CALCULATED MAGNETICS (nT)
CALCULATED MAGNETICS (nT)
the case of salt), but simple criteria such

100
100
100
100

as that can cause problems if the inter-

preter is not careful.

0
0
0
0

-100
-100

-100
-100

On the other hand, some aspects of the

magnetic method are quite straightfor-

-200
-200

-200
-200

ward. One is the direct mathematical

60000 70000 80000 90000 100000 110000 120000 130000 140000 150000 160000 60000 70000 80000 90000 100000 110000 120000 130000 140000 150000 160000
relationship between anomaly wave- DISTANCE (METERS) DISTANCE (METERS)
length and source depth. Another impor-
tant relationship is that basement lith-

0
0

1000
ology changes typically have amplitudes

2000
2000
2000

of hundreds of nanotesla (nT), but struc-

DEPTH (METERS)

DEPTH (METERS)
DEPTH (METERS)
DEPTH (METERS)

S=0

3000
tural changes usually cause anomalies of

4000
4000
only tens of nT (see Gibson’s magnetic
4000

S=2500

5000
susceptibility versus structure illustration

6000
later in this volume).
6000

6000

7000
S=5000
It is also important to note that, depend- S=2500
60000 70000 80000 90000 100000 110000 120000 130000 140000 150000 160000 60000 70000 80000 90000 100000 110000 120000 130000 140000 150000 160000
ing on magnetic inclination, a fault and a DISTANCE (METERS) DISTANCE (METERS)
lithology change can have distinctly dif-
Figure 1. A 2-D magnetic model of an extremely simplified, straight-sided graben filled with nonmag- Figure 2. The same simple model from Figure 1 with a basement susceptibility change of 5000 micro-cgs
ferent magnetic signatures. One might
netic sediments. The single basement polygon has a susceptibility of 2500 micro-cgs. Magnetic anom- (2500 micro-cgs contrast) included in the center of the basin. A third magnetic anomaly (this time sym-
cause a symmetrical anomaly while the alies are caused by the “shoulders” of the basin. Depth estimates would give depth information at about metrical) is created. Depth estimation would give depths for the shoulders as well as about 6000 m ± 10%
other causes an asymmetrical anomaly. 500 m ± 10% for the shoulders only. in the center of the basin.
This underscores the importance of pre-
02 General Planning 4/16/07 4:03 PM Page 41
GMS: “My point exactly. Did you do any
depth work on the edges of the low?”
E: “Yes, an interpretation came with the
data [digs it out of a file at the bottom of a
pile of seismic sections]. It says 4000 ft.”
GMS: “At what depth did the well
penetrate granite?”
E: “3900 ft, but that’s way over
there. . . .”
GMS: “Magnetics are telling you that
you have two edges at about 4000 ft
within a few miles on either side of your
prospect. You have a linear low-suscepti-
bility zone . . . could be sedimentary,
could be low-susceptibility igneous . . . at
about 4000 feet . . . that doesn’t make
any sense with your seismic interpreta-
tion. Inspection of the magnetic map
shows that all the anomalies are roughly
the same wavelength across your area of
interest . . . high versus low is immater-
ial; the magnetics were trying to tell us
something. . . .”
Moral: Sometimes, just having the data
is not enough, and just looking at it is
not enough, either. You also must ana-
lyze the data. How much analysis
depends on the geologic problem, the
staffing levels, and the time available for
interpretation.
The Concept of Continuation
Magnetic anomalies (on profiles or maps)
become broader (longer wavelength) as
the distance between the geologic source
and the magnetometer increases. This
concept of “continuation” is key to the
interpretation of magnetic data. As L. L.
Nettleton put it in his classic 1976 text:
. . . it is usually possible by simple
inspection of a magnetic map over a
broad area covering a wide range of
depths to separate the map into areas of
shallow, intermediate, and large base-
ment depth by noting the areas of
sharp, intermediate, and broad magnetic
anomalies. . . . The whole process of
basement-depth determination depends
on devising quantitative measurements
of this relationship of sharpness to
depth.
Basement Rocks
From Nettleton (1976), 369–370:
. . . actual basement rocks may be com-
pared with a mildly stirred-up matrix of
components with varying magnetite con-
tent. The condition may be generally
likened to an old-fashioned marble cake
with dark and light batter lightly stirred
together and covered with frosting. The
two colors of batter correspond to base-
ment materials with relatively high and
low magnetite concentrations and the
frosting to the overlying sediments.
Such concepts would seem to make
numerical calculations from geometric
models not very relevant. This is not
true, however, apparently because at
depths comparable with the horizontal
dimensions of [the] units . . . effects
from bodies with irregular boundaries
are effectively simulated by models with
simple geometric forms. Therefore cal-
culated effects from simple models can
be very useful in understanding the mag-
netic effects observed in nature.
Magnetic susceptibility is a measure of
how susceptible the rock material is to
being “magnetized” by the Earth’s mag-
netic field (induced magnetization). This
is also a measure of the amount of mag-
netite (or other minor magnetic miner-
als) in the rocks. Susceptibility is a
Geologic Applications of Gravity and Magnetics: Case Histories
dimensionless unit, expressed in SI or
cgs units. Journals prefer the more
formal SI units, while practitioners and
most software vendors use micro-cgs
(cgs × 10–6) units for modeling. The fol-
lowing comments use micro-cgs units.
An “average” granite contains about 1%
magnetite, which equates to about 2500
micro-cgs units. For comparison, the fol-
lowing table gives a simplistic rule of
thumb for rock susceptibilities:
(cgs × 10–6)
Sedimentary rocks 0–600
Acidic basement rocks 600–5000
Volcanics, mafic rocks 3000–10 000
Pure magnetite 30 000
On a few occasions in this author’s
twenty-two years of magnetic interpreta-
tion, real measurements from basement
outcrop or cores were available for mod-
eling purposes. It was impossible to use
the actual measured susceptibilities to
model geologic structure. This is proba-
bly because outcrops are weathered and
magnetite oxidizes fairly easily, losing
magnetization. Published accounts of
basement from oil fields in the Gulf of
Suez, for example, document wells
encountering tens of meters of “frac-
tured, weathered granite” over “fresh
granite.” The good news in this story is
that the susceptibility values required for
modeling varied within the range
expected of the rock type, and the sus-
ceptibility contrasts between adjacent
units were definitely comparable to the
measured contrasts.
It is impossible to perfectly decipher the
“marble cake” of basement susceptibili-
ties. Try to think of the “simple geomet-
ric forms” in magnetic models as aver-
aged zones of magnetite content. If the same as in Figure 5. The 6000-m
shapes remain simple and comparable in (about 20 000-ft) continuation distance
a suite of models, then conclusions are was chosen to simulate the depth of a
possible. Keep the susceptibilities in the sedimentary basin. Hence, if the granitic
accepted range of rock materials; it is rocks of the Canadian Shield were buried
probably more nearly correct, and it will under 6000 m of nonmagnetic sediment
be easier to “sell” to management. and there were no structures on the
basin floor, then the magnetic map
Basement Continuation Example would look like Figure 6. The magnetic
relief has diminished from 1500 nT to
Figure 3 shows a geologic map of Quebec,
about 800 nT, the major lithologic trends
over part of the Canadian Shield (latitude
are still visible, but the wavelengths have
55–60°N, longitude 70–75°W). There is
no structural development here, only
broadened (smoothed).
outcropping basement rocks. The map
The same upward-continued map is dis-
represents the area with a single pink
played in Figure 7 with a 100-nT con-
color and describes “undifferentiated
tour interval, to give a better represen-
granitic basement.” Another geologic
tation of individual anomalies. The wave-
map (Figure 4) shows more detail and
lengths across the map are roughly
describes “mostly granites, syenites,
equivalent, and calculations would pro-
monzonites, diorites and their gneissic
vide depths of about 6000 m for this
equivalents.” Topographic relief is mini-
hypothetical flat-bottomed basin.
mal, a few hundred meters, or essentially
flat for the purpose of this discussion.
If structural changes were imposed on
this susceptibility regime, anomaly wave-
Figure 5 is a total-magnetic-intensity
lengths would differ across the map, and
anomaly map from the public-domain
depth estimation would provide varying
compilation for the Decade of North
depths that could be contoured as a map
American Geology (DNAG). The contour
of structural (magnetic) basement. This is
interval is 500 nT. The color interval on
illustrated more easily in profile view
this and all the figures in this paper is the
(Figures 8 and 9).
gravity/magnetic convention of purples,
blues, and greens for the magnetic min-
Figure 8 is a cross-section model through
ima, and yellows, oranges, and reds for
the study area. Basement lithology
the magnetic maxima. The magnetic
changes have been “simulated by model-
anomalies average about 1500 nT over
ing simple geometric forms” (after Net-
this area. A visual comparison of Figures
4 and 5 shows a striking similarity in
tleton, 1976). Because this is to illustrate
a point, the curve fitting between the
trends. Even some of the interpreted
observed and calculated fields is only
fault/fracture trends are clear in the
approximate. In Figure 9, structure (a
magnetic map.
straight-sided, 6000-m-deep “basin”) has
been introduced to the same susceptibil-
An upward continuation of 6000 m was
ity scheme. Note the difference in anom-
applied to the total-intensity data (Figure
aly wavelengths on the shoulders of the
6), showing how the data would appear
“rift” versus the deeper basin center.
if the magnetometer was 6000 m above
basement. The contour interval is 500 nT,
41
02 General Planning 4/16/07 4:03 PM Page 42

General/Planning/Large Scale
42

Figure 3. Geologic map of Quebec, Canada, over part of the Canadian

Shield (latitude 55-60°N, longitude 70-75°W). The pink color is described
Figure 4. A more detailed geologic map of Quebec, Canada, over part of
as “undifferentiated granitic basement.”
the Canadian Shield (latitude 55-60°N, longitude 70-75°W). The legend de-
scribes “mostly granites, syenites, monzonites, diorites and their gneissic
equivalents.”
Figure 5. Total-magnetic-intensity anomaly map of Quebec, Canada, over part of the Canadian
Shield (latitude 55-60°N, longitude 70-75°W), from the public-domain compilation for the
Decade of North American Geology (DNAG). The contour interval is 500 nT. The colors repre-
sent purples, blues, and greens for the magnetic minima, and yellows, oranges, and reds for the
magnetic maxima. The magnetic anomalies average about 1500 nT over this area. A visual com-
parison of Figures 4 and 5 shows a striking similarity in trends.

The goal of any magnetic interpretation
is to gain information about the subsur-
face geology. Given planned and well-
acquired aeromagnetic data over an area,
an interpreter should be able to map tec-
tonic (lithologic) provinces, if necessary.
The interpreter also can calculate depth
to magnetic basement and contour it.
With high-resolution aeromagnetics, the
interpreter also may be able to detail a
potential trapping structure.
Whether an explorationist is involved in
prospect analysis or in frontier basin eval-
uation, it may be important to know that
the basement rocks represent the Ken-
oran orogeny, undergoing folding and
granitic intrusion during the Late Arch-
aean. This may impact source rock and
reservoir quality, or thermal history. In
short, knowledge of the regional geology
and tectonics can be gained from mag- Reference
netic data, along with a contoured depth Nettleton, L. L., 1976, Gravity and magnetics in
map of the area being explored. Analysis oil prospecting: McGraw-Hill Book Co.
of the data and an understanding of the
geology (or proposed geologic model) are
key to a successful interpretation.
Analysis is required!
02 General Planning 4/16/07 4:03 PM Page 43

Figure 6. Total-magnetic-intensity anomaly map of Quebec (same area as Figure 5). The contour interval is 500 nT. An upward Figure 7. The same upward-continued total-intensity magnetic anomaly data as in Figure 6, but with a 100-nT contour interval to
continuation has been applied, showing how the data would appear if the magnetometer was 6000 m above basement. The contin- give a better representation of individual anomalies.
uation distance was chosen to simulate the depth of a sedimentary basin. Hence, if the granitic rocks of the Canadian Shield were
buried under 6000 m of nonmagnetic sediment and there were no structures on the basin floor, then the magnetic map would look
like this. The magnetic relief has diminished from 1500 nT to about 800 nT, the major lithologic trends are still visible, but the
wavelengths have broadened.

43
Geologic Applications of Gravity and Magnetics: Case Histories
02 General Planning 4/16/07 4:03 PM Page 44

General/Planning/Large Scale
44

Figure 8. A 2-D cross-section magnetic model through the study area from southwest to northeast across the center of Figure 9. A 2-D cross-section magnetic model (same as Figure 8) with the addition of a simple, straight-sided, 6000-m-
the area. Basement susceptibility changes have been simulated by simple geometric forms. The observed (acquired) deep “basin.” It is the same susceptibility scheme as in Figure 8. Note that the anomaly wavelengths coming from the
magnetic curve is green, while the magnetic curve calculated by the model is red. The attempt at curve fitting is only center of the basin are broader and lower amplitude than those on the shoulders of the “rift.” At least two anomalies are
approximate, but seems to represent the actual data well. distinct enough to provide depth information for the “basin floor.”
03 Basin Studies 4/16/07 4:16 PM Page 45

3
The St. George Basin, Alaska, COST #1 Well: An Example of the
Need for Integrated Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
David A. Chapin, LaCoste & Romberg
SubbaRao V. Yalamanchili, PGS Seres
Paul H. Daggett, ARCO Alaska Inc.
Interpretation of Magnetic Anomalies at Low Latitudes: Potential Pitfalls . . . . . . . . . . . 53

Basin
D. E. Bird, Bird Geophysical
S. A. Hall, J. F. Casey, University of Houston
P. S. Millegan, Marathon Oil Company
Aeromagnetic Constraints on the Basement Structure of the Tunghai Shelf
and the Okinawa Trough in the East China Sea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
S. Okuma, T. Nakatsuka, M. Makino, and R. Morijiri
Geological Survey of Japan
Aeromagnetic Interpretation of Southwestern Continental Shelf of Korea . . . . . . . . . . 63
Czango Baag, Kangweon National University
Studies
Chang-Eob Baag, Seoul National University
Integrated Gravity Modeling of Salt Features in the Mississippi Salt Basin . . . . . . . . . . 68
Rhonda L. Schenk, International Paper Petroleum Co.
John J. Morris, Savannah Oil and Gas L.L.C.
Stuart A. Hall, University of Houston
03 Basin Studies 4/16/07 4:16 PM Page 46
Gravity Applications
E. K. Biegert
Shell Exploration and Technology
Company, USA
Conventional gravimeters are extremely
sensitive weighing devices routinely
capable of distinguishing anomalies less
than 1 part in 107 of the Earth’s ambient
field, and gradiometers, which measure
the first derivative of the gravitational
field, to about 1 part in 3000. Obser-
vations of the Earth’s gravity field are
made with a variety of instruments from
different types of platforms, each with its
own characteristic measurement accu-
racy, spatial resolution, and cost of acqui-
sition. Applications of gravity within the
oil industry are just as varied, impacting
all phases of the business from global tec-
tonic studies to frontier exploration to
prospect selection to seismic processing
to well logging to reservoir monitoring to
site hazard and engineering studies to
environmental analysis. Advances in
three technologies, satellite altimetry,
Global Positioning System (GPS), and
gravity gradiometry, have dramatically
improved our ability to acquire higher
resolution gravity data. This, in turn,
has increased the impact that analysis
of gravity data has on exploration
applications.
In southeast Asia, inexpensive satellite
gravity predicts previously undiscovered
minibasins for the frontier exploration
group, allowing it to relocate its seismic
lines to the more prospective areas.
In the deep-water Gulf of Mexico,
drilling operations are suspended while
Eddy Whopper, a large vortex of swirling
water shed from the Loop Current,
passes the drill ship. The drillers, warned
by predictions from the ocean circulation
model constrained by satellite gravity
determinations of the local geoid, are
ready to resume drilling operations with-
out unnecessary loss of time when the
eddy is predicted to pass.
In the Black Sea, satellite gravity is the
only gravity data available to constrain
geologic hypotheses and to help design a
seismic survey program.
In Peru, airborne gravity and aeromag-
netic surveys enable the consortium of
companies to evaluate the large lease
acreage in the remote, dense jungles.
Seismic programs are scheduled over the
anomalies detected by the airborne sur-
veys, avoiding unprospective areas of the
basin.
In Michigan, an inexpensive land-gravity
survey locates a reef trend along the
flanks of the basin. A 3-D seismic survey
is relocated to evaluate the new prospects.
In the Gulf of Mexico, a 3-D seismic
survey is acquired over a salt dome to
delineate traps along the flanks. A high-
resolution gravity survey acquired simul-
taneously with the 3-D seismic survey is
processed and interpreted before the seis-
mic data are released for processing. The
gravity interpretation predicts a large salt
overhang on one side of the dome and a
nearly vertical face on the other. This salt
model guides the seismic processors in
selecting appropriate prestack depth mi-
gration velocity models, and later guides
3-D seismic interpreters as they work
their prospects closer and closer to the
salt face.
In the North Sea, the top salt reflection is
invisible on the seismic sections. Marine
3-D gravity surveys successfully provide
additional geophysical information, used
Basin Studies
46
to determine the depth to the top of salt, and a potentially dry hole is transformed
thereby reducing structural uncertainties into a commercial success.
and improving the migration velocity
model. In situ measurements of the conversion
of the remaining oil saturation to gas in a
During rough weather in the deep-water Middle Eastern field are unobtainable
Gulf of Mexico, a marine vessel with a with the requisite accuracy from time-
navy gravity gradiometer continues to lapsed pulsed-neutron capture logging.
record high-resolution gravity data to Borehole gravity is able to measure the
interpret the thickness of the tabular salt gas saturation deep into the formation.
bodies for a subsalt exploration play.
Gas suddenly is being produced from a
In the refinery near Houston, a leaking well in Prudhoe Bay. Borehole gravity
buried pipeline excavates a large cavity. detects the extent of coning and moni-
Ground-penetrating radar is unable to tors its size as reservoir production
see through the saturated ground, but an and injection changed in attempts to
engineering gravity survey easily detects control it.
the cavern for repair crews. Seismic sta-
tic corrections derived from gravity data
acquired along seismic lines enabled the
crew to correct its seismic data enough
to obtain refraction and reflection statics
solutions for its lines.
In Oman, a borehole gravity deep-read-
ing density tool monitors a steam flood
in a viscous oil reservoir.
In south Texas, a borehole gravity meter
run in old, cased wells detects bypassed
gas sands. Perforating the zones revital-
izes an old well and adds additional
reserves.
In the swamps of Louisiana, a borehole
gravity survey constrains the salt-flank
geometry away from the small portion of
the salt dome imaged by a vertical seis-
mic profile (VSP) salt-proximity survey.
In drilling locally tight pinnacle reefs in
Michigan, shallow-reading nuclear-den-
sity tools do not allow detection of
nearby favorable porosities. A borehole
gravity survey detects a more porous
zone in the reef, the well is kicked over,
03 Basin Studies 4/16/07 4:16 PM Page 47
The St. George Basin, Alaska,
COST #1 Well: An Example of the
Need for Integrated Interpretation
David A. Chapin
LaCoste & Romberg, Austin, Texas, USA
SubbaRao V. Yalamanchili
PGS Seres, Houston, Texas, USA
Paul H. Daggett
ARCO Alaska Inc., Anchorage, Alaska, USA
EDITOR’S NOTE: This famous exploration “horror story” resulted
from ignoring gravity and magnetic data. An expensive lesson
to learn, this single event changed the way the oil industry
does geophysical integration. It is not enough just to have
gravity and magnetics; you must evaluate it and use the
results! Listen up—the next time it may be you.
Abstract
The 1976 St. George Basin COST #1 well, Bering Sea, Alaska, produced
unexpected results. Twenty oil companies participated in the drilling of
this expensive stratigraphic test, and the well was expected to encounter
a thick nonigneous sedimentary section in the deepest portion of the
basin. Instead, the well drilled through more than 1000 m (3300 ft) of
Tertiary basalt and other volcanic extrusives before it was abandoned.
Gravity and magnetic data were available before the well site was chosen,
but the information apparently did not impact the site selection. A recent
reexamination of the data available in 1975 reveals that significant grav-
ity and magnetic anomalies occur over the well site. These data point to
an interpretation that the St. George COST #1 well is located on a col-
lapse feature associated with an igneous calderalike structure. The expe-
rience gained from this well demonstrates the need to integrate all
available types of data prior to making exploration decisions.
Introduction
In July 1976, a Continental Offshore Stratigraphic Test well (COST well)
spudded in the St. George Basin in the southern Bering Sea of Alaska
the drilling. This interpretation would have had an impact on the location
of the COST #1 well.
Available Data and Interpretation Methods
By the fall of 1975, two major potential-fields data sets existed for the St.
George Basin:
1) Marine gravity and magnetic data which were collected along seismic
lines in nonexclusive or group-shoot seismic surveys by Petty-Ray
(1971), Seiscom-Delta (1975), and Dresser-Olympic (1975). The grav-
ity and magnetics data all were acquired by Tidelands Geophysical (G.
Kelo, pers. comm.) and were available from the various seismic con-
tractors on a nonexclusive basis. Each of these three surveys covered
the area of the COST #1 well. Austin and Hopkins (1975) had pre-
sented to the public an interpretation of the Petty-Ray data, a regional
data set, prior to the drilling of the COST #1 well.
Figure 1. Index map of the St. George Basin, Alaska. 2) Public-domain marine gravity and magnetics collected and published
by NOAA (Pratt et al., 1972; see Figures 2 and 3). These data were
acquired in 1970 along single-fold seismic lines spaced 8 km apart.
(Figure 1). Atlantic Richfield Company was the operator for a twenty-part-
Pratt et al. (1972) make the following comment regarding the area
ner consortium of oil companies. (The other partners were Amerada
that later became the site of the St. George COST #1 well:
Hess, Amoco, British Petroleum, Cities Service, Conoco, Exxon, Getty,
Gulf, Marathon, Mobil, Pennzoil, Phillips, Shell, Skelly, Standard Oil of
California, Sun, Tenneco, Texaco, and Union Oil of California)
(Anonymous, 1976). The purpose of the St. George COST #1 well was to
test the stratigraphic section in this previously unexplored offshore basin,
in preparation for a federal outer continental shelf (OCS) lease sale origi-
nally scheduled for 1977. The COST well was located on a seismically
imaged synclinal structure (Worrall, 1991) predicted to contain a com-
plete Mesozoic to Cenozoic stratigraphic section for the basin. The well
was scheduled to drill 4571 m (15 000 ft) of section. It penetrated 3163
m (10 380 ft) of Tertiary marine clastic rocks and 1034 m (3391 ft) of
Middle Eocene and older volcanic rocks before it was abandoned at about
4197 meters (13 771 ft) (Figure 8). This $16.5 million well (1976 dol-
lars) was one of the most expensive drilling operations undertaken up to
that time. The OCS sale subsequently was postponed and a second COST
well was not drilled in the St. George Basin until 1982.
By late 1975, several gravity and magnetic data sets existed for the St.
George Basin. Unfortunately, these data seem to have had little influence Figure 2. Free-air gravity map from Pratt et al. (1972). Acquired by the NOAA ship
on the predrill interpretation of the area. We will show that if the gravity
Rainier in 1970. Tracklines were 8 km apart northeast-southwest, tie lines were 24 km
apart north-south. Contour interval is 10 mGal. Notice the large 50-mGal positive
and magnetic data had been integrated with the seismic data, the volcanic anomaly at the future COST #1 well location (blue dot). This map provided an early
rocks encountered in the well might have been predicted in advance of hint that something was unusual at the well location.
47
Geologic Applications of Gravity and Magnetics: Case Histories
03 Basin Studies 4/16/07 4:16 PM Page 48

Basin Studies
48

Figure 3. Marine magnetic map from Pratt et al. (1972). Contour interval is 100 and 300 nT. Notice the 591-nT anomaly at the
future COST #1 well location (blue dot). Comparing this map to Figure 5a, it is clear that most of the major magnetic features can
be seen in this data set. This map provided a clear warning sign that unusual magnetic character existed at the future well loca-
tion. Pratt et al. (1972) noted this specific location in their article as “conspicuous” and warned that it “may reflect an underlying
igneous structure.” The drilling results four years later proved that Pratt et al. correctly interpreted this area.
Figure 4. Compiled marine magnetic map from Bailey et al. (1976), available before the COST #1 was spudded. The well is posted on
this map for reference (blue dot). Notice the large, 400-nT anomaly at the well site. The ship tracks shown in green on this map come
from public-domain data collected between 1964 and 1973. Comparing this map to Figure 5a, it is clear that most of the major magnetic
features can be seen in this sparse data set. Once again, the magnetics indicate unusual magnetic character at the future well site.

The only conspicuous anomaly in the south is a peak of +591 g in the COST #1 well was spudded, contained public-domain data col- Table 1. Specifications of Aero Service St. George Basin
the SW corner which may be part of a weak E-W trend. This peak lected between 1964 and 1973. Even at a regional scale, the magnetic Aeromagnetic Survey.
corresponds to a gravity high and may reflect an underlying igneous anomaly at the COST #1 well site is quite noticeable (see Figure 6d). Survey number: 76-2
structure.
Date flown: June 6, 1976 – August 10, 1976
2) Aeromagnetic data flown by Aero Service in 1976 (see Figures 5a and
This statement proved to be an accurate prediction of the subsurface Magnetometer Tail-mounted alkali vapor magnetometer
5b). These data were of special interest because they were acquired
geology as encountered by the COST #1 well, drilled four years later.
with modern, high-resolution specifications (Table 1). Acquisition Navigation Loran C, radar and barometric altimeters, Doppler radar
took place during the drilling of the COST well, and therefore had no Flight altitude 305 m (1000 ft) barometric
Two additional magnetic data sets became available in 1976: impact on the well-site decision.
Line spacing 3 km traverses at 28°, 12 km ties at 118°
1) Marine magnetic data collected and published by the U.S. Geological Along-line sampling Digitally recorded, 46 m (152 ft)
Survey (Bailey et al., 1976). This map, released a few months before
Linear coverage 35 225 line km (21 888 statute miles)
03 Basin Studies REV copy:03 Basin Studies 5/18/09 3:45 PM Page 49
Figure 5a. Regional aeromagnetic map from Aero Service survey 76-2. Hot colors are magnetic highs; cool colors are mag-
netic lows. Here we have the clearest picture yet of the intense magnetic anomalies at the well site. When the COST #1 well
was drilled, these proved to be the signature of igneous rocks.
Figure 5b. Aeromagnetic map from survey 76-2 showing a close-up of the COST #1 and #2 well locations and profile shown
in Figure 6.
Digital-processing techniques commonly were used for gravity and
magnetic data interpretation in 1975. For instance, Werner deconvo-
lution profile plots routinely were generated to determine magnetic
source depths from a series of dike and contact geometry solutions
(Werner, 1953; Hartman et al., 1971). Dip and susceptibility infor-
mation also can be obtained from the Werner deconvolution picks in
addition to depth estimates (e.g., Figures 6b and 7). Two-dimensional
forward-gravity and magnetic-modeling programs also were widely
available in 1975 (e.g., Talwani, et al., 1959; Talwani and Heirzler,
1964) for estimating lithology, depth, and geometry of subsurface
bodies. Even without digital processing,
several manual techniques also were
available to determine the depth to an
anomalous body from analog magnetic
data (e.g., Peters, 1949).
Conclusions
The peak-to-trough magnetic anomaly
located at the St. George COST #1 well is
nearly 1000 nT—one of the largest mag-
netic anomalies in the entire Bering Sea
(DNAG, 1987). The synform imaged in
the St. George seismic appears to be a col-
lapse feature over a caldera or other type
of igneous center (see Figure 6d). The lay-
ered events within the feature proved to
be sills and flows, not sedimentary hori-
zons, as originally interpreted. The gravity
data display a high at this location (Figure
6a), indicating a very dense body. The
magnetic data also show an intense high,
suggesting the presence of a large amount
of magnetite-rich rocks—a probable indi-
cation of igneous rocks. Depth estimates
from both Werner deconvolution (Figures
6b and 7) and forward modeling (Figure
6c) place an igneous source at a depth
corresponding to the top of the igneous
rocks found in the well.
If the potential-fields data had had a
more prominent role in the interpretation
of the St. George Basin before the well
was drilled, this large igneous center
might have been recognized in the prognosis. Unfortunately, the gravity
and magnetics data sets apparently were overlooked during the hectic
period when the well-site decisions were being made. In any event, the
unexpected result of the COST #1 well pointed to a need for fully inte-
grating all available data prior to making final exploration decisions. The
use of seismic data alone in frontier basins clearly can lead to erroneous
and expensive drilling errors.
In the 1970s, common-depth-point seismic-reflection profiling
became the geophysical exploration tool of choice because it offered bet-
Geologic Applications of Gravity and Magnetics: Case Histories
ter depth resolution compared with any other geophysical method. As a
result of the increased reliance on seismic data, the exploration usage of
gravity and magnetic data declined. The results of the St. George COST
#1 well reminded many oil companies of the pitfalls of relying too heav-
ily on a single geophysical tool. Although seismic-reflection profiling does
offer better depth resolution than other exploration methods, impedance
contrasts caused by igneous layers frequently cannot be distinguished
from nonigneous sedimentary rocks.
Today, 3-D seismic-reflection surveys are used more frequently for
detailed hydrocarbon exploration. With the advent of workstation tech-
nology, explorationists now can construct complex structural and strati-
graphic interpretations based largely on seismic data. Once more, we see
a dramatic decline in the use of potential-fields data, as the industry relies
heavily on a single geophysical tool. Has our industry forgotten the
lessons of the past? Are further unexpected surprises awaiting us because
of our failure to integrate all available data in exploration decisions?
Acknowledgments
We would like to thank ARCO Alaska Inc. and ARCO Exploration and
Production Technology for allowing us to publish this paper. We also
would like to thank GEONEX Aero Service for allowing us to use and
display its aeromagnetic data. Finally, we wish to acknowledge the con-
tributions of the late Jim Schwartz, who gave us the inspiration for this
case study.
References
Anonymous, 1976, ARCO spuds south Bering Sea strat test: Oil and Gas J., July 12, 36.
Austin, C. T., and Hopkins, M. G., 1975, Geophysics and geology of the Bering Sea
(abs.): Geophysics, 40, 167. Presented at the 44th Ann. Internat. Mtg., Soc. Expl.
Geophys.
Bailey, K. A., Cooper, A. K., Marlow, M. S., and Scholl, D. W., 1976, Preliminary resid-
ual magnetic map of the eastern Bering Sea Shelf and parts of western Alaska, USGS
MF-716, 1:2,500,000.
DNAG, 1987, Decade of North American Geology magnetic map of North America,
1:5,000,000, Geol. Soc. Am.
Hartman, R. R., Tesky, D. J., and Friedberg, J. L., 1971, A system for rapid digital aero-
magnetic interpretation: Geophysics, 36, 891–918.
Peters, L. J., 1949, The direct approach to magnetic interpretation and its practical appli-
cation: Geophysics, 14, 290–320.
Pratt, R. M., Rutstein, M. S., Walton, F. W., and Buschur, J. A., 1972, Extension of
Alaskan structural trends beneath Bristol Bay, Bering Shelf, Alaska: J. Geophys.
Res., 77, 4994–4999.
49
03 Basin Studies 4/16/07 4:16 PM Page 50

Basin Studies
50

Figure 6c. A 1976 postmortem 2-D modeling interpretation by J. Schwartz at Gulf Oil Corp. (Schwartz, pers. comm.), using the
marine gravity and magnetic data collected, along with seismic data used to site the well.

Figure 6a. Profile plotted to display Seiscom-Delta marine gravity and magnetics (Schwartz, pers. comm.), as well as the Aero
Service aeromagnetics.

Figure 6d. The 1991 seismic interpretation by D. Worrell (1991). Figures 6a–6d are all scaled the same horizontally.

Figure 6b. A 1995 Werner deconvolution plot and interpretation by S.V. Yalamanchili, performed without benefit of well or seismic
data. Although this is a modern plot, it is an attempt to reproduce the technology available in 1976. Notice how remarkably close the
interpretation is to the other interpretations.
03 Basin Studies 4/16/07 4:16 PM Page 51

Magnetic Reduction-to-the-
Pole at Low Latitudes
W. C. Pearson, Pearson Technologies
Lakewood, Colorado, USA

EDITOR’S NOTE: The appearance of any

gravity or magnetics map is of pri-
mary importance to the interpreter.
Map displays are the single most
important tool in passing interpreted
information along to management
or a client. Improper or inexact data
reductions or enhancements can
Figure 7. A close-up of Figure 6b—Werner deconvolution and interpretation. The automated solutions labeled “x” give wrong impressions, as all view-
are dike solutions; those labeled “–” are contact solutions. Averaged solutions have bigger symbols and a dip ers will be making their own spur-
arrow. Selected magnetic susceptibilities, in μcgs units, are annotated in italics. At greater than 2000 μcgs units, of-the-moment interpretations.
the susceptibility of interpreted volcanic intrusives near the well site is more than double the magnetic-basement
susceptibilities of approximately 1000 μcgs. This indicates that the magnetic anomaly at the well site is caused by Although a rule of thumb that sug-
a more magnetite-rich intrusive rock than that of the host rock (basement). These two groups of susceptibilities gests a reduction to the equator
are consistent with values for mafic igneous rocks such as basalts. would be appropriate, the resulting
map (heavily biased by east-west
trends) might obscure the true geo-
logic trends. Pearson’s discussion is
Talwani, M., and Heirtzler, J. R., 1964, Computation of magnetic anomalies caused by a technical guide to the proper han-
two-dimensional structures of arbitrary shape: in Parks, G.A., Ed., Computers in the
mineral industries: Stanford Univ. Press, 9, 464–480.
dling of reduction-to-the-pole at low
Talwani, M., Worzel, J. L., and Landisman, M., 1959, Rapid gravity computations for
magnetic latitudes.
two-dimensional bodies with application to the Mendocino Submarine Fracture
Zone: Jour. Geophys. Res., 64, 40–59.
The distortion of magnetic anomalies near
Werner, S., 1953, Interpretation of magnetic anomalies at sheet-like bodies: Sveriges
the magnetic equator, where the earth field
Geol. Undersok., Ser. C. Arsbok. 43 (1949), no. 6.
vector is close to horizontal, is not well
Worrall, D. M., 1991, Tectonic history of the Bering Sea and the evolution of Tertiary understood in all corners of the exploration
strike-slip basins of the Bering Shelf: Geol. Soc. Am. Spec. Paper 257. Figure 8. Lithology log of the St. George COST #1 well, from Worrell (1991). Notice
industry. Unfortunately, a proper correction
that the well never penetrated below the Tertiary and that more than 1000 m (3000 ft)
of volcanic rocks was drilled before the well was abandoned. for anomaly distortion has not been adop-
ted fully. Therefore, many forms of mag-
netic interpretation ignore or disguise the
problem, causing a misleading picture of
low-latitude magnetic-line profile or map
displays.

Unsatisfactory applications of magnetic sur-

veys in low-latitude areas are particularly
disappointing because much regional geo-
logic and tectonic work needs to be per-

51
Geologic Applications of Gravity and Magnetics: Case Histories
03 Basin Studies 4/16/07 4:16 PM Page 52

Basin Studies
52
formed in the band near the magnetic
-4000.
-2000. 0. 10000. 20000. 30000. -4000.
-2000. 0. 10000. 20000. 30000.
equator, where low-cost, easy-access
aeromagnetic data should be used.
Although forward modeling of magnetic 30000. 30000. 30000. 30000.
profiles and map grids can be performed - 10 1 0
1 0

with relative success, many of the most - 10 - 20 - 20

0
2 0

-1
3

- 30

10
0

- 210
0- 1200 0

0
0

common interpretation tools, such as 0 6078900

3
4
5 0- - 1
0
- 4
0
0 -2 0
2 0
- 3 0
1 0 - 1 0 0
0 -4 0 -2 0
- 5
20

- 3 0

20
- 5 0 - 6 0 - 7 0
20000. 20000. 20000. 20000.

-010
- 8 0

-20
90 - 60

-
0
- 7

01
map filtering with color and shaded relief

0
9
782

0
31
-10

-3

-30
-43-010

65420
- 5 0

0
00
- 0

561000
- 30 4

70
0020

0
-1

-0
2

-20
80

-3
90
8
90
displays and depth estimation by Werner

0 -20

- 10
0

- 20

1- 0
- 40

90
0 7 6043200
301 0 - 8
-2 20 05
- 1 0 - 2 0
-3 0

00
0 - 11 1 0 0 2 0 - 1 0 0
1 0

and Euler (particularly 3-D Euler) tech- 3 0 0

4 5 0 6 0 0
3

-2
10000. 10000. 10000. 0 0
10000.
-3

0
-1
7 0

40

20
0
6 0 0

- 20

20
5

0
0 4 0

- 10
30

niques, are compromised or totally mis-

0
1
1
- 10 0

0
2

leading. Figure 1 is a 3-D block model

0
1

0. 0. 0. 0.
computed at the magnetic pole. Anomaly -2000. -2000. -2000. -2000.
-4000. -4000. -4000. -4000.
gradients correspond to block edges or
faults. If a color display cannot be cre- -4000.
-2000. 0. 10000. 20000. 30000. -4000.
-2000. 0. 10000. 20000. 30000.

ated where highs are over the targets

and linear trends are above faults, irre- Scale 1:4000 Scale 1:4000
spective of strike, the value of magnetic 2000. 0. 2000.4000.6000.8000.10000. 2000. 0. 2000.4000.6000.8000.10000.
data is lost. Magnetic results must be
transformed to appear the same, no mat-
ter what the latitude. Otherwise, geolo- Figure 1. Three-dimensional magnetic block model at the magnetic poles, I = 90°. Figure 2. Three-dimensional magnetic block model at low latitude, I = 10°.
gists and seismologists rightfully will
-4000.
-2000. 0. 10000. 20000. 30000.
continue to eye magnetic data with a 1) Design the survey with survey lines buffer area around the actual survey and
certain amount of suspicion. close enough so that the mapped smearing it across the survey as north-
total-magnetic-intensity grid is a south stripes. 30000. 30000.
10
Proper low-latitude reduction-to-the-pole good representation of the field. 0

1
20
(RTP) really is a two-stage operation. Once the survey edges are extended in a 304
0

10
0 4 0 503 0
First, the magnetic grid must be ex- 2) If possible, avoid ending the survey controlled manner to allow a good fast 0 2--04
10 1 0
0- 20--65
3000 2 0 - 1
20000. 0 0 20000.
- 0

- 30 0
- 50 -4
tended carefully, and then a frequency- over large anomalies. Fourier transform, then the standard RTP

- 10

-3
- 20
0
- 40
domain filter can be applied. Some formula can be applied. Actual surveys - 50

10
- 4600
00 - 2 0

-
0 -230 10-40-10
-
20
0 50 - 3
algorithms work in some cases and mis- 3) Orient survey lines and tie lines at and a map-grid power spectrum of low- 10000.
0 50 20 0 10000.

0
40 3

10
10
erably fail in others, even though the an oblique angle to the Earth’s decli- latitude magnetic models have notches 20
10
earth field vector is about the same. nation. where the spectrum represents faults par-
0. 0.
Even RTP filtering of a relatively high-lat- allel to the magnetic declination. On a -2000. -2000.
-4000. -4000.
itude magnetic survey can be disappoint- 4) If possible, compute a few 2-D and low-latitude total-magnetic-intensity map,
ing if the survey is small and contains 3-D models to understand the any faults nearly parallel in strike to the -4000.
-2000. 0. 10000. 20000. 30000.

large anomalies on the edge that are
sampled only partially.
The first half of the RTP operation in-
volves extending the survey to a full rec-
tangle in a manner that avoids improper
edge effects—especially low-frequency
ones. Some rules of thumb on the sur-
vey-design side of the picture can help:
effect of the low-magnetic vector Earth’s declination are almost invisible.
inclination. In fact, low-latitude (or reduced-to-equa- Scale 1:4000
tor) magnetic maps seem to have a series 2000. 0. 2000.4000.6000.8000.10000.
Perhaps the most robust and universally of near east-west ovals with no definitive
available approach to edge-effect condi- north-south gradients.
Figure 3. Three-dimensional magnetic block model at the magnetic equator (or reduced to the equator),
tioning involves extrapolation over a lim-
I = 0°.
ited distance to a flat level that is the Figure 2 is the same block model as
mean of the survey grid. In this way, the Figure 1 but computed at a 10° inclina-
RTP operation will avoid taking low-fre- tion. This is because the magnetic
quency noise from the extrapolation domains in the rock line up head to tail
03 Basin Studies 4/16/07 4:16 PM Page 53
along the faults paralleling the earth field
vector and have very little chance to con-
centrate on body corners to create anom-
alies. Reduction-to-the-pole filtering is an
amplitude- and phase-changing operation
that assumes nothing about depth, body
shape, or altitude (but does assume a
known remanence). The weak spots in
the spectrum are boosted in amplitude
by the RTP filter, presenting the most
sensitive part of the operation. Because
noise is present in any survey, boosting
parts of the spectrum by factors of 4 to
50 or more when the signal-to-noise ratio
is weak will cause the catastrophic arti-
fact of north-south striping so common to
many RTP algorithms. This problem can
be avoided by limiting the amount of
amplification.
An occasionally applied process of
“reduction-to-the-equator” only worsens
a bad anomaly distortion and should be
avoided. Figure 3 is the block model at
the magnetic equator. Low-cut filtering
or reduction part of the way to the pole
has been tried unsuccessfully. They leave
the resulting grid, and certainly the even-
tual high-pass filtered displays, with
incorrect and misleading anomalies.
Once the above approaches have been
followed and the grid has been extended
properly and filtered with an appropriate
amount of whitening, how can the inter-
preter be sure the map is right? Although
this question can be answered by apply-
ing the same operator to 3-D theoretical
models with known geology, the fact that
the question is so common underlies the
need to improve and standardize the
low-latitude RTP filter so potential-field
experts and their geologic/geophysical
customers can trust and use the results.
Interpretation of Magnetic Anomalies
at Low Latitudes: Potential Pitfalls
D. E. Bird, Bird Geophysical, Houston, Texas, USA
S. A. Hall, University of Houston, Houston, Texas, USA
J. F. Casey, University of Houston, Houston, Texas, USA
P. S. Millegan, Marathon Oil Company, Houston, Texas, USA
EDITOR’S NOTE: Many people are “afraid” of magnetics because it
seems complicated by inclination and declination effects. It is
very important for the magnetics interpreter to understand
what to expect for the geologic and geographic setting. This
paper deals with pitfalls inherent to data coverage and incli-
nation/declination effects. Often you cannot understand the
tectonics from a casual glance at a magnetic map; analysis is
mandatory.
Abstract
The vector nature of the Earth’s magnetic field dictates that interpreters
must take care to understand pitfalls related to the orientation of the field
(i.e., magnetic inclination and declination), and the relationship of the
magnetic field to a region’s geology. The case history presented here
demonstrates one such pitfall. Present models for the formation of the
Grenada Basin vary from north-south extension to northeast-southwest
extension to east-west extension. Gridded magnetic anomalies over the
basin provide a picture of the Earth’s field that contributes to this spec-
trum of possible extensional origins.
The Grenada Basin is a back-arc basin located near the eastern edge
of the Caribbean Plate. The basin is bounded on the east and west by the
roughly north-south-trending active Lesser Antilles and remnant Aves
Ridge Island Arcs, respectively. Although this physiography, as well as
gravity data, supports formation by near east-west extension, magnetic
anomalies over the basin exhibit predominantly east-west trends. The
crust of the Grenada Basin and of other back-arc basins forms similarly to
the crusts of ocean basins. If the observed magnetic anomalies over the
basin are produced by sea-floor spreading, then the orientation of exten-
sion may be complex. Extension in most back-arc basins is roughly nor-
mal to their trenches and subduction zones, but some basins appear to
Geologic Applications of Gravity and Magnetics: Case Histories
exhibit oblique extension. A careful interpretation of magnetic profiles
reveals low-amplitude magnetic anomaly trends, oriented subparallel to
the island arc, over the southern part of the Grenada Basin, which sup-
ports a model for basin development by near east-west extension.
Introduction
Theories on the development of the Grenada Basin generally agree
that it was formed by sea-floor spreading in early Cenozoic time
(Bouysse, 1988; Boynton et al., 1979; Donnelly, 1975; Pindell and
Dewey, 1982; Shurbet, 1976; Uyeda and Kanamori, 1979; Uyeda, 1982;
and Westbrook, 1975). Late Tertiary tectonic forces caused the northern
half of the Lesser Antilles Island Arc to bifurcate into an older outer arc
and younger, inner arc. In general, back-arc basins are formed in response
to a reorganization of tectonic plates. Extensional forces resulting from
this reorganization can cause an island arc to split, followed by the accre-
tion of back-arc crust by sea-floor spreading.
The direction of extension shown in existing models for the formation
of the Grenada Basin varies from north-south (Pindell and Barrett, 1990)
to northeast-southwest (Bouysse, 1988) to east-west (Tomblin, 1975).
The eastern Caribbean Plate boundary between the North and South
American Plates is a subduction zone that is oriented generally
north-south (Figure 1). Similarly, the trends of the Aves Ridge, Grenada
Basin, and Lesser Antilles region are oriented north-south. East-west
extension is suggested by these trends (Tomblin, 1975). However, mag-
netic anomalies over the Grenada Basin exhibit predominantly east-west
trends, suggesting that the basin may have formed by north-south exten-
sion (Figure 2).
Magnetic anomaly patterns over some of the world’s back-arc basins
are organized poorly, indicating changing patterns of sea-floor spreading;
however, the patterns over other back-arc basins are well defined, indicat-
ing preferred directions of sea-floor spreading. A review of back-arc basins
located along the west Pacific margin demonstrates this relationship
(Weissel, 1981). The orientation of most linear anomaly trends over back-
arc basins is generally subparallel to their associated subduction zones.
Tomblin (1975) describes a possible scenario for east-west extension
that involves a shift of the Aves Ridge westward, relative to the subduc-
tion zone, with the formation and subsequent spreading from a north-
south-oriented median ridge (Figure 3a). He reports that no such ridge
has been observed.
Pindell and Barrett (1990) describe a model in which the Leeward
Antilles have been coupled to the northern edge of the South American
Plate (Figure 3b). North-south spreading in the Grenada Basin occurs as
a result of oblique convergence with the South American Plate. In this
53
03 Basin Studies 4/16/07 4:16 PM Page 54

Basin Studies
54
model, the Leeward Antilles was part of the Aves Ridge prior to the for-
mation of the basin and represents fragmentation of the arc as the
Caribbean Plate progressed eastward. Pindell and Barrett (1990) suggest
that the general east-west orientation (i.e., perpendicular to the island arc)
of magnetic anomalies over the basin supports north-south extension.
Bouysse (1988) describes a possible mechanism for extension, similar
to Pindell and Barrett’s, in which coupling of the southern part of the
Lesser Antilles with the South American Plate also precedes opening of
the basin (Figure 3c); however, the mechanism of back-arc spreading in
this model is similar to a mechanism described by Poehls (1978). Bouysse
(1988) further suggests that sea-floor spreading in the Grenada Basin was
oriented northeast-southwest at the onset of the Cenozoic, and was seg-
mented in a manner such as described by Tamaki (1985) for the Sea of
Japan Basin. Initial spreading was in the southernmost part of the basin,
gradually progressing northward as the Caribbean Plate moved eastward
between the North and South American Plates.

Data
The database for this study includes gridded gravity, bathymetry, and mag-
netic data as well as shipborne magnetic profiles. Gravity and total-inten-
sity magnetic anomalies were compiled in 1987 by the Geological Society
of America Decade of North American Geology (DNAG) Committees on
the Magnetic and Gravity Anomaly Maps of North America, and were
gridded at 6 and 2 km (Figures 4 and 2), respectively. Gridded bathyme-
try data were extracted from the ETOPO5 data set and regridded from 5
minutes to 9 km (Figure 5). Magnetic anomaly profile locations are
shown in Figure 6. All data used in this study are available from the
National Geophysical Data Center (NGDC).
The Pitfall
The shape of a magnetic anomaly depends on several factors, including
magnetic inclination and declination, source-body geometry, and the
orientation of the source body (Sharma, 1976). The following discussion
assumes that source bodies are symmetric, have stronger magnetizations
than surrounding media, and have no components of remanent magneti-
zation. These assumptions are not necessarily exactly correct, but in the
absence of evidence to the contrary, they are reasonable and typical of
those made by interpreters in this kind of analysis.
At either magnetic pole, a magnetized source body will produce an
anomaly high centered over it. At the magnetic equator, a source body
will produce a symmetric anomaly low over it. Elsewhere in the world, a
source body will produce an asymmetric high-low anomaly pair. The loca-
Figure 1. Physiography of the eastern Caribbean with 2-, 4-, and 5-km isobaths con-
toured (after Bouysse, 1984). The outline of the study area, trace of the subduction
zone, and strike-slip fault zones which define the North American/Caribbean and
South American/Caribbean Plate boundaries are displayed. Heavy dashed lines indi-
cate probable locations for plate boundaries. The inner and outer arcs are repre-
sented by dashed and dotted lines, respectively.
Figure 2. Total-intensity magnetic anomalies over the study area. The contour interval
is 50 nT. Gridded data (2 km) were compiled in 1987 by the Geological Society of
America Decade of North American Geology Committee on the Magnetic Anomaly Map
tion of the source body beneath this anomaly pair depends on the mag-
of North America.
netic inclination (or latitude). For example, if the source body is at 45°
magnetic latitude, then the center of the body would be located beneath
the inflection point between the high-low anomaly pair. This anomaly
trate this dependence, four profiles have been calculated, using two geo-
would shift toward an anomaly low over the source as the latitude
magnetic inclinations and two strike directions for a 2-D model (Figure 7).
decreases from 45°, and would shift toward an anomaly high as the lati-
Most plate reconstructions place the leading edge of the Caribbean plate
tude increases from 45°. In the northern magnetic hemisphere, the
at approximately 12° latitude at the time the Grenada Basin formed
“high” part of the high-low pair is located south of the “low” part, and in
(Duncan and Hargraves, 1984; Ghosh et al., 1984; Pindell et al., 1988;
the southern magnetic hemisphere, it is located north of the “low” part.
Ross and Scotese, 1988). To simulate a remanent inclination and, it is
Note that magnetic latitude is not coincident with geographic latitude,
hoped, to resolve anomalies produced by this remanent field, a paleo-
except in a general way.
magnetic inclination of 23° is calculated (Sharma, 1976). Present mag-
The relationship between magnetic inclination and source body is
netic inclination in the Grenada Basin is 43°. Two profiles were oriented
complicated further by source-body geometry and orientation. To illus-
03 Basin Studies 4/16/07 4:16 PM Page 55
Figure 3. (a) Possible east-west extension caused by a westward shift of the Aves
Ridge for the opening of the Grenada Basin, as proposed by Tomblin (1975). Large
arrows indicate the relative motions of the North American, Caribbean, and South
American Plates. Small arrows indicate the directions of extension for the formation of
the basin. (b) Possible north-south extension for the opening of the basin, as pro-
posed by Pindell and Barrett (1990). (c) Possible northeast-southwest extension for
the opening of the basin, as proposed by Bouysse (1988).
Figure 4. Free-air gravity anomalies over the study area. The contour interval is 10
mGal. Gridded data (6 km) were compiled in 1987 by the Geological Society of
America Decade of North American Geology Committee on the Gravity Map of North
America.
south-north over an east-west-trending body, and the other two were ori-
ented west-east over a north-south-trending body.
Using an inclination of 43°, calculated amplitudes decrease about 57%
(from about 300 to 130 nT) from south-north to west-east. A more dra-
matic decrease, and more important to this study, is observed from south-
north to west-east calculations using a 23° inclination. The decrease in
amplitude is about 86% (from about 290 to 40 nT).
Interpretation
Figure 8 displays selected magnetic anomaly profiles over the southern
part of the basin, with some of the major anomaly trends indicated. Note
Geologic Applications of Gravity and Magnetics: Case Histories
Figure 5. Bathymetry of the study area. The contour interval is 100 m. Gridded data
(5 minute) were compiled by Lamont-Doherty Geological Observatory of Columbia
University.
that trends, although discontinuous, are oriented generally north-south,
or subparallel to the island arc and the trench line of the subduction zone.
Anomalies correlated from the observed magnetic profile data over the
southern part of the basin exhibit amplitudes near 40 nT. There does not
appear to be structural relief on the acoustic basement surface which
would produce these anomalies (Speed et al., 1984). Furthermore, in this
comparison of relative anomaly amplitudes, the smaller amplitude over a
north-south-oriented body suggests that magnetization of the body is
caused primarily by the hypothetical remanent field (or 23° inclination).
That is because any effect of the inducing field would only increase the
55
03 Basin Studies 4/16/07 4:16 PM Page 56

Basin Studies
56

Figure 8. Magnetic anomaly trends over the study area from profile data. Anomaly
highs are indicated by solid lines and anomaly lows are indicated by dashed lines.

Figure 6. Shipboard magnetics data coverage for the study area. Anomaly profiles for
west-northwest-oriented ship tracks outlined by the dashed box are displayed in
Figure 8.
amplitude, because the north-south-oriented body produces a larger-
amplitude anomaly at 43° magnetic inclination. The effect of magnetiza-
tion contrasts caused by geomagnetic polarity reversals also increases the
relative amplitude of anomalies.
Figure 7. Magnetic anomalies calculated for two profile directions (south-north and
west-east). For profiles 1 and 3, present-day magnetic inclination (43˚) and declination produced by sea-floor spreading and indicate a near east-west direction of
(–11˚) are used, and paleomagnetic inclination (23˚) and declination (0˚) are used for
profiles 2 and 4. In each calculation, the same 2-D causative body is used: 5 km thick,
extension and opening of the Grenada Basin. Although these anomalies
at 12 km depth, and 8000 micro-cgs units susceptibility magnetization. exhibit amplitudes of about 40 nT, confidence in these trends is high. This
confidence is supported by two aspects of the region and the magnetic
field. First, the trends were correlated using data from a single cruise
At low geomagnetic inclinations, the ends of offset-spreading ridge
(U. S. Navy WI932010) with eighteen lines spaced approximately 8 km
segments or east-west-trending features such as transform faults that have
apart. Second, the acoustic basement surface in this part of the basin is
been injected with magnetized material would produce anomalies of sev-
relatively smooth (Speed et al., 1984), suggesting that intrabasement
eral hundred nT. In contrast, north-south-trending ridge segments would
sources are responsible for the anomaly trends observed in profiles.
produce anomalies of only a few tens of nT. At the magnetic equator, a
Trends over the northern part of the basin may have resulted from the tec-
north-south ridge segment would produce no anomaly if magnetization of
tonic event responsible for the bifurcation of the northern Lesser Antilles.
the source body is caused solely by the inducing field.
That is, the original magnetic signature is thought to have been disrupted
The north-northeast to north-south orientation of anomaly trends,
by faulting and possible strike-slip motion (Bird, 1991; Bird et al., 1993).
seen in profile data, over the basin south of 14°N are interpreted to be
03 Basin Studies 4/16/07 4:16 PM Page 57
Conclusion
After understanding the magnetic field over the Grenada Basin, it is a rel-
atively easy task to piece together the geologic events related to its for-
mation. In his discussion regarding the magnetic anomalies over the
Grenada Basin, Bouysse (1988) points out that the great depth to the
oceanic basement, combined with a possible location near the geomag-
netic equator of the eastern Caribbean, may blur the original anomaly pat-
tern. These factors are the primary reason for the confusion regarding the
magnetic field over the Grenada Basin.
Interpretation of magnetic anomalies at low geomagnetic inclinations
depends on the strike of the geologic features and the anomaly patterns
they produce. The magnetic anomaly patterns over the Grenada Basin
and our interpretation of them demonstrate this dependence. The
Grenada Basin is interpreted to have formed by near east-west extension
in the Early Tertiary (Figure 9). This conclusion is supported by the basin
morphology, gravity data, and subtle magnetic anomaly trends over the
southern part of the basin (Bird, 1991). These low-amplitude anomalies
are produced by roughly north-south-oriented spreading centers near the
geomagnetic equator. The chaotic, patchy anomalies over the northern
part of the basin are thought to have formed by sea-floor spreading also,
but they were disrupted later by the Late Tertiary event responsible for
the bifurcation of the Lesser Antilles (Bird et al., 1993).
Figure 9. Two-step model for the formation of the Grenada Basin via east-west exten-
sion. Large arrows indicate directions of relative plate motion. Small arrows indicate
directions of extension and basin formation. (a) The basin formed fairly uniformly by
sea-floor spreading. (b) Late Tertiary compressional tectonism disrupted the northern
portion of the basin (indicated by long, northeast-oriented arrows).
Because many of the world’s prospective basins are in regions charac- Poehls, K. A., 1978, Intra-arc basins: A kinematic model: Geophys. Res. Lett., 5,
terized by low magnetic latitudes, the work performed here can be used 325–328.
as an analog for understanding the magnetic field, and the geology which Ross, M. I., and Scotese, C. R., 1988, A hierarchical tectonic model of the Gulf of
Mexico and Caribbean region: Tectonophysics, 155, 139–168.
produces it, in such regions.
Sharma, P. V., 1976, Geophysical methods in geology: Elsevier Science Publ. Co., Inc.
Acknowledgment Shurbet, D., 1976, Definition of the Caribbean plate by Sn waves from earthquakes:
Trans. 7th Conf. Geol. Caraibes, 87–91.
The authors would like to thank Marathon Oil Company for contributing
the data used in this study as well as the use of Marathon’s computer sys- Speed, R. C., et al., 1984, Lesser Antilles arc and adjacent terranes, Atlas 10, Ocean
Margin Drilling Program, Regional Atlas Series: Mar. Sci. Int.
tem and software.
Tamaki, K., 1985, Two modes of back-arc spreading: Geology, 13, 475–478.
References Tomblin, J. F., 1975, The Lesser Antilles and Aves ridge, in The ocean basins and mar-
gins, Nairn, A. E. M., and Stehli, F. G., Eds.: Plenum Press, 3, 467–500.
Bird, D. E., 1991, An integrated geophysical interpretation of the Grenada Basin: M.S.
thesis, Univ. Houston. Uyeda, S., 1982, Subduction zones: An introduction to comparative subductology:
Tectonophysics, 81, 133–159.
Bird, D. E., Hall, S. A., Casey, J. F., and Millegan, P. S., 1993, Interpretation of magnetic
anomalies over the Grenada Basin: Tectonics, 12, 1267-–279. Uyeda, S., and Kanamori, H., 1979, Back-arc opening and the mode of subduction: J.
Geophys. Res., 84, 1049–1061.
Bouysse, P., 1988, Opening of the Grenada back-arc basin and evolution of the
Caribbean plate during the Mesozoic and Early Paleocene: Tectonophysics, 149, Weissel, J. K., 1981, Magnetic lineations in marginal basins of the west Pacific: Philos.
121–143. Trans. Roy. Soc. London, A300, 223–247.
Boynton, C. H., Westbrook, G. K., Bott, M. H. P., and Long, R. E., 1979, A seismic Westbrook, G. K., 1975, The structure of the crust and upper mantle in the region of
refraction investigation of crustal structure beneath the Lesser Antilles island arc: Barbados and the Lesser Antilles: Geophys. J. Roy. Astr. Soc., 43, 201–242.
Geophys. J. Roy. Astr. Soc., 58, 371–393.
Donnelly, T. W., 1975, The geological evolution of the Caribbean and Gulf of Mexico:
Some critical problems and areas, in Nairn, A. E. M., and Stehli, F. G., Eds., The
ocean basins and margins, 3: Plenum Press, 663–689.
Duncan, R. A., and Hargraves, R. B., 1984, Plate tectonic evolution of the Caribbean
region in the mantle reference frame, in The Caribbean-South American plate
boundary and region tectonics, Bonini, W. E., et al., Eds.: Geol. Soc. Am. Mem.,
162, 81–93.
Ghosh, N., Hall, S. A., and Casey, J. F., 1984, Seafloor spreading magnetic anomalies in
the Venezuelan basin, in The Caribbean-South American plate boundary and
regional tectonics, Bonini, W. E., et al., Eds.: Geol. Soc. Am. Mem., 162, 65–80.
National Geophysical Data Center (NGDC), Global Relief Data CD-ROM: National
Oceanic and Atmospheric Administration (NOAA), U.S. Department of Commerce.
National Geophysical Data Center (NGDC), Marine Geological and Geophysical Data
CD-ROM: National Oceanic and Atmospheric Administration (NOAA), U.S.
Department of Commerce.
Pindell, J. L., and Barrett, S. F., 1990, Geological evolution of the Caribbean region: A
plate tectonic perspective, in Dengo, G., and Case, J. E., Eds., The Caribbean
region, decade of North American geology: Geol. Soc. Am., H, 405–432.
Pindell, J., and Dewey, J. F., 1982, Permo-Triassic reconstruction of western Pangea and
the evolution of the Gulf of Mexico-Caribbean region: Tectonics, 1.
Pindell, J. L., Cande, S. C., Pitman III, W. C., Rowley, D. B., Dewey, J. F., LaBrecque, J.,
and Haxby, W., 1988, A plate-kinematic framework for models of Caribbean evolu-
tion: Tectonophysics, 155, 121–138.
57
Geologic Applications of Gravity and Magnetics: Case Histories
03 Basin Studies 4/16/07 4:16 PM Page 58
General Philosophy 2
Gravity and Magnetics
Community
Edited by David A. Chapin
LaCoste & Romberg
Austin, Texas, USA
The foundation of good interpretive
work is a sound understanding of the
geology.
It sounds trite, but you just can’t get any-
where blundering around manipulating
data without some idea of what to look
for in the results.
I agree with the concept of systematically
suppressing the things you aren’t inter-
ested in. The counterpart is systemati-
cally enhancing the features you are
interested in, and that’s where under-
standing the geology is crucial.
I think potential-fields interpretation is
enriched greatly by the wide variety of
technology in use. In many branches of
geophysics, I see pretty much everyone
doing the same thing. The fact that this
isn’t the case in gravity and magnetics
may not comfort management, but it
makes the business much more
interesting.
Richard Hansen
Pearson, deRidder, & Johnson
Lakewood, Colorado, USA
I like the idea about our varied
approaches to the interpretation and use
of potential-fields data. Indeed, the fun-
damental fact is:
No single action can be taken to under-
stand the data.
Most potential-fields geophysicists enjoy
the challenge of deciphering the anomaly
patterns.
Many geophysicists and managers feel
more comfortable with seismic interpre-
tation because they are viewing a repre-
sentation that “looks like” a cross-section
of the Earth (velocity versus depth and
acoustic interfaces versus lithologic
boundaries aside). Potential-fields data
are a summation of effects measured at a
reference surface, and they consist of an
accumulation of overlapping anomalies
and/or superimposed anomalies. Very
seldom are we blessed with single, well-
formed anomalies. Even if we were, the
representation of the data—a map or
profile—is not a cross-section, it is a
wavy line or wavy surface. Hence, the
visualization process is much different.
Potential-fields interpreters become
involved in anomaly separation processes
and modeling schemes that attempt to
approximate a physical representation of
the subsurface geology (often in both
two and three dimensions). This is an
attempt to visualize the subsurface in
relation to the waveforms in the data.
Thus, many view us as wizards who
“make up” geology to fit the data. Again,
we are confronted with the importance
of the real geologic input, the real geo-
logic control in the interpretation
problem.
Pat Millegan
Worldwide Exploration Services
Marathon Oil Company
Houston, Texas, USA
It has been said that we should keep our
geologic objectives in mind during pro-
Basin Studies
cessing and interpretation. I would add
acquisition to this.
We should keep our geologic objectives
in mind during the acquisition, process-
ing, and interpretation.
Too often, the geologic objective requires
500-m resolution when the station con-
trol is 2 km apart. Gravity and magnetics
have a reputation as a low-resolution
tool. Although we still cannot cheat on
basic laws of nature, we often give up
too easily in striving for higher resolution.
My approach to interpretation is to start
with defining the geologic problem even
before acquisition. I model various possi-
ble scenarios and from there decide the
acquisition parameters that will be
needed to answer the problem. Some-
times the answer is that gravity and
magnetics cannot help with current
accuracies (forming a justification to
enhance the technology to new accuracy
limits). Sometimes the station or line
spacing required is uneconomic, and the
survey should not proceed. I would,
however, prefer to know this beforehand.
This “feasibility study” is also important
when working with existing data, in
deciding whether the data are “good
enough” to solve the geologic problem or
whether new acquisition will be required
(or whether the problem is insoluble).
In defining the various possible model
scenarios, it is also vital to include other
possible disturbing factors. The fre-
quently abused rules of thumb of need-
ing a station or line spacing as a fixed
proportion of the depth to objective
ignore the possibility (probability?) that
shorter-wavelength features from shal-
lower bodies in the section will be
aliased by the sampling. Interpretation on
aliased data (which is often very difficult
to recognize) will give wrong results,
even for the deeper objectives.
Paul Featherstone
Shell International Explorational Production B.V.
The Hague, Netherlands
I would like to support the comment
about aliasing caused by short-wave-
length variations, caused by some shal-
low source unrelated to our target. In
Canada, we run into this all the time
with the glacial drift. We insist on rela-
tively tight gravity station spacing to ade-
quately sample the density contrast at
the base of the drift. With close station
spacing, we usually can define this sur-
face quite reliably, and it often causes far
more amplitude than the deep signal
which we are really interested in.
John Peirce
GEDCO
Calgary, Alberta, Canada
The clearly expressed statements about
keeping the geologic problem in mind
are really the bottom line, and the con-
cept of acquisition parameters leads right
into this: What different kinds of infor-
mation can we get from data sets at dif-
ferent scales?
I remember working on a high-quality
survey over a tiny area, with no regional
data to provide a context. A gigantic gra-
dient at one corner of the survey was a
real question (unresolved by us at the
time). But later, with the publishing of a
regional data set, it became clear that the
58
feature was part of a wide dikelike body,
presumably in the basement. The point is
simply that the broader context may be
more informative, even if the regional
data are of lower quality.
I think the tectonic and structural infor-
mation (both blatant and subtle) present
in the excellent continental-scale compi-
lations is almost incredible sometimes
and often overlooked or underused. But
unless geologists looking at such maps
are cognizant of the data and grid spac-
ings, they will be confused, distressed,
and may not return to use such valuable
data when appropriate. The simple point
is this: Different valuable information is
in data sets at different scales.
Richard I. Gibson
Gibson Consulting
Golden, Colorado, USA
Try to move from 2-D to 3-D in your
interpretation.
Usually I model in 2.5-D initially. Once
the model is “complete,” I then use the
potential-field data to expand the inter-
pretation to 3-D. In most cases, the
potential-field data are the only geophysi-
cal data available in map view, so they
are very useful for extending the inter-
pretation of 2-D seismic or electrical
data.
Simple geologic maps of the subsurface,
containing faults, folds, basins, and geo-
logic units, can be derived from the avail-
able data sets. Use of the maximum
horizontal gradient, pseudogravity, and
other techniques facilitates the map-
making. Cordell and McCafferty (1989)
developed the “terrace” method to try to
03 Basin Studies 4/16/07 4:16 PM Page 59
direct potential-field interpretation
towards subsurface geologic maps. In this
way, potential-field data could gain the
kind of credibility accorded to seismic
reflection data, because the seismic data
are presented to look like geologic cross-
sections.
References
Cordell, L., and McCafferty, A. E., 1989, A ter-
racing operator for physical property mapping
with potential-field data: Geophysics, 54,
621–634.
Carol Finn
U. S. Geological Survey
Denver, Colorado, USA
Keep in mind that seismic anomalies are
indicative of vertical changes in rock
properties, whereas potential-field anom-
alies are indicative of lateral changes in
rock properties. Indeed, assumptions
related to processing seismic data actu-
ally suppress identification of these lat-
eral changes, such as: The Earth is made
up of horizontal layers of constant veloc-
ity, the source waveform is known, noise
is predictable, the wavelet is minimum
phase, etc.
Seismic data, unlike potential-field data,
can’t always tell what is happening geo-
logically in a lateral sense.
Dale Bird
Bird Geophysical
Houston, Texas, USA
Aeromagnetic Constraints on the Basement
Structure of the Tunghai Shelf and the
Okinawa Trough in the East China Sea
S. Okuma, T. Nakatsuka, M. Makino, and R. Morijiri
Geological Survey of Japan
Tsukuba, Ibaraki, Japan
EDITOR’S NOTE: Depth-to-basement inversions are fairly routine in
regional scale exploration. It is important to note that rela-
tively simple modeling assumptions can yield valuable con-
straints on basement structure.
Remanant magnetization “relaxes” over geologic time. Pre-
cambrian granites worldwide seem to display only induced
magnetization and can be modeled effectively without a
remanant effect.
Introduction
Aeromagnetic surveys offshore the western Nansei Islands, Japan, were
conducted by the Geological Survey of Japan (GSJ) from 1982 to 1989
(Okuma et al., 1991). The surveys cover the area from the Tunghai Shelf
in the East China Sea to the Okinawa Trough west of the Nansei Islands
(Ryukyu Arc) (Figures 1 and 2). Although the area, especially the Tunghai
Shelf, was regarded as a potential field for hydrocarbon resources, few
explorations were conducted in the area. The aeromagnetic surveys,
therefore, were conducted. Then we performed a 3-D, two-layer mag-
netic modeling to obtain further detailed information about the basement
structure of the area. This paper outlines the results of the modeling and
describes characteristics of the magnetic anomalies in the area.
Geology
In the study area, two major geologic structures are known for having
hydrocarbon potential—the Tunghai Shelf and the Okinawa Trough
(Figure 2).
Wageman et al. (1970) suggested the existence of a large sedimentary
basin, the Taiwan Basin, in the Tunghai Shelf, which is bounded on the
northwest and southeast by the Fukien-Ryeongnam Belt (Fukien-Reinan
Massif) and Goto-Senkaku Belt (Taiwan-Sinzi Folded Zone), respectively
(Figure 3). Since then, the basin has been regarded as a potential field for
Geologic Applications of Gravity and Magnetics: Case Histories
oil and gas resources. The Taiwan Basin is divided into three basins: the
Goto Basin, the Tokai Basin, and the Senkaku Basin from northeast to
southwest (Kizaki, 1986; Figure 3). Thick sediments have been deposited
almost continuously since the Paleogene, with a hiatus in the Middle
Miocene (Aiba and Sekiya, 1979). Recently, hydrocarbon explorations
such as seismic reflection surveys and drilling have been conducted,
mainly in the Goto Basin and the Tokai Basin, resulting in shows of oil
and gas. Oil and gas shows also have been confirmed in several drill holes
close to the continent in the East China Sea (Guangding, 1989).
However, the detailed basement structure of the area, especially the
Senkaku Basin, remains unclear because of insufficient explorations.
The Goto-Senkaku Belt lies along the outer edge of the Tunghai Shelf
between Kyushu Island and Taiwan and consists of two parts, the Goto
Belt and the Senkaku Belt (Aiba and Sekiya, 1979; Figure 3). The Goto
Belt comprises Paleogene to Early Miocene sediments associated with
welded tuff and granitic rocks of Miocene age, and it is overlain mostly
by Pliocene sediments (Kizaki, 1986). Miocene granitic rocks also are
exposed in the Senkaku Islands (Matsumoto and Tsuji, 1973). The Goto-
Senkaku Belt was uplifted in the Middle Miocene and, in some parts, to
Pliocene time and dammed flows of sediments from the continent.
The Okinawa Trough is a back-arc basin associated with northwest-
ward subduction of the Philippine Sea Plate at the Ryukyu Trench (Figure
2) and is bounded on the northwest and southeast edges by the Goto-
Senkaku Belt and the Ryukyu Arc, respectively (Figure 3). The bathyme-
try of the trough becomes deeper toward the southwest and exceeds
2000 m offshore northern Ishigaki Island (Figure 2). Hydrocarbon explo-
rations have been conducted mainly in shallow-water areas—the north-
eastern part of the trough and the Ryukyu Arc area (Aiba and Sekiya,
1979). Geologic and geophysical investigations (e.g., Oshima et al., 1988)
also have been conducted to reveal the origin of the trough. Many step
faults and their resultant roll-over anticlines commonly are observed on
seismic profiles in the west flank of the trough north of Okinawa Island
(Aiba and Sekiya, 1979). Covering the trough are thick Pliocene and
Pleistocene sediments, and the maximum depth of acoustic basement
exceeds 5000 m.
Characteristics of Magnetic Anomalies
Previous magnetic data in the East China Sea are mainly from widely
spaced ship-track data (e.g., Wageman et al., 1970). The aeromagnetic
surveys by GSJ were flown at an elevation of 460 m (1500 ft) above sea
level with line spacing of 5 km, especially 2.5 km over the Senkaku
Islands (Okuma et al., 1991). Therefore, our new magnetic data (Figure
4) are more detailed than the previous ones.
59
03 Basin Studies 4/16/07 4:16 PM Page 60
Figure 1. Location of the study area. The area bounded by solid red lines shows the study area. Vertical exaggeration is
12.5:1. Topography data were compiled from the terrain data “KS-110” by Geographical Institute of Japan, bathymetry data
by the Hydrographic Department, Maritime Safety Agency, Japan, and ETOPO5 data by NOAA.
Generally, magnetic anomalies trend northeast to southwest along the
topography (Figure 4). Magnetic anomalies, which have long wavelengths
of more than several tens km and small amplitudes of less than 50 nT,
are distributed predominantly in the Tunghai Shelf and west flank
of the Okinawa Trough. From the total-intensity magnetic anomalies
(Figure 4), a reduction-to-the-pole anomaly map was made through a
method in frequency domain (Figure 5). This map clearly shows a corre-
lation between the bathymetry and magnetic highs at the margin of the
Tunghai Shelf. A large number of magnetic highs (H2) are dominant over
the Goto Belt, whereas few intense anomalies are obvious over the
Senkaku Belt, except in and around Sekibi-sho Island. This implies a dif-
ference in past volcanic activities between the Goto Belt and the Senkaku
Belt, with more intense activities on the Goto Belt. Magnetic lows (L1)
lie west of the Goto-Senkaku Belt in the shelf, where the Tokai and
Senkaku basins are located. A magnetic high area (H1) is distributed
north of Uotsuri Island in the shelf, which is inferred to correspond to
local uplifts of the basement of the Senkaku Basin.
Basin Studies
Figure 2. Bird’s-eye view map of the topography of the study area. See also Figure 1.
A magnetic low area (L2) ranges along the west flank of the Okinawa
Trough but is limited to the areas from the northeastern edge of the sur-
vey area to the northwest of Okinawa Island.
Magnetic highs (H3, H4) can be observed in the Okinawa Trough
near the southwestern and northeastern edges of the survey area. Some
of these anomalies at the axis of the trough correspond to knolls which
are assumed to be magnetized in the direction of the present Earth’s mag-
netic field (Oshima et al., 1988).
Magnetic Modeling
As we described before, long-wavelength and small-amplitude magnetic
anomalies are distributed predominantly in the Tunghai Shelf and west
flank of the Okinawa Trough. These anomalies may correspond to uplift
and subsidence of the geologic basement, composed probably of granitic
rocks. Therefore, we applied a 3-D, two-layer model inversion (Gerard
and Debeglia, 1975; Okuma et al., 1989) to interpret magnetic anomalies
of the study area. First, an average depth of the interface, 3.9 km below
60
sea level, between the nonmagnetic upper layer and magnetic basement
was estimated from a slope of radially averaged power spectrum of the
reduction-to-the-pole anomalies (Figure 5). Then the deviation from the
average depth (i.e., magnetic basement depth) was calculated over the
survey area by an iterative convergence method, using pseudogravity and
reduction-to-the-pole anomalies. We assumed the magnetic basement is
magnetized in the direction of the present Earth’s magnetic field (inclina-
tion = 38.5°N, declination = 4°W) in the area, with a uniform magneti-
zation intensity. We employed 2.0 A/m for the magnetization intensity,
taking account of magnetic anomalies caused by volcanic rocks in the
Goto-Senkaku Belts and the Okinawa Trough in addition to those caused
by granitic rocks. As magnetization intensities of rocks actually vary from
point to point, resultant magnetic basement depths in this study do not
always coincide with the acoustic basement. In our case (Figures 6 and
7), the magnetic basement in the southern Okinawa Trough is deeper
than the top of the 4.5–6.0 km/s layer, the acoustic basement of a multi-
channel seismic reflection profile, and shallower than the top of the
6.0–6.4 km/s layer (Hirata et al., 1991), probably a granitic layer.
03 Basin Studies 4/16/07 4:16 PM Page 61
Figure 3. Outline of the geologic structure in and around the East China Sea (after
Japan Natural Gas Association and Japan Offshore Petroleum Development
Association (1982): (1) Fukien-Reyongnam Belt; (2) Goto Belt; (3) Senkaku Belt; (4)
Ryukyu Geanticline; (5) Ryukyu Trench; (A) Goto Sedimentary Basin; (B) Tokai
Sedimentary Basin; (C) Senkaku Sedimentary Basin; (D) Okinawa Trough; (E) Ryukyu
Fore-arc Sedimentary Basin.
The magnetic basement attains a depth of less than 2 km below sea
level on the Goto Belt (Figure 6 and cross-sections C-D and E-F in Figure
8). The result implies the existence of magnetized bodies near the sea
bottom in this area. They may be related to Middle Miocene volcanic and
intrusive rocks of the Goto Belt. Around Sekibi-sho Island, the magnetic
basement also rises to a depth of shallower than 2 km below sea level
(Figure 6 and cross-section G-H in Figure 8). However, the uplifted area is
limited to near the island, compared with the one in the Goto Belt. This
result suggests a difference in geology between the Goto and Senkaku Belts.
Figure 4. Shaded-relief map of IGRF-residual total-intensity aeromagnetic anomalies
of the study area. The aeromagnetic surveys by GSJ were flown along northwest-
southeast flight lines at an altitude of 460 m above sea level and spaced 5 km apart,
especially 2.5 km apart over the Senkaku Islands (Okuma et al., 1991). Contour inter-
val is 25 nT. Bathymetric contours are in meters. Inclination of the light source, 45°;
declination, 0°.
The magnetic basement is situated at a depth of 1 km below sea level
to the north of Uotsuri Island in the Tunghai Shelf (Figure 6 and cross-sec-
tion G-H in Figure 8). This magnetic basement high corresponds to one of
the large anticlinal structures with gentle dips in the Senkaku Basin, sug-
gested by seismic reflection data (Tokai Univ., 1969). Hydrocarbon accu-
mulations are expected in those structures, but intensive seismic
investigations are necessary to confirm more detailed structures in this area.
The magnetic basement also rises up to near the sea bottom along the
east flank of the Okinawa Trough, especially offshore northern Okinawa
Island (cross-section C-D in Figure 8) and near the Miyako Seamount
(cross-section G-H in Figure 8). They correspond to volcanic knolls and
intrusive rocks beneath the sea bottom.
The magnetic basement subsides mainly in the Tunghai Shelf and the
west flank of the Okinawa Trough (Figure 6). A magnetic basement low
extends in the north-northeast–south-southwest or northeast-southwest
direction at a maximum depth of 10 km below sea level in the Tunghai
Geologic Applications of Gravity and Magnetics: Case Histories
Figure 5. Shaded-relief map of reduction-to-the-pole anomalies, calculated from the
IGRF-residual total-intensity aeromagnetic anomalies (Figure 4). H1-H4, magnetic
highs; L1-L2, magnetic lows. See also Figure 4.
Shelf (cross-section E-F in Figure 8), although a small magnetic basement
high, 100 km north of Sekibi-sho island, interrupts the extension locally.
This corresponds to the southern edge of the Tokai Basin and the eastern
part of the Senkaku Basin.
The maximum depth of the magnetic basement reaches 10 km below
sea level in the west flank of the Okinawa Trough, 150 km offshore north-
west Okinawa Island (cross-section E-F in Figure 8). A magnetic basement
low traverses the Goto Belt in the northwest-southeast direction and cor-
responds to a subsidence of the acoustic basement in the northern part of
this area, where thick deltaic or submarine fan sediments are deposited
(Aiba and Sekiya, 1979). There is a possibility of hydrocarbon accumula-
tions in those thick sediments because of high heat flows, although the
sediments seem to be too young to generate them.
Conclusions
A 3-D, two-layer model inversion was applied to analyze magnetic anom-
alies in the Tunghai Shelf and the Okinawa Trough. The resultant mag-
netic basement map gave constraints of the basement structure of the
area. This map shows the existence of a local uplifted zone in the
61
03 Basin Studies 4/16/07 4:16 PM Page 62
Figure 6. Magnetic basement depth map of the study area, estimated by a 3-D, two-
layer model inversion (Gerard and Debeglia, 1975; Okuma et al., 1991). The magnetic
basement is assumed to be magnetized in a direction of the present Earth’s magnetic
field (inclination = 38.5°N, declination = 4°W) in the area, with a uniform magnetiza-
tion intensity of 2.0 A/rn. The absolute mean error is 14.9 nT. Contour interval is
1000 m. See also Figure 4.
Figure 8. Cross-sections (C-D), (E-F), and (G-H) on the magnetic basement depth map (Figure 6).
Basin Studies
Figure 7. Cross-section (A-B) on the magnetic basement depth map (Figure 6). The
4.5–6.0 km/s layer, corresponding to the acoustic basement of a multichannel seismic
reflection profile, and the 6.2–6.4 km/s layer, probably a granitic layer, are from a seis-
mic refraction profile along a 2-D P-wave velocity model (Hirata et al., 1991) of the
1988 DELP cruise.
62
Senkaku Basin, and magnetic basement depths of 10 km below sea level
in the Tokai Basin and in the west flank of the Okinawa Trough north of
Okinawa Island.
Acknowledgment
The authors would like to express their appreciation to Carol Finn, U.S.
Geological Survey, for her helpful reading of the manuscript.
References
Aiba, J., and E. Sekiya, 1979, Distribution and characteristics of the Neogene sedimen-
tary basins around the Nansei-Shoto (Ryukyu Islands): J. of the Japanese Association
of Petroleum Technologists, 44, 329–340 (in Japanese with English abstract).
Gerard, A., and Debeglia, N., 1975, Automatic three-dimensional modeling for the inter-
pretation of gravity or magnetic anomalies: Geophysics, 40, 1014–1034.
Guangding, L., 1989, Geophysical and geological exploration and hydrocarbon
prospects of the East China Sea: China Earth Sciences, 1, 43–58.
Hirata, N., Kinoshita, H., Katao, H., Baba, H., Kaiho, Y., Koresawa, S., Ono, Y., and
Hayashi, K., 1991, Report on DELP 1988 cruises in the Okinawa Trough, part III,
crustal structure of the southern Okinawa Trough: Bull. of the Earthquake Research
Institute, Univ. of Tokyo, 66, 37–70.
Japan Natural Gas Association and Japan Offshore Petroleum Development Association,
1982, Oil and natural gas resources in Japan (in Japanese).
Kizaki, K., 1986, Geology and tectonics of the Ryukyu Islands: Tectonophysics, 125,
193–207.
03 Basin Studies 4/16/07 4:16 PM Page 63

Matsumoto, Y. and Tsuji, K., 1973, Geology of the Uoturi-jima, Kita-kojima and Minami- Magnetic Frequency-depth
ko-jima in the Senkaku Retto: Bull. of Faculty of Liberal Arts, Natural Science,
Nagasaki Univ., 14, 43–57. Relationship
Okuma, S., Makino, M., and Nakatsuka, T., 1989, Two-layer model inversion of mag-
Shallow density and susceptibility con-
netic anomalies using pseudogravity and reduction to the pole: An application to the
analysis of aeromagnetic anomalies over Izu-Oshima: Butsuri-Tansa (Geophysical trasts produce narrow, or high-frequency,
Exploration), 42, 82–96 (in Japanese with English abstract). anomalies, while deeper contrasts result
in broader, low-frequency anomalies. We
Okuma, S., Nakatsuka, T., Makino, M., and Morijiri, R., 1991, Determination of base-
can use this characteristic of the gravity
ment structure offshore of the western Nansei Islands from aeromagnetic data:
Butsuri-Tansa (Geophysical Exploration), 44, 202–214 (in Japanese with English and magnetic fields to determine depths
abstract). to sources, both qualitatively and quanti-
Oshima, S., Takanashi, M., Kato, S., Uchida, M., Okazaki, I., Kasuga, S., Kawashiri, C., tatively. Because gravity signatures come
Kaneko, Y., Ogawa, M., Kawai, K., Seta, H., and Kato, Y., 1988, Geological and geo- from everywhere, from the air to the
physical survey in the Okinawa Trough and the adjoining seas of Nansei Syoto: Earth’s core, such inferences are often
Report of Hydrographic Researches, Maritime Safety Agency, Japan, 24, 19–43 (in not straightforward or reasonably unique.
Japanese with English abstract). But most magnetic anomalies come from
Tokai Univ., 1969, Cruise report on a submarine geological survey in and around the
Senkaku Islands (in Japanese).
only a few rock types, such as volcanics,
intrusives, and basement rocks. Magnetic
Wageman, J. M., Hilde, W. C., and Emery, K. O., 1970, Structural framework of East data therefore can be used to estimate
depth to basement—a classic use for
China Sea and Yellow Sea: AAPG Bull., 54, 1611–1643.
such data. The illustration shows how
the anomalies change from narrow over
shallow basement to broad over deeper
basement.
R. I. Gibson
Aeromagnetic Interpretation
of Southwestern Continental
Shelf of Korea
Czango Baag
Department of Geophysics, Kangweon National University
Kangween-Do, Republic of Korea
Chang-Eob Baag
Department of Geological Sciences, Seoul National University
Seoul, Republic of Korea
EDITOR’S NOTE: Baag and Baag show how analysis of even quite
old low-quality data sets can provide a powerful tool in tec-
tonic analysis. The work resulted in the prediction of a new
sedimentary basin whose origin can be understood in terms
of geologic structural elements that are well expressed in the
magnetics data.
Abstract
Analysis of the Project Magnet aeromagnetic data acquired by the U. S.
Navy in 1969 permits us to predict a new sedimentary basin, Heuksan
Basin, south of the known Gunsan Basin in Block II, offshore South
Korea. The basin appears to consist of three subbasins trending north-
northwest–south-southeast. The results of our analysis provide not only
an independent assessment of the Gunsan Basin, but also new important
information on the tectonic origin and mechanism for the two basins as
well as for the entire region. The basin-forming tectonic style is inter-
preted as rhombochasm associated with double-overstepped left-lateral
wrench faults. From magnetic evidence, a few northeast-southwest-trend-
ing major onshore faults are extended to the study area.
We also interpreted the faults to be left-lateral wrenches. This new
gross structural style is consistent with the results of recent Yeongdong
Basin analysis (Lee, 1990). The senses of fault movement also are sup-
ported by paleomagnetic evidence that the Philippine Sea has experi-
enced an 80˚ clockwise rotation since the Eocene. Based on a 2.5-D
model study, the probable maximum thickness of the sediments in the
Gunsan Basin is approximately 7500 m. We believe that the new
Heuksan Basin was left unidentified because a high-velocity layer may

63
Geologic Applications of Gravity and Magnetics: Case Histories
03 Basin Studies 4/16/07 4:16 PM Page 64
overlie the basin. Because the overall structural configuration of the
Heuksan Basin appears to be favorable for hydrocarbon accumulation, a
detailed airborne magnetic survey is recommended in the area, to verify
the magnetic expressions of both this thick basin and the tectonic style.
A detailed marine gravity survey acquired in conjunction with a high-
energy source seismic survey also is recommended, to delineate the sed-
imentary section and to acquire data supplemental to the magnetics.
Introduction
Since geologic conditions onshore Korea proved to be unfavorable for
petroleum, hydrocarbon exploration efforts have been targeted at Korea’s
vast continental-shelf areas. The western area offshore Korea (South
Yellow Sea), known to contain thick nonmarine sedimentary sequences,
is of especially strong economic interest because of the discovery of very
prolific hydrocarbons in similar nonmarine sediments in China.
As a result of a significant amount of marine seismic data acquired
from the area and with three subsequent exploratory wells, two sedi-
mentary basins were delineated (Saltmarsh, 1989; Korea Institute of
Energy and Resources, 1984, 1985, 1990). One is Gunsan Basin, the
other North-South Heuksan Basin. (Note: The North-South Heuksan
Basin is not shown in Figure 3 because it is located outside the aeromag-
netic survey area in the far western part of Block III.) However, the ori-
gin and nature of the basins are not fully understood. The basins were
interpreted to originate from a complex fault system (Korea Institute of
Energy and Resources, 1984, 1985), and only recently, wrench-fault con-
cepts have been mentioned (Chun, 1989).
Because marine seismic lines are confined to the basin areas, no
regional tectonic studies were possible. In light of the paucity of regional
geophysical data, the aeromagnetic data of 1969, although of reconnais-
sance nature, can be very useful in delineating basement surface configu-
ration and structural pattern. A magnetic basement interpretation can
lead to a better understanding of structures of the overlying sedimentary
rocks.
Objectives of this study were: (1) to calculate thickness of sediments;
(2) to delineate magnetic basement surface configuration; (3) to outline
possible magnetic basement rock provinces; (4) to distinguish possible
intrasedimentary volcanic areas; (5) to construct 2-D and/or 3-D models;
(6) to select and/or recommend potential hydrocarbon areas, if any; and
(7) to recommend future geophysical investigation in the area.
In this investigation, we not only formulate a gross structural style of
the area, but we also predict a possible new basin which appears to be a
northeastward extension of the North Heuksan Basin.
Basin Studies
Magnetic Data
Offshore Korea was flown over by the U.S. Naval Oceanographic Office
in February and March of 1969 (Figure 1). The project was carried out
within the framework of its Project Magnet. More than 40 000 line-kilo-
meters were flown, covering approximately 200 000 km2 (Bosum et al.,
1971). Three flight elevations were employed: 183m (600 ft) over the
sea, 1067 m (3500 ft) along coastal areas, and 2195 m (7200 ft) over
Cheju Island. Flight-line spacing was 8 km in an east-west direction.
It is reported that much wider flight-line spacing sometimes was used.
No tie lines were flown. Although a Doppler-LORAN system was used, the
possible error estimate in plotting of the aeromagnetic data appears to be
as large as ± 2 km (Bosum et al., 1971).
The data originally were recorded on analog charts. The original total-
intensity maps were contoured manually at 50-gamma intervals.
Figure 1. Location map of the study area.
64
Data Quality
Quality of the aeromagnetic data was judged to be poor because the orig-
inal object was for a reconnaissance survey for Project Magnet, a world-
wide aeromagnetic survey project. Widely spaced lines without any tie
lines inevitably can cause large position errors. Moreover, contour maps
do not show flight lines, and errors from splining can be suspected.
Therefore, interpretation often can be misleading and basement depths
quite wrong, and a false trend can be made.
Qualitative Magnetic Interpretation
Because of the reconnaissance nature of the data, no attempts were made
to estimate magnetic susceptibilities.
The distinctive magnetic anomaly trends, patterns, and amplitude val-
ues strongly suggest that the study area may consist of four magnetic base-
ment-rock provinces.
Basement-rock Provinces
1) A magnetically complex, short-wavelength, high-amplitude province
is distributed most widely throughout the study area, particularly off-
shore Korea north of Cheju Island and in the southwestern area off-
shore Gunsan city and the coastal areas connecting the two locations.
In this province, magnetic anomalies trend predominantly northeast-
southwest. In the area farther southwest of this province, they trend
northwest-southeast. It appears that some of the strong lineaments
may have been generated falsely because of the wide flight-line spac-
ing. In this province, intrasedimentary flows, dikes, and extrusive
flows are distributed ubiquitously (Figure 2).
2) In the northern part of the study area, low-amplitude, high-frequency
anomalies exhibiting wavy contour lines are interpreted to be the
magnetic expression of Yeoncheon Group of Precambrian age. This
group is well known for its banded iron formation. Anomalies trend
northeast-southwest, but the trend gradually changes to a north-south
direction toward the Gunsan Basin.
3) The shorter-wavelength anomalies around Cheju Island are inter-
preted to be lava flows and injected lavas.
4) The southern part of the study area south of approximately 33°N lat-
itude shows characteristics of medium-wavelength and medium-
amplitude anomalies. The lower magnetic relief with only vaguely
03 Basin Studies 4/16/07 4:16 PM Page 65
defined strike directions may indicate the magnetic expression of
granitic or gneissic origin.
Magnetic Basement Configuration
The interpreted magnetic basement structure map shows a relief of about
8000 m within the study area, ranging from sea level to –7500 m in the
Gunsan Basin.
The interpreted magnetic basement surface appears to be rather
smooth in the study area, except in the two deeply buried basement
areas.
Gunsan and Heuksan Basins
The most pronounced and economically appealing features interpreted
from the magnetic data are two areas of deeply buried magnetic base-
ment. Located in the northwestern part of the study area, they occupy the
western part of Block II and a significant portion of the adjoining north-
western part of Block III (Figures 2 and 3). The Gunsan Basin already has
been defined and delineated by the seismic method and by three wells
drilled in the basin. The magnetic characteristics of the newly identified,
deeply buried magnetic basement are very similar to those of the Gunsan
Basin. This basinlike deep magnetic basement area abuts the Gunsan
Basin, from which it is separated by faults. This deep magnetic basement
area appears to be a northeastern extension of the seismically defined
North Heuksan Basin.
Nevertheless, a large areal distribution with granitic rocks and
gneisses (granitic) possessing very low-magnetic susceptibility values (e.g.,
subaluminous granite which lacks hornblende, biotite, and magnetite)
may exhibit similar magnetic characteristics. However, granitic rocks usu-
ally occur in a contaminated or mixed form containing varying amounts
of mafic minerals within an area. Geologically, a region as large as the
newly discovered deep magnetic basement province, occupied by homo-
geneously nonmagnetic igneous rocks, is considered very rare. Therefore,
the most plausible explanation, since no earlier seismic sections had iden-
tified the area as a sedimentary basin, is that the area is overlain by a non-
magnetic, high-velocity layer.
In view of the above geologic and geophysical considerations, the
new deep magnetic basement surface can be interpreted best as a deep
sedimentary basin, named the Heuksan Basin. The Heuksan Basin is com-
posed of three smaller subbasins, which are interpreted to have favorable
structures for hydrocarbon accumulation.
Two subbasins, east and central (Saltmarsh, 1989), comprise the
Gunsan Basin in the study area. The central subbasin appears to be struc-
turally complex compared with the east. Saltmarsh
(1989) states that horsts, grabens, and other rotated fault
blocks originate from both extensional (earlier-time) and
compressional (later-time) tectonic episodes. The east
subbasin is oriented north-northwest–south-southeast,
whereas the central has no predominant orientation.
Genesis of the Gunsan and Heuksan Basins was con-
trolled by the regional fault system.
Tectonic Style—Gross Structural Style
The basin-forming tectonic style for the Gunsan and
Heuksan Basins is interpreted to be rhombochasm (i.e.,
divergent overstep; pull-apart), associated with double-
stepped left-lateral wrenches (Figures 3, 4, and 5). The
interpreted magnetic basement map shows that all three
subbasins of the Heuksan Basin and the East Gunsan
Subbasin are roughly elongated northwest-southeast per-
pendicular to the northeast-southwest-striking wrenches.
The shape and trends of the subbasins with respect to the
faults strongly indicate that the origin of the basins results
from overstepped divergent wrench faults. We believe
that a wrench originating from onshore Korea or from far-
ther east terminates at the northern margin of the
Gunsan Basin in Block I. Then the fault trace steps left,
forming areas of tension, i.e., the pull-aparts of the
Gunsan Basin (Figures 3 and 4). The fault steps left a sec-
ond time, forming the Heuksan Basin, and finally the
trace continues to the Tienmushan-Baijishan wrench of
China (Xu et al., 1987).
We find no evidence to support or deny the inter-
pretation that horsts in the Gunsan and Heuksan Basins
originated from compressional stress believed to have
been applied since Late Neogene in the study area
(Saltmarsh, 1989). However, from our interpretation, it
is possible to say that the north-northwest–south-south-
east horsts are original basement left behind by grabens
formed by left-stepping. It should be noted that our inter-
preted senses of wrench movement are in agreement
with those in China (Xu et al., 1987). Furthermore, the
formation and development of the basins hardly can be
explained by any right-lateral wrenches. The results of
study of the Yeongdong Basin (Lee, 1990) support our Figure 2. Aeromagnetic anomaly map of southwestern continental shelf of Korea. Digitized by LCT Inc. from
the original map of Project Magnet Aeromagnetic Survey, 1969.
explanation.
65
Geologic Applications of Gravity and Magnetics: Case Histories
03 Basin Studies 4/16/07 4:16 PM Page 66

Basin Studies
66
Although the aeromagnetic data are of recon-
naissance quality, the data were very useful in mak-
ing the following valuable interpretations:

1) An extensive magnetic lineament indicates a

major, extensive, basement-involved fault system.

2) Distinctively different basement rocks are

found, with differing structural patterns juxta-
posed across the faults.

3) Displacement of antithetic faults and apparent

left-lateral displacement of interpreted acidic
basement rocks along the fault system are found
in the southwestern part of the study area.

4) Many of the interpreted numerous submarine

igneous intrusives and extrusives in the area can
be ascribed to have intruded into tensional
cracks in the fault system.

5) Both the Gongju and Gwangju-Yeongdong

Faults are left-lateral wrenches.

2.5-D Magnetic Model (Figure 6)

A model profile, A–A1, was made to supplement the
magnetic basement surface configuration study.
Two-and-a-half-dimensional calculations for the
total-magnetic-intensity anomaly were carried out.
The model magnetic depth configurations sug-
gest that the East Gunsan Subbasin is as deep as
7500 m below sea level. The subbasin shows a sym-
metrical graben structure with a few sets of normal
faults. The sedimentary rock-body geometry is arbi-
trary, because no gravity data were available.
Figure 6 shows that magnetic susceptibilities
decrease gradually toward the west. This strongly
supports the interpretation results, in that materials
beneath the South Yellow Sea are highly fractured
Figure 3. Interpreted magnetic basement surface map showing simplified structural features such as basins and
faults of southwestern continental shelf of Korea: W-1, Changle-Nanao Fault; W-2, Lishui-Haifong fault; W-3, and softened by the high heat flow (Jun and
Tienmushan-Baijishan fault (Xu et al., 1987); B-1, Jincheon Basin; B-2, Yeongdong Basin. Roman numerals indi- Kulhánek, 1991).
cate block number.
Figure 4. Sketch map showing horsts and grabens (i.e., pull-aparts) associated with
double-overstepped left-lateral wrenches.
Discussion
Because the magnetic data show no suitable anomalies to use for mag-
netic-depth computation from Heuksan Basin and because the data are
incomplete in the west of the basin, no meaningful modeling could be
performed or appropriate magnetic-depth estimates made.
Pull-apart basins associated with single-overstepped wrench fault sys-
tems are common (e.g., Segall and Pollard, 1980; Harding et al., 1985).
However, we believe that our interpreted double-overstepped wrenches
are the first of their kind.
03 Basin Studies 4/16/07 4:16 PM Page 67
Figure 5. Schematic diagram showing horsts which can be associated with double-
overstepped right-lateral wrenches.
It is very likely that the interpreted volcanic rocks south of the INGA
well (Figure 3), across the southern border fault, may overlie a thick sed-
imentary section of the Heuksan Basin. If this is the case, the Gunsan and
Heuksan Basins can be interpreted to abut each other along a common
border fault, as they do at the western part of the basin. Consequently,
the two basins could be interpreted as originating from a single-over-
stepped wrench fault.
Contrary to the Chinese observation (Xu et al., 1987), displacement
senses for the faults in the Korean portion of the South Yellow Sea were
interpreted by Chun (1989) to be right-lateral. Chun based his interpre-
tation mainly on seismic data, of which coverage is limited only to the
hydrocarbon prospects.
Lee (1990) showed geologic evidence indicating that the Yeongdong
Basin onshore Korea originates as a pull-apart basin by left-lateral motion
on the Gwangju-Yeongdong wrench system. He also suggested that sev-
eral other basins (e.g., Jincheon Basin, Figure 1) onshore Korea may have
been formed and developed by similar mechanisms (personal communi- Conclusions
cation, 1991). Results of Lee’s (1990) study strongly support our inter-
1) The results of our aeromagnetic interpretation show that a new thick
pretation. A wrench fault must be consistent in terms of lateral
sedimentary basin, Heuksan Basin, is situated south of the Gunsan
movement; otherwise, the fault may not be called a wrench.
Basin across its southern border fault. The Heuksan Basin appears to
Saltmarsh (1989) states that the border faults along the northern and
consists of three subbasins trending north-northwest–south-southeast.
southern flanks of the Gunsan Basin show left-lateral movement.
The basin appears to be a northeastward extension of the North
Moreover, paleomagnetic results (Hanston and Fuller, 1991) from the
Heuksan Basin, which had been identified by previous seismic data.
Philippine Sea plate indicate that the plate has rotated as much as 80˚
clockwise since Eocene, which also supports our interpretation.
Our interpreted magnetic basement map does not show any evidence 2) We infer that the Heuksan Basin is overlain by a nonmagnetic high-
for the “Cheju volcanic belt” (e.g., Frazier et al., 1974; Chough, 1983; velocity (i.e., a high-reflection coefficient) layer so that the possibly
Han, 1984; Li, 1984; Chun, 1989). From our investigation, it is difficult thick sedimentary basin may not have been identified by earlier seis-
to see a basement high trending east-west that spans the South Yellow Sea mic surveys.
between Cheju Island and China.
3) The Gunsan Basin comprises East
and Central Subbasins (Saltmarsh,
1989). The East Subbasin appears to
show a very deep magnetic basement
and the Central Subbasin a shallower
one. The Central Subbasin can be
divided into three smaller basins.
4) The basin-forming tectonic style for
both the Gunsan and Heuksan Basins is
interpreted to be pull-apart basins asso-
ciated with double-overstepped left-lat-
eral wrenches.
5) Traces of major onshore northeast-
southwest wrenches such as Gongju
Fault and Gwangju-Yeongdong Fault
are extended to the offshore area. From
magnetic interpretation, we define
these faults as left-lateral wrenches. The
left-lateral senses are consistent with
those of wrenches onshore China.
6) The results of our investigation do
Figure 6. 2.5-D magnetic model along profile A-A1 for East Gunsan Subbasin (see Figure 2 for profile location): F (total intensity), not support the reported Cheju Volcanic
D (declination), and I (inclination) values are calculated from IGRF for March 1, 1969; κ is magnetic susceptibility in unrational-
ized SI unit; Mr, Ir, and Dr, are assumed values for remanent-magnetization intensity, inclination, and declination, respectively;
Belt (e.g., Frazier et al.,1974; Chun,
1989).
“y = ± 10 Km” indicates body length along the axis perpendicular to the page surface.
67
Geologic Applications of Gravity and Magnetics: Case Histories
03 Basin Studies 4/16/07 4:16 PM Page 68
Recommendations
1) All seismic sections, if any, that cross the newly interpreted Heuksan
Basin area should be reprocessed and rechecked carefully, bearing in
mind an overlying high-velocity layer.
2) Because the newly interpreted Heuksan Basin seems to have favor-
able structures for hydrocarbon accumulation, a detailed aeromag-
netic survey is recommended strongly, to verify the magnetic
expression of deeply buried basement surface.
3) At least a few lines of shipborne gravity survey are recommended, in
conjunction with the detailed aeromagnetic survey, to identify inde-
pendently any thick sediments in the basin. If both gravity and mag-
netic surveys show favorable, encouraging results, a carefully
designed and effective (e.g., high-energy source) seismic survey
should be carried out to generate prospects.
Acknowledgments
This work was carried out originally under a contract with the Korea
Petroleum Development Corporation (KODECO). The authors would like
to express their sincere gratitude to KODECO for permitting us to publish
this paper. We wish to thank Richard I. Gibson and Patrick S. Millegan for
their valuable comments and suggestions, which improved our manu-
script significantly.
References
Bosum, W., Kind, E. G., and Koo, J. H., 1971, Aeromagnetic survey of offshore areas
adjoining the Korea Peninsula: U. N. ECAFE, CCOP, Technical Bulletin, 4, 1–21.
Chough, S. K., 1983, Marine geology of Korean Seas: International Human Resources
Development Corp.
Chun, S. S., 1989, Sedimentary processes, depositional environments and tectonic set-
tings of the Cretaceous Uhangri Formation, Southwest Korea: Ph.D. dissertation,
Seoul Nat. Univ.
Frazier, S. B., Choi, S. 0., Kim, B. K., and Schwartz, D., 1976, Marine petroleum explo-
ration of Huksan Platform, Korea: AAPG Mem., 25, 268–275.
Han, J. H. 1984, Basin evolution and genetic stratigraphy of sediments in the central
Yellow Sea: Korea-U.S. Seminar and Workshop on Marine Geology and Processes of
the Yellow Sea, Abstracts, 13–15.
Hanston, R. B., and Fuller, M., 1991, Paleomagnetic data from the Philippine Sea Plate
and their tectonic significance: J. Geophys. Res., B. Paper 90JB02700.
Harding, T. P., Vierbuchen, R. C., and Christie-Blick, N., 1985, Structural styles, plate-
tectonic settings, and hydrocarbon traps of divergent (transtensional) wrench faults,
Basin Studies
in Biddle, K. T., and Christie-Blick, N., Eds., Strike-slip deformation, basin forma-
tion, and sedimentation: SEPM Soc. for Sed. Geol.., Special Pub. No. 37, 52–77.
Jun, Myung-Soon, and Kuihánek, 0., 1991, Source parameters of earthquakes in and
around the Korean Peninsula deduced from spectral analysis: Physics of Earth and
Planetary Inter., 65, 255–266.
Korea Institute of Energy and Resources, 1984, Geophysical data interpretation in Block
II, offshore Korea.
——— 1985, Geophysical data interpretation in Block IV, offshore Korea.
——— 1990, Geological and geophysical characteristics of the Yellow Sea.
Lee, Dong-Woo, 1990, Sedimentation and tectonic evolution of the Cretaceous
Yongdong Basin, Korea: Ph.D. dissertation, Korea Univ.
Li, D., 1984, Geologic evolution of petroliferous basins on the continental shelf of China:
AAPG Bull., 68, 993–1003.
Saltmarsh, A. C., 1989, Final well report, INGA-1, Block II, Yellow Sea, South Korea:
Marathon Petroleum Korea, LTD.
Segall, P., and Pollard, D. D., 1980, Mechanics of discontinuous faults: J. Geophys. Res.,
85, 4337–4350.
Xu, J., Zhu, G., Tong, W., Cui, K., and Liu, Q., 1987, Formation and evolution of the
Tancheng-Lujiang wrench fault system: A major shear system to the northwest of
the Pacific Ocean: Tectonophysics, 134, 273–310.
68
Integrated Gravity Modeling of Salt
Features in the Mississippi Salt Basin
Rhonda L. Schenk
International Paper Petroleum Co.
Houston, Texas, USA
John J. Morris
Savannah Oil and Gas L.L.C.
Houston, Texas, USA
Stuart A. Hall
Department of Geosciences, University of Houston
Houston, Texas, USA
EDITOR’S NOTE: An excellent example of an integrated interpreta-
tion, this case study uses gravity data to constrain the seismic
interpretation. The resulting increase in inferred subsurface
complexity provides a much better understanding of the struc-
tural traps.
Abstract
Recent exploration in portions of the Mississippi Salt Basin indicates that
some models used in the interpretation of surface and subsurface data
may be inadequate to explain several salt features. Models are important
in interpretation because of the depth of the objectives and the complex-
ity of the salt ridges. Significant oil and gas reserves trapped by the salt
and related faults in Upper Jurassic reservoirs were discovered from the
late 1960s to the early 1980s. Early models show concordant overburden
on the flanks of a simple, symmetrical salt anticline. However, with fur-
ther drilling, Upper Jurassic formations and the underlying salt were not
encountered where predicted by the anticlinal model, suggesting that
many of the salt ridges in the basin are more complex. Well-log data indi-
cate that the flanks of such salt features may dip at rates greater than
those proposed by the symmetrical anticline model.
A study using an integrated approach to test the viability of a more
complex salt-ridge model was conducted to aid in the interpretation of sur-
face and subsurface data in the Mississippi Salt Basin. Gravity data in con-
junction with seismic reflection data and well-log information form the
database from which salt models were created. Seismic reflection data and
well control provide a reasonable estimate of the depth of the sources of
03 Basin Studies 4/16/07 4:16 PM Page 69
the gravity anomalies. The gravity data, on the other hand, have been used
to help constrain the interpretation of the seismic data by providing a bet-
ter geometric framework for the salt bodies observed on the seismic data.
Eucutta Field and Diamond Field were selected as test areas for the
salt models, based on the availability of data. The two fields are located
along different types of parallel salt ridges. Eucutta is located deeper in the
basin, in an area of large salt features, and Diamond is located closer to
the rim of the basin, where smaller salt features are more common.
Results of the study indicate that an asymmetrical pattern exists in
which the flanks of the ridges may dip at different rates. Both the Eucutta
and Diamond salt ridges exhibit a steeply dipping flank on the basinward
side of the salt ridge and a more gently dipping landward flank. The salt
ridges also appear to have a complex variation along strike in the struc-
ture of the salt. The variations of the salt features in both the strike and
dip directions appear to reflect the variations in the way the overlying sed-
iments were deposited.
Introduction
Many of the fields discovered in Mississippi during the 1960s and 1970s,
such as Nancy, Patchuta Creek, and Prairie Branch Fields, were drilled on
the assumption that Upper Jurassic reservoirs were overlying a symmetri-
cal salt anticline and were dipping at the same rate. During that time,
Hughes (1968) described four types of salt anticlines in the basin. They
Figure 1. A generalized Jurassic stratigraphic column of the Mississippi Salt Basin
(after Schenk, 1994).
are, beginning with the structure closest to the
updip limit of the basin: (1) salt ridges parallel
to the wedge-out of salt, (2) low-relief salt pil-
lows, (3) intermediate-relief salt anticlines
with a primary graben, and (4) high-relief salt
anticlines with a primary and secondary
graben forming on the crest. Drilling during
the 1970s and 1980s indicated that the rela-
tionship between the salt features and the
reservoirs is much more complex than the
anticlinal model previously proposed.
Structural models sometimes are used in the
interpretation of seismic data in this area
because the salt-sediment interface is hard to
image on seismic data due to the steeply dip-
ping flanks of salt ridges.
An integrated study using gravity data in
conjunction with seismic data and well data
was conducted to investigate an alternative
salt model for exploration of Upper Jurassic
reservoirs in the Mississippi Salt Basin. Figure
1 is a generalized Jurassic stratigraphic column
for the Mississippi Salt Basin. Two fields were
used, Eucutta and Diamond, both in Wayne
County. Figure 2 shows the location of the
fields and salt ridges in the study area. Eucutta
Field is located in the zone that Hughes
described as high-relief salt anticlines, and
Diamond Field is located in the intermediate-
relief zone. The fields were selected based on Figure 2. Regional map of Jurassic fields and salt ridges in Wayne, Clarke, Jasper, and Jones Counties, Mississippi, with
the availability of data and the problems with the study area and analogous models highlighted.
interpreting the salt-sediment interface on 2-D
seismic data. By using gravity data in combina- gram. Ten tops were chosen on the logs and were used to create different
tion with seismic reflection data, better geometric constraints can be sediment packages or blocks overlying the salt in the gravity modeling.
placed on viable models of salt features. For this study, a sediment package contains the sedimentary layers
from the top of a formation chosen on the well log to the top of the for-
Methodology and Results mation below it. Doing this results in nine sedimentary packages, which
The study used thirty-seven density logs and dual induction logs, three are, from the most recent to the basement, (1) Quaternary (Qt), (2) Eutaw
seismic profiles oriented perpendicular to the salt ridges, and a residual (Eu), (3) Lower Tuscaloosa (L/T), (4) Rodessa (Rod), (5) Cotton Valley
gravity map covering Wayne County. The seismic data were interpreted (CV), (6) Haynesville (HV), (7) Smackover/Norphlet (Smk), (8) Louann
using the well logs and velocity control from each field. The well logs Salt, and (9) Werner-Paleozoic basement (Pz basement). The density well
served to identify the salt limits and the type and depth of the sedimentary logs then were used to determine an average density for each sedimen-
layers. The interpretation then was used in a forward gravity modeling pro- tary package or block.
69
Geologic Applications of Gravity and Magnetics: Case Histories
03 Basin Studies 4/16/07 4:16 PM Page 70
Figure 3. Residual gravity map over Eucutta Field with seismic lines superimposed. Contour intervals are 1 gravity unit for values
between –20 and 20, and 2 gravity units for values greater than 20 and less than –20 (after Schenk, 1994).
Because well-log data and seismic data are used in the study, the
depth of the sources for the gravity anomalies caused by the salt are
known in this area. For this study, salt is assumed to have a uniform den-
sity of 2.2 g/cc and, at the depths of the salt features found in the basin,
is less dense than the overlying sediments, thus causing a negative anom-
aly on the gravity data. The gravity anomaly over Eucutta Field (Figure 3)
is much larger and is elliptical in shape, and the anomaly over Diamond
Field is smaller and circular (Figure 4).
The model generated from each seismic profile then was used in a for-
ward gravity modeling program. A gravity profile using residual values
from stations lying along the seismic profiles was entered into the model-
ing program, which calculates gravity values given the model and densi-
ties used. The points creating the outline of the salt and the overlying
Jurassic sediments on the model then were adjusted to fit calculated grav-
ity values to the observed gravity profile. This was done on all the seismic
lines over Eucutta and Diamond Fields.
Basin Studies
Figure 4. Residual gravity map over Diamond Field with seismic lines superimposed . Contour values are 1 gravity unit for values
between –20 and 20, and 2 gravity units for values greater than 20 and less than –20 (after Schenk, 1994).
In Eucutta Field, the profiles cross the salt feature at each end and
through the center. The points making up the flanks of the salt were
moved into a much more steeply dipping flank to fit the calculated grav-
ity data to observed gravity data on all three seismic lines that crossed the
salt feature. However, the points in the model over the center of the
feature had to be moved to a steeper position than the points in the two
models at the end of the feature. The center of the feature is interpreted
to be much more complex than the edges, where the salt begins to lose
elevation. The final model displays an asymmetrical salt feature with a
steeper-dipping flank on the basinward side and a less steeply dipping fea-
ture on the landward side. Figure 5 shows the resulting model and grav-
ity profile over the center of Eucutta Field.
In Diamond Field, only two of the three seismic profiles crossed the
salt feature, one through the middle and one on the north end. Again,
points in the model describing the salt flank were moved into a steeper
position, compared with the original interpretation from the seismic data.
70
The salt ridge at Diamond is interpreted to be asymmetrical in shape with
a steeper dipping flank on the basinward side and a less steeply dipping
flank on the landward side. The salt feature appears to be more complex
in the center and loses complexity as the salt loses amplitude to the north
and south. Figure 6 shows the resulting model and gravity profile over the
center of the Diamond salt ridge.
Conclusions
The modeling suggests that salt ridges in the Mississippi Salt Basin are a
composite or hybrid feature of steeply dipping salt flanks and gently dip-
ping salt flanks. The two fields used in the study indicate that Upper
Jurassic reservoirs beds have been truncated by a steeply dipping salt face
with no concordant overburden on the basinward side, while the land-
ward-facing limb is less steeply dipping, with concordant overburden. The
salt features appear to vary in amplitude in the strike direction, with the
center of the salt features having the largest amplitude and a more com-
03 Basin Studies 4/16/07 4:16 PM Page 71
Figure 5. Forward gravity model over Eucutta Field after final iterations showing observed and calculated gravity val-
ues and density blocks (after Schenk, 1994).
plex faulting structure. Parallel ridges in the basin are interpreted to be
nonuniform in amplitude, as seen when comparing Diamond with the
Cypress Creek Ridge close by. The resulting forward model generated in
this study fits all data available—well logs, seismic data, and gravity data.
Although this study suggests that the more steeply dipping flank of the
salt feature is on the basinward side, a feature that is symmetrical or with
a steeply dipping flank on the landward side is possible. The causal mech-
anism for the shape of the salt is interpreted to be deposition of the over-
lying Haynesville and Cotton Valley sediments. The growth of the feature
is related directly to the availability of the salt source from all sides.
Figure 6. Forward gravity model after final iterations over Diamond Field showing calculated and observed gravity
values with different density blocks (after Schenk, 1994).
The juxtaposition of Upper Jurassic Smackover and Norphlet reservoir
beds to Louann Salt is important because the resulting trap geometry may
provide ideal sealing conditions. The potential exists for significant hydro-
carbon accumulations in the Upper Jurassic Smackover and Norphlet,
sealed laterally by salt and capped by thick lower Haynesville evaporitic
sections. Example fields found to date include West Chaparral in Wayne
County, Mississippi, and Hatters Pond in Mobile County, Alabama.
Geologic Applications of Gravity and Magnetics: Case Histories
References
Hughes, D., 1968, Salt tectonics as related to several Smackover fields along the north-
east rim of the Gulf of Mexico Basin: Transactions of Gulf Coast Association of
Geological Society, 18, 320-–30.
Schenk, R. L., 1994, Integrated gravity modeling of salt features in the Central
Mississippi Salt Basin, master’s thesis, Univ. of Houston.
71
03 Basin Studies 4/16/07 4:16 PM Page 72
Mississippi Salt Basin
In early 1937, a salt dome was defined
by Gulf Oil in Lamar County, Mississippi,
the first to be found in that state. Oil dis-
coveries related to salt domes identified
by gravity surveys proceeded at an aver-
age rate of one per week for more than
twenty weeks, according to Eckhardt, so
it can be said with justification that Gulf
Oil discovered the Mississippi Salt Basin
through gravity surveying.
The Mississippi Salt Basin is a regional
gravity high, and the Wiggins Arch,
south of the basin, is a regional gravity
low. One explanation for this calls on
crust-mantle effects, as illustrated here.
The basin would be a “neck” of thinned
crust associated with uplift of dense man-
tle, and the Wiggins Arch would be
interpreted as a block of low-density
crust left behind as the Gulf of Mexico
opened when Yucatán rotated away.
R. I. Gibson
Basin Studies
72
Kentucky Basement Note from the published USGS Geologic Map
of the United States. Linear magnetic
anomalies and gradients correlate nicely
It is true, as Betty Johnson says, that
with many faults and fault zones in the
“integration is more than getting the
nonmagnetic sedimentary section at the
maps at the same scale.” However, the
surface, implying a relationship between
exercise of comparing diverse data sets at
basement features and structures in the
the same scale is a simple way of discov-
prospective section. If nothing else, this
ering possible relationships that should
suggests that some of the magnetic
be studied further. This example shows
anomalies could be used predictively to
the magnetic map of Kentucky (from
suggest the locations of additional fault-
USGS GP-954-A), with surface faults,
ing or fracturing.
R. I. Gibson
04 Plays-Field Studies 4/16/07 4:34 PM Page 73

4
An Integrated 3-D Gravity and Seismic Study of the Ouachita Frontal Thrust Belt,
Pittsburg and Latimer Counties, Oklahoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Holly Hunter-Huston, Hunter 3-D Inc.
E. F. Greene, Gene Greene & Associates,
C. L. V. Aiken, University of Texas at Dallas
Interpretation of Rift-stage Faulting in the West Siberian Basin from
Magnetic Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Richard I. Gibson, Gibson Consulting
Case History: Reddell Oil Field, Evangeline Parish, Louisiana . . . . . . . . . . . . . . . . . .
79

82
Plays/
C. T. Austin, Austin Exploration, Inc.
Magnetic Anomalies Associated with the North and South
Morecambe Fields, U. K. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Richard Morgan, World Geoscience UK Ltd.
High-resolution Aeromagnetic Interpretation over Sierra and Yoyo Reefs,
85 Field Studies
Northeastern British Columbia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
John Peirce, Erwin Ebner, and Nathalie Marchand
Geophysical Exploration and Development Corporation (GEDCO)
04 Plays-Field Studies 4/16/07 4:34 PM Page 74
Integration of Gravity and
Magnetic Methods in the
Risk-weighted Exploration
Decision Process: Prospect
Definition Stage
Robert Pawlowski
Integrated Geophysics Corporation, USA
(formerly with Amoco Production Company)
Exploration-risk parameters evaluated
through the geologic integration of grav-
ity and magnetic data during the
prospect definition stage include:
1) definition of structure in seismically
difficult-to-image areas
2) structural focusing of hydrocarbon
migration
3) porosity enhancement
4) nonhydrocarbon amplitude-variation-
with-offset (AVO) response
5) improving initial prestack migration
velocity models
6) verification of seismic structural inter-
pretation
At the prospect definition stage, a com-
pany has completed the fundamental
basin reconnaissance work, has selected
a prospect (or prospects), and is doing
detailed evaluation of the prospect before
deciding whether to commit drilling
money. Gravity and magnetic methods
can be applied again at this point to con-
tribute to the evaluation of various risk
parameters.
The most common application of gravity
and magnetics at this stage of exploration
is to better define geologic structure in
areas of poor seismic data quality. For
autochthonous sedimentary structures,
magnetic data allow the geometry and
depth of underlying basement structures
to be determined. This is useful in struc-
turally complex areas involving over-
thrusting and folding, such as that
occurring in the Carpathian Mountains
of Romania and the Dinaride Mountains
of Yugoslavia. Potential-field methods
often permit deep structure to be deter-
mined where the deep seismic image is
poor because of the often high acoustic
impedances and steep dips of the con-
torted structure within the overthrust
units.
Similarly, where intrasedimentary
igneous material occurs, as with the
basalt flows of the Chaco-Parana Basin of
Argentina, potential-field methods can be
used to peer beneath the seismically
reflective barrier presented by the basalt.
Gravity and magnetic data also can be
used to resolve seismic no-data zones
caused by igneous intrusions or fossil vol-
canoes, such as those offshore the
Netherlands in the North Sea (see also
the St. George Basin study by Chapin et
al. in this volume).
Another example is the case of the so-
called bald high, a basement-cored sedi-
mentary structure whose top is devoid
of a significant portion (if not all) of
prospective sedimentary strata. Depth-to-
magnetic-basement mapping is useful for
interpreting many such features in cases
where the seismic imaging of basement
is nebulous, such as in portions of the
Norwegian North Sea.
In seismic subsalt imaging, such as in the
Gulf of Mexico or offshore the west
African coast, gravity data can be used to
determine whether a salt mass is rooted
or detached. Where a widespread layer
of salt makes the imaging of subsalt
structures (e.g., grabens and horsts) diffi-
cult, gravity and magnetics can be used
Plays/Field Studies
to define interesting basement- and non-
basement-rooted sedimentary structures,
and the relation of such structures to
possible depocenters. The mapping of
subsalt structures and depocenters per-
mits the location of probable hydrocar-
bon migration fairways to be interpreted,
both beneath and through the salt,
where the latter is breached or absent.
The geophysical detection of porosity
enhancement is desirable for evaluating
hydrocarbon migration pathways as well
as hydrocarbon reservoir quality. Porosity
enhancement is not directly detectable
using surface gravity and magnetic mea-
surements. Porosity detection via bore-
hole gravimetry is, however, well
documented (see the application to
Michigan reefs documented by
Rasmussen, 1975). As an alternative to
detecting porosity directly, gravity and
magnetic interpretation techniques can
be employed to detect features contribut-
ing to the development of enhanced
porosity. One way in which enhanced
porosity develops is through fracturing
and the attendant chemical dissolution of
the host rock along the fractures (so-
called fracture porosity). Such fracturing
often is related to basement tectonics,
often occurring along basement-tectonic
hinge zones. The mapping of basement-
tectonic blocks is one of the classic uses
of aeromagnetic data. Accordingly, the
prediction of enhanced porosity zones
(and their relation to reservoir quality
and even migration efficiency) can be
used to reduce drilling risk, particularly
in exploitation (extensional exploration)
problems.
In cases of seismic AVO (amplitude varia-
tion with offset) anomalies, magnetic
anomaly analysis continues to be useful
for differentiating intrasedimentary
74
igneous material (e.g., volcanic flows, Lastly, prospect risk can be reduced by
ash, sills, etc.) from gas-bearing sedimen- verifying the seismic structural interpre-
tary strata, thus aiding in the identifica- tation through prospect-scale gravity and
tion of nonhydrocarbon-sourced seismic magnetic modeling, using the seismic
AVO anomalies. The caveat should be interpretation and any available well
added that the magnetic technique information as constraints. When neces-
works for this purpose when the igneous sary, the seismic interpretation can be
material is magnetic (and thus detect- revised to honor the additional con-
able). It always must be borne in mind straints imposed by the gravity and mag-
that certain geologic processes can result netic data, thus reducing exploration
in the alteration/destruction of a rock’s risk.
ferromagnetic minerals, thus rendering it
nonmagnetic (and thus not detectable). Conclusion
Furthermore, many such seismic AVO
Clearly, a principal exploration benefit
anomalies clearly are seen over more
arising from the geologic integration of
deeply seated igneous plutons or feeder
gravity and magnetic data is the quantifi-
stocks after gravity and magnetic data are
cation and refinement of certain of the
integrated in the interpretation process.
exploration-risk model variables, both at
the scale of basin reconnaissance and at
Three-dimensional seismic prestack
the scale of prospect definition, which
migration is currently popular for imag-
ultimately improves the estimated
ing complex structures such as sediments
chance of success (or failure) of finding
underlying the salt sills and overhangs in
commercial quantities of hydrocarbons.
the Gulf of Mexico. Although such tech-
Careful gravity and magnetic data analy-
nology is impressive in its imaging capa-
sis enhances the application and interpre-
bility, it is equally impressive from the
tation of subsequent exploration data
point of view of its cost. Common to the
sets, such as seismic data, helping to
process is the construction of an initial
reduce risk even further. Gravity and
velocity model. The initial velocity model
magnetic exploration methods yield cost-
typically is characterized as being quite
effective, supplemental constraining
smooth and general in its detail. The
information, their cost of implementation
velocity model is refined through subse-
being fractional in comparison with cor-
quent iterations of the prestack migration
responding seismic work program expen-
process. Gravity and magnetic structural
ditures.
modeling can be used to obtain an initial
starting model (e.g., by better defining
the salt-feature geometry, basement geo-
Reference
metry, etc.) which is closer to reality, Rasmussen, N. F., 1975, The successful use of
while still having the required degree of the borehole gravity meter in northern
Michigan: The Log Analyst, September-
smoothness. The additional expected October.
benefit is that the required number of
migration iterations can be reduced, thus
saving time and money which can be in-
vested elsewhere in evaluating the pros-
pect and its risk. This application of gra-
vity and magnetic analysis is underused.
04 Plays-Field Studies 4/16/07 4:34 PM Page 75
An Integrated 3-D Gravity and Seismic
Study of the Ouachita Frontal Thrust Belt,
Pittsburg and Latimer Counties, Oklahoma
Holly Hunter-Huston
Hunter 3-D Inc., Houston, Texas, USA
E. F. Greene*
Gene Greene & Associates, Houston, Texas, USA
C. L. V. Aiken
University of Texas at Dallas, Texas, USA
* Deceased
EDITOR’S NOTE: This paper shows how even high-quality 3-D seis-
mic data can benefit from integration with gravity data. Simple
residuals along with 3-D and 2.5-D models provide significant
insight into the subsurface of this complex thrust belt.
Introduction and Tectonic Setting
The study area is located in the Ouachita frontal thrust belt between the
Choctaw thrust fault and the Ti Valley thrust fault in Pittsburg and
Latimer Counties, Oklahoma (Figure 1). Although previous studies are
available on the compressional structures in this area, few studies have
analyzed the deep Ordovician extensional structures we focused on. A
detailed gravity survey was acquired to complement an existing 3-D seis-
mic survey shot by Texaco and Weeks Exploration. The seismic data set
is ambiguous at the Ordovician level; 3-D gravity modeling proved to be
critical in confirming large, deep structural highs within these strata.
Deep extensional structures of the Ouachita frontal thrust belt were
developed during Precambrian rifting that created the Arkoma Basin
shelf. Reactivation along these basement faults tilted and truncated the
section during post-Hunton time, causing erosion of the Hunton Group in
many parts of the Arkoma Basin and Ouachita overthrust belt (Tilford,
l990). These faults are characterized by localized horst and graben fea-
tures. On seismic data, some of these faults appear to be high-angle thrust
faults. The throw probably was reversed during the period of Ouachita
Mountain overthrusting (Tilford, l990; Valderrama, 1994).
Seismic Data
The 233-km2 3-D seismic data set was used to inter-
pret the compressional and extensional structures.
Interpretation of the seismic data was supplemented
by well logs, dipmeter surveys, velocity surveys, sur-
face geology, and gravity data. Key marker horizons
are the tops of the Shallow Spiro Thrust, Middle
Spiro, Deep Spiro, and the Viola strata. Structure
maps at these horizons were constructed. During
initial 2-D gravity modeling of the study area, the
Middle and Deep Spiro compressional structures
appear invisible to the gravity data, with a lateral
density contrast of +.05 gm/cc. These horizons lie
at an average depth of 3.0 to 3.6 km and have an
average thickness of 60 m. Structure maps of the
Shallow Spiro Thrust (Figure 3) and Viola Formation
(Figure 4) were used to help constrain gravity mod-
eling for this study. In areas where the seismic data
were ambiguous because of complex faulting or low
fold, gravity modeling assisted in the generation of
these structure maps.
Figure 1. Location map showing the study area in Oklahoma. The surface trace of the main faults in the frontal
belt is shown for reference.
Gravity Data
More than 1100 gravity stations, at a spacing of 268
Compressional structures of the Ouachita frontal thrust belt are com-
m, were collected along thirty-two seismic receiver lines (Figure 5). These
prised of deep and shallow regimes. These structures formed during the
gravity profiles are 686 m apart, 9 km long, and oriented perpendicular
compression of the Ouachita Mountains overthrust orogeny in late
to the southwest-trending structural grain. Processing of the gravity data
Atokan through Des Moinesian time (Arbenz, l989). The shallow com-
included standard corrections for tides, instrument drift, latitude, eleva-
pressional structures consist of asymmetric north-verging faults. Most of
tion, density, and topography. The triplet method was used to derive an
these faults become listric at depth and parallel the Choctaw and Ti Valley
average density for the Bouguer correction. Densities ranged from 2.45
thrust faults. Deeper compressional structures are extremely complex and
gm/cc to 2.55 gm/cc; an average density of 2.50 gm/cc was chosen for
consist of closely spaced imbricate thrust sheets stacked in multilevel
the Bouguer correction. Terrain corrections were applied by Photo
duplexes and tight overturned recumbent folds (Tilford, l990). The Spiro
Gravity Inc. of Houston, Texas. Gravity data were gridded on a 305-m
and Wapanucka comprise the main formations that record the extensive
grid node spacing using the inverse-distance method of gridding.
compressional deformation.
A first-order polynomial trend surface was fitted to the complete
This area is part of the South Hartshorne gas field, with production in
Bouguer gravity anomaly map. Subtraction of the regional surface from
the Atoka, Spiro, and Wapanucka Formations. To the northeast, the
the Bouguer gravity data generated a first-order residual (Figure 6). The
Wilburton gas field also produces from the Arbuckle Formation. The
first-order residual map shows several gravity highs and lows. Two promi-
stratigraphy is summarized in Figure 2.
nent long-wavelength gravity highs are in the northeast and southwest
corners of the survey. Smaller-wavelength gravity highs are superimposed
on the longer wavelengths. Gravity lows are in the southeast and north-
west corners of the survey.
75
Geologic Applications of Gravity and Magnetics: Case Histories
04 Plays-Field Studies 4/16/07 4:34 PM Page 76
Figure 2. Stratigraphic column for the frontal zone of the Ouachita Mountains (modi-
fied from Suneson and others, 1990).
A 3-D forward gravity model (Figure 7) was generated from the Viola
Formation structure map to separate anomalies of interest and to deter-
mine source depth. The 3-D model was constrained with structure and
density derived from seismic and well data. In the southwest portion of
the survey, the seismic data are ambiguous at the Viola horizon. The mod-
eled structures of the Viola Formation were altered and recalculated to
obtain a better fit between the calculated anomaly and first-order residual
for the long-wavelength anomaly in the southwestern portion of the sur-
vey. This procedure was repeated through several iterations until a best fit
was found, while remaining within the constraints of the seismic data and
subsurface data. The 3-D gravity anomaly map in Figure 7 shows two
long-wavelength highs, to the northeast and southwest of the survey,
with an amplitude of approximately 2.60 mGal. These anomalies corre-
Plays/Field Studies
Figure 3. Structure map on top of the Shallow Spiro Thrust showing location of wells penetrating the
thrust; 2.5-D gravity profiles and seismic lines are labeled (lines 1–10).
Figure 4. Structure map on top of the Viola Formation showing surface location of wells penetrating the
horizon.
76
late well with the Viola Formation structure map and the
first-order residual map.
The calculated 3-D gravity anomaly map was subtracted
from the first-order residual, creating a second-order residual
map (Figure 8). Several gravity lows in the second-order
residual map appear to reflect surface outcrops of major fault
zones mapped at the Shallow Spiro Thrust horizon. Gravity
anomaly lows L1, L2, and L3 coincide with surface projec-
tions of the C and Choctaw thrust faults. This series of grav-
ity minima, probably caused by fault-zone brecciation,
appears to merge where thrust faults are close together.
Gravity anomaly lows along major thrust faults in the
Ouachita overthrust belt have been reported by Leonhardt
(l983), Lasley (l987), and Ragland (l988). Gravity anomaly
lows L4 and L5 coincide with a lateral ramp mapped at the
Shallow Spiro Thrust horizon (Figure 3), and a large north-
south-trending normal fault mapped at the Viola horizon
(Figure 4). Evidence of a major transverse structure is sup-
ported at the surface between Gaines and Buffalo Creeks
(Suneson, l990). Within this same location, Wilkerson and
Wellman (1993) mapped a tear fault at the Deep Spiro hori-
zon. Interpretation of the 3-D seismic data and well data sug-
gests that this fault within the Viola Formation controls
lateral ramping of shallow and deep thrusts.
Gravity anomaly highs are also evident on the second-
order residual map. Gravity anomaly H1 results from a high-
density lateral contrast along vertically dipping beds south of
the Ti Valley thrust fault. Gravity anomaly H2 corresponds to
the Hartshorne Syncline. Lasley (l987) reported that syn-
clines north of the Choctaw thrust fault appear as large, sub-
circular gravity highs because of higher density rock within
the synclinal core.
Forward gravity models in 2.5-D (Lai, l982) were gener-
ated over the study area. Each model coincides with a seis-
mic line. These lines are labeled on the Shallow Spiro Thrust
and Viola Formation structure maps. Densities for gravity
models were obtained from well-log bulk densities. Densities
for strata from 91 to approximately 914 m deep ranged
between 2.51 and 2.57 gm/cc and were chosen based on
best fit of the calculated gravity to the observed profile. The
strata are composed of shale, with minor amounts of sand-
stone from the Atoka Formation. Densities within these
strata gradually increase with depth to 2.65 gm/cc. The
04 Plays-Field Studies 4/16/07 4:34 PM Page 77
Atoka Formation is less dense within the Shallow Spiro
Thrust and has a bulk density of 2.64 gm/cc. The Caney and
Woodford Formations have a bulk density of 2.60 gm/cc and
the Viola and Arbuckle Formations have bulk densities rang-
ing between 2.81 and 2.83 gm/cc. Construction of the 2.5-
D gravity models started with a lateral contrast of +.18
gm/cc between the Atoka and Viola-Arbuckle Formations.
The calculated 2-D anomaly was compared with the calcu-
lated 3-D gravity anomaly. Adjustments were made as needed
in the strike dimension to tie with the 3-D gravity anomaly.
Models in 2.5-D in the dip direction (lines 1-8) were tied to
models in the strike direction (lines 9-10) before attempting
to model the shallower thrusts.
At the northwest end of the gravity profile (Figure 9a), a
Viola structural high (Figure 9b) produces the effect of a long-
wavelength gravity high with a superimposed high related to
the Hartshorne Syncline. Small-wavelength gravity lows
related to thrust faulting are located north of the trace of the
Choctaw and C thrust faults.
At the northwest end of the gravity profile (Figure 10a), a
Viola structural low (Figure 10b) produces the effect of a
long-wavelength gravity low, with a superimposed low
Figure 5. Elevation map showing location of gravity stations.
Figure 6. Complete Bouguer gravity first-order residual map showing 2.5-D gravity Figure 7. Calculated 3-D gravity anomaly map from the Viola structure model show-
and seismic profiles (lines 3 and 7). ing 2.5-D gravity and seismic profiles (lines 3 and 7).
Geologic Applications of Gravity and Magnetics: Case Histories
related to the proximity of the Choctaw, C, and B thrust faults. In the cen-
ter of the profile, thrusting of the Viola Formation produces a moderate
gravity high with superimposed highs related to shallow thrusting of rocks
between the B and Pine Mountain thrust faults. A slight gravity low in the
center of this high is related inversely to topography. The Bouguer cor-
rection was overcomputed by +.05 mGal for this small area. The small-
wavelength gravity low at the southeastern end of the profile is related to
a 15-m thick lens of alluvium deposited along Elm Creek.
Conclusions
In this study, structures that could not be imaged with a 3-D seismic data
set were identified by 3-D gravity modeling. Densities derived from well
data and structure maps constructed from well and seismic data provided
input to our 2.5-D and 3-D gravity modeling system. Modeled output pro-
vided a solution which honored all observations. This work shows that
integration of 3-D gravity modeling, well logs, and 3-D seismic data can
result in a better, more comprehensive interpretation.
References
Arbenz, J. K., 1989, Ouachita thrust belt and Arkoma basin, in Hatcher, R. D. Jr., and
Viele, G. W., Eds., The Appalachian-Ouachita orogen in the United States: Geol.
Soc. Amer., The Geology of North America, F-2, 621–634.
Figure 8. Complete Bouguer gravity second-order residual showing surface trace of
major faults and 2.5-D gravity and seismic profiles (lines 3 and 7).
77
04 Plays-Field Studies 4/16/07 4:34 PM Page 78

Plays/Field Studies
78

Figure 9a. 2.5-D model for gravity profile (line 3).

Figure 9b. Interpreted time-migrated seismic dip section. Figure 10b. Interpreted time-migrated seismic dip section.

Lai, S. F., Peoples, W., and Aiken, C. L. V., 1982, Generalized Tilford, M. J., l990, Geological review of the Ouachita Mountains thrust belt play,
linear inversion of 2.5-dimensional gravity and magnetic western Arkoma basin, Oklahoma, in Suneson, N. H., Campbell, J. A., and
anomalies, Expanded Abstracts, 52nd Ann. Internat. Mtg., Tilford, M. J., Eds., Geology and resources of the frontal belt of the western
Soc. Expl. Geophys., 352–354. Ouachita Mountains, Oklahoma: Oklahoma Geological Survey Special Publication
90-1.
Lasley, B. A., l987, A detailed study of the thrust fault zones and
structural relationship of the Penitentiary and Carbon thrust Valderrama, M. H., Nielsen, K. C., McMechan, G. A., and Hunter, H., l994, Three-
faults, Krebs Quadrangle, Oklahoma: Unpub. master’s the- dimensional seismic interpretation of the triangle zone of the frontal Ouachita
sis, Univ. Texas at Dallas. Mountains and Arkoma Basin, Pittsburg County, Oklahoma, in Suneson, N.H.,
Hemish, L. A., Eds., Geology and resources of the eastern Ouachita Mountains
Leonhardt, F. H., 1983, Deformational history of Black Knob
frontal belt and southeastern Arkoma basin, Oklahoma: Oklahoma Geological
Ridge, southeastern Oklahoma: Unpub. master’s thesis,
Survey Guidebook 29, 225–241.
Univ. Texas at Dallas.
Wilkerson, M. S., and Wellman, P. C., 1993, Three-dimensional geometry and kine-
Ragland, B. C., l988, Gravity investigation of a portion of the
matics of the Gale-Buckeye thrust system, Ouachita fold and thrust belt, Latimer
Ouachita central zone, southeastern Oklahoma and western
and Pittsburg counties, Oklahoma: AAPG Bull., 77, 1082–1100.
Arkansas: Unpub. master’s thesis, Univ. Texas at Dallas.
Suneson, N. H., Campbell, J. A., and Tilford, M. J., 1990,
Geology and resources of the frontal belt of the western
Ouachita Mountains, Oklahoma: Oklahoma Geological
Figure 10a. 2.5-D model for gravity profile (line 7). Survey Special Publication 90-1.
04 Plays-Field Studies 4/16/07 4:34 PM Page 79
Gravity Resolution
E. K. Biegert
Shell Exploration and Technology Company,
Houston, Texas USA
Estimation of the accuracy and precision
associated with a gravity survey is not an
easy task. We must estimate not only the
accuracy and precision associated with
each observation of the Earth’s gravity
field, but also, because most modern sur-
veys are acquired from moving platforms
(or individual land stations are leveled as
interconnected networks), we must esti-
mate the spatial resolution of the survey.
In one of the first attempts to understand
this problem, Alan Herring of EDCON
prepared the following graph, summariz-
ing the estimated range of accuracies and
spatial resolution of gravity acquisition
systems. Individual estimates taken from
specific surveys agree fairly well with this
envelope, but vary depending on actual
survey conditions. This graph plots sur-
vey accuracy as a function of shortest
observable wavelength for various types
of gravity acquisition systems.
Perhaps the best way to estimate survey
accuracy and spatial resolution is to
record “repeat lines,” that is, to make
multiple observations of the gravity field
over the same traverse. Fourier analysis
of the repeat lines provides a good esti-
mate of the spatial reproducibility of the
survey and allows characterization of the
noise statistics.
More commonly, gravity surveys are
acquired as a series of intersecting pro-
files. Harmonization of the line intersec-
tions required to “level” the survey and
the resultant “mis-tie” histograms often
are quoted as indicative of the resolution
of the survey, but they provide little
information about the spatial resolution
or its variability. Interpretation of mis-tie Interpretation of Rift-stage Faulting in the
analysis data is complicated further by
lack of consistency in its application West Siberian Basin from Magnetic Data
between surveys and among contractors.
Richard I. Gibson
Estimates of survey accuracy also are Gibson Consulting
made sometimes by inspecting gravity
Golden, Colorado, USA
maps and profiles to determine the mini-
mum detectable anomaly. This subjective
technique provides spatial information EDITOR’S NOTE: This paper provides an example of both lithologic
and is intuitively appealing, but it and structural interpretation of magnetic data. The anomalies
depends heavily on the skill and experi- described are fundamentally lithologic in nature (basalts), but
ence of the interpreter. their tectonic setting (a rift) permits direct interpretation of
normal faults that are important in hydrocarbon trapping.
Abstract
Regional aeromagnetic data in the northern part of the West Siberian
Basin of Russia show elongate highs with amplitudes on the order of 200
to 500 nT. These highs are interpreted to represent basalt-filled grabens of
Triassic age. The geometry of gradients along the flanks of the magnetic
highs permits the interpretation of specific faults as well as accommoda-
tion zones, and these features clearly are related to existing gas produc-
tion. This tectonic interpretation can be used to define details within
known productive areas as well as to point out analogous locations for
exploration.
Introduction
The data set used consists of a published compilation (1:2,500,000 scale)
of aeromagnetic data by Makarova (1974). The original data probably
were acquired at 2-km line spacing, about 400 m aboveground. The data
were digitized for Gibson Consulting, and this interpretation is part of a
comprehensive review of the data set for the entire former Soviet Union.
The analysis was revised somewhat for this paper.
Geologic Setting and Interpretation
The West Siberian Basin is underlain by probable Paleozoic (Hercynian,
Late Paleozoic, or Baikalian, Early Paleozoic and Late Precambrian) crys-
talline basement. The region was accreted to the East European and
Kazakh Cratons during the Permo-Carboniferous assembly of Pangea,
recorded in this part of the world by the Ural Orogenic Belt. Post-tectonic
extension, associated with voluminous extrusion of trap basalts in the
79
Geologic Applications of Gravity and Magnetics: Case Histories
04 Plays-Field Studies 4/16/07 4:34 PM Page 80

Plays/Field Studies
80

Figure 1. Location of Figures 4, 5, and 6.

East Siberian Craton, began in the Triassic and continued into the
Jurassic. In the West Siberian Basin, the extension was marked by graben
formation on a scale comparable to the Precambrian Mid-Continent Rift
System of North America (Figure 2). As in that rift system, the grabens
were filled extensively with basaltic material. Subsequent postrift deposi-
tion (Jurassic and especially Cretaceous fluviatile) and marginal marine
sediments (Figure 3) buried these grabens.
Post-Neocomian rejuvenation of the graben faults created broad
arches in the Cretaceous sediments; maturation of Jurassic source rocks
resulted in migration of hydrocarbons into these huge traps. The coinci-
dence of magnetic lows with horsts and productive fields (Figures 5 and
6) is evidence for the rejuvenation of preexisting faults, if the assumption
that the magnetic highs result from basalt-filled grabens is valid.
Figure 4. Magnetic map of part of the north- Figure 5. Interpretation showing normal faults
The fault interpretation (Figure 5) shows numerous specific inferred of Triassic extensional origin. They were reac-
ern West Siberian Basin. Data from Makarova
normal faults. These are presumed to have Triassic-Jurassic heritage and (1974). Contour interval—100 nT. tivated during post-Neocomian time. Thin
green lines, generally northeast-southwest,
are inferred accommodation zones.

to have been reactivated at least slightly in post-Neocomian time. Accom-

modation zones (thin green lines), probably of similar age, clearly segment
the inferred grabens and horsts. In this area, accommodation zones appear
to define limits on structures and on production. In some locations around
the world, such zones may be preferred areas of fracturing and develop-
ment of small-scale structures, which may be prospective.
The interpretation of the magnetic highs as basalts is supported by
their coincidence with gravity highs (Shaposhnikov and Samoletov,
1986), a further analogy with the Mid-Continent Rift System of North
Figure 3. Generalized stratigraphic column (after Clarke, et al., 1977).
America.

Figure 2. Mid-Continent Rift of North America and West Siberian Rift at same scale.
04 Plays-Field Studies 4/16/07 4:34 PM Page 81

References Mag Susceptibility Contrast

Clarke, J. W., Girard, O. S. Jr., Peterson, J., and Rachlin, J., 1977, Petroleum geology versus Structure
of the West Siberian Basin and detailed description of the Samotlor oil field: USGS
Open-File Report 77–871. Magnetic methods are known to be
Makarova, Z. A., 1974 (published 1977), Map of the anomalous magnetic field of the “nonunique” in their interpretation, but
USSR: VSEGEI, 1:2,500,000 scale. certain generalities are usually true. For
Shaposhnikov, K. K., and Samoletov, M. V., 1986, Use of combined gravimetric and example, it is fair to say that most high-
aeromagnetic data in exploration for oil and gas pools in West Siberia: Petroleum amplitude magnetic anomalies indicate
Geology, 23, 362–368 (English summary). large contrasts in susceptibility, which in
turn indicate large variations in intrabase-
ment rock type. “High amplitude” means
anomalies on the order of 150 or 200 nT
(gamma) or more. The figure here shows
the difference between a simple suscepti-
bility contrast (about a 300-nT anomaly)
and a large structure (1219-m, or 4000-
ft, fault, producing a 25-nT anomaly). If
such features are superimposed on each
other, it can be very difficult to separate
the effects of structures and susceptibility
contrasts. Inasmuch as major intrabase-
ment contacts are often weak zones that
may be rejuvenated by later tectonism,
this separation is important, even though
it may be difficult.

R. I. Gibson

Figure 6. Magnetic map showing gas fields.

Note coincidence of fields with magnetic
lows. Urengoy Field is the largest gas field in
the world.
Figure 7. This magnetic model shows how the presence of magnetic basalt (V pattern; susceptibility
contrast = +0.01 to +0.018) in grabens in basement (random-dash pattern; susceptibility contrast =
+0.005) can produce the 300- to 400-nT anomalies observed in the Urengoy area. Sediments have
zero susceptibility in this model.

81
Geologic Applications of Gravity and Magnetics: Case Histories
04 Plays-Field Studies 4/16/07 4:34 PM Page 82

Plays/Field Studies
82
Case History: Reddell Oil Field,
Evangeline Parish, Louisiana
C. T. Austin
Austin Exploration, Inc., Houston, Texas, USA

EDITOR’S NOTE: A classic gravity case history, this study high-

lights the need to incorporate geologic thinking into all
phases of gravity data processing, interpretation, and
modeling.

The purpose of this paper is to give the gravity meter credit as the main
geophysical method in the discovery of the Reddell Oil Field in Evangeline
Parish, Louisiana.
As in most cases, no one procedure or person is responsible for a dis-
covery, but a culmination of several disciplines working in unison. There
always must be the consideration of geology, including all known facts rel-
ative to hydrocarbon entrapment.
In late 1971, LL&E (INEXCO) asked me to model the salt structure at
the Reddell Salt Dome from gravity data. The gravity data were obtained
from Conoco, Inc., and tribute is given to Conoco for allowing this data
to be used in this paper. I want to thank LL&E for allowing me to publish
this paper and for assisting in the preparation of the maps. I also appreci-
ate Tobin Research Inc. for letting me use its base map of the area.
The object of the study was to define the salt-dome structure so the
truncation of three (3) prospective sands in the Wilcox, Lower Eocene,
could be determined.
Figure 1. Figure 2.
The Bouguer gravity map (Figure 1) shows the minimum anomaly
from the Reddell Dome combined with the anomaly from the Pine Prairie Once a regional gravity (Figure 2) was determined, it was subtracted 2.2 grams per cubic centimeter, was plotted at –305 m (–1000 ft). The
Salt Dome. There is a distinct difference in the character of the gravity from the Bouguer gravity map to produce the residual gravity map (Figure density contrast between salt and sediments may be observed from the
over the two domes because the Reddell Salt Dome, at the southern part 3). Minor adjustments were necessary in the residual gravity map to plot, and this was input for our 3-D modeling program.
of the anomaly, is 3658 m (12 000 ft) deep, and Pine Prairie Dome, at the remove anomalies from shallow sources and leave the anomalies caused Another input for the modeling program was a reference surface, in
northern part of the anomaly, is near ground surface. by the more deep-seated salt structure. This map shows only the mini- this case, the top of Mother Salt, or Louann, Lower Jurassic. We used a
The regional gravity and its removal from the Bouguer gravity are mum gravity associated with the salt below the density crossover depth, reference surface of –6096 m (–20 000 ft). It was also necessary to know
always an interpretive procedure. In more general cases, this can be between salt and sediments, because our interest was in the deep salt all the well control where salt and sediments were drilled on the tops and
approximated by using low-frequency filters, polynomial fitting, etc. In structure. The residual gravity was input for our 3-D computer modeling flanks of the salt domes to be modeled.
this case, the gravity data were fitted to a second-order polynomial. This program. The residual gravity input for the program was in milligals from a
surface then was checked by comparing the Bouguer and regional gravity We used all available gamma-gamma density logs to derive the sedi- square grid pattern appropriate for the size of anomalies mapped and the
along several profiles. It is necessary that the regional gravity be very pre- ment densities shown on this plot on the top-of-salt map (Figure 4). The cross- gravity control available. The grid system for the residual input estab-
cise in making detailed model studies. over density depth where the sediments have the same density as salt, or lished the grid pattern used by the reference base. This grid became a sys-
04 Plays-Field Studies 4/16/07 4:34 PM Page 83
Figure 3.
tem of prisms contained in a rectangular area. All prisms were identical
except for the heights, which dictated the surface of the model. A gravity
value was associated with the center of each prism. These values formed
the data grid used in an iterative procedure to compute a 3-D salt model.
The iterative process used in computing the 3-D model began by
approximating a model using the slab formula and the input gravity data.
The gravity was computed for this model and was compared, grid inter-
section by grid intersection, with the input gravity data. Prisms at each
intersection were modified as dictated by the difference between the
input gravity and the computed gravity. A single iteration of the program
involved calculating the gravity at each grid intersection and modifying
the height of each prism at each intersection. The root mean square (rms)
error between the calculated gravity and the input gravity was calculated
Figure 4.
at the end of each iteration. This rms error tends to become smaller and
smaller after each iteration. If this error was to reach zero, then the 3-D
model calculated by the program produces a gravity response identical to
the input gravity. In the real case, this rms value never will go to zero.
However, the magnitude of this error does indicate how good the model
was, and ten iterations were sufficient to establish a good model.
A useful capability of the modeling program used the well control,
and we input salt control at both domes.
After we had computed a reasonable shape for the two salt domes, we
added the well control as input for the modeling program. Our main salt-
control point located near the top of Reddell Dome was –3753 m
(–12 314 ft) from a drill site at NE, NW, SE, Section 21, T4S, R1W. The
salt-control point for Pine Prairie Dome was considered to be at sea level
Geologic Applications of Gravity and Magnetics: Case Histories
Figure 5.
at the center of the dome in the southeastern part of T3S, R1W.
Adjustments are made to the density function if the model does not agree
with the known well information. Iterations with continued well check-
ing and density function adjustments were made until the model reached
a suitable fit with the well information.
When we had completed the two above computer programs, includ-
ing the first with structure as a variable and the second with the density
function as a variable, we had modeled the Reddell and Pine Prairie Salt
Domes to our satisfaction. The model salt structure then was used with
geologic information to drill a producing well on the Reddell Dome.
Other producing wells were drilled at Reddell Field and, later, Pine Prairie
Field was drilled using the same salt model from gravity data. Reddell
Field and Pine Prairie Field (Figure 5) show the producing area. The
83
04 Plays-Field Studies 4/16/07 4:34 PM Page 84
Plays/Field Studies
cumulative production as of September 27, 1993, at Reddell Field was Classic Salt Dome
6.6 million bbls condensate and 172.2 billion cubic feet of gas; at Pine
Prairie Field, it was 5.8 million bbls oil and 39.1 billion cubic feet of gas. The classic salt dome anomaly shows a
The gravity study at the Reddell and Pine Prairie Fields was definitely sharp, intense gravity low, reflecting the
typically large density contrast between
successful. In situations such as this, it usually is recommended that seis-
salt and surrounding sedimentary rocks.
mic control be acquired before drilling. When seismic data are available,
the gravity data should be used to confirm that all data fit the interpreta- In the Gulf Coast, shallow cap rock may
produce an even sharper gravity high at
tion. When one set of geophysical data, either seismic, gravity, magnetic,
or near the center of the low, as seen in
or other, does not fit the interpretation, it is usually the interpretation that
the illustration on the left. This is an
is incorrect. indication of high-density materials such
as anhydrite or limestone. Many exam-
ples of such anomalies in real data can
be found in the older literature.
A similar effect occurs when the salt is
shallower than the density crossover
depth, i.e., when the salt is more dense
than the sediments. These alternatives
are resolvable when integrated with the
top of salt from seismic, density control
from nearby wells, and/or densities cal-
culated from sedimentary velocities.
Cap-rock anomalies are nothing to be
ignored. Unocal has used carefully ana-
lyzed gravity to interpret hydrocarbon-
filled cap-rock reservoirs.
Reference
Johnson, E. B., 1995, Caprock oil: Presented at
the Gravity Magnetics Workshop, Soc. Expl.
Geophys.
R. I. Gibson
Nash Dome
Nash Dome was the first oil field to be
found through the use of any geophysical
technique. Its discovery in 1924, in Fort
Bend and Brazoria Counties, Texas,
demonstrated the capability of torsion-
balance surveying to define subsurface
features. Barton ((1949) reports that con-
siderable luck was involved in the dis-
covery and that similar salt domes would
not be found by the reconnaissance
methods in use in the late 1930s. This
dome is also a classic example of the
gravity signature of a salt dome with a
cap rock. A fairly broad low contains a
relatively sharp high within it. The broad
low is the anomaly caused by low-den-
sity salt, and the sharp high is caused by
dense cap rock (e.g., limestone, anhy-
drite) at shallow depth. A generalized
profile of this classic salt-dome anomaly
is given here.
Reference
Barton, D.C., 1949, Gravity anomalies of Nash
and Damon Mounds, Fort Bend and Brazoria
Counties, Texas, in Geophysical Case
Histories: Soc. Exp. Geophys., 1, 35–42.
R. I. Gibson
General Philosophy 3
edited by David A. Chapin
LaCoste & Romberg,
Austin, Texas, USA
Concerning Modeling
The trend toward modeling is a healthy
change from the days when “prospects”
were located solely on the basis of clo-
sures on residual maps of mysterious and
sometimes secret origin. Although resid-
ual maps are valuable for helping us
notice unusual relationships, the business
of exploration has become less tolerant
of anything that raises more questions
than it answers. A flurry of residual maps
is a pretty good example of such a thing.
Once an appealing target is noticed in a
residual, its geologic origin ought to be
84
questioned and justified in relation to the
basic, measured data.
Alan T. Herring
EDCON
Denver, Colorado, USA
To simplify modeling, use free-air gravity.
Why does everyone regard Bouguer grav-
ity as the prime form of preserving and
presenting gravity data? Because it pro-
vides profiles or grids which do not show
much elevation effect and because to do
anything else risks killing a sacred cow.
But modeling a Bouguer profile or grid is
fraught with problems because you have
to change densities at sea level (onshore)
and may need to change densities at
seabed (offshore).
So it’s much simpler to model using free-
air gravity and effectively incorporate the
Bouguer and topography corrections in
the model.
Be careful with models; they can be mis-
leading.
I believe modeling is oversold. I’m frankly
uncomfortable most of the time I’m doing
modeling, for a variety of reasons.
1) The old one about nonuniqueness. I
continue to see reputable practition-
ers showing models which are pretty
complex, fitting a bland or complex
profile, without much (or any) dis-
cussion of the sensitivities. Once you
have a fit which is geologically credi-
ble, you have done only the first part
of the job. The next part is to investi-
gate just how much you can change
it while retaining both the fit and the
geologic credibility. The problem is
04 Plays-Field Studies 4/16/07 4:34 PM Page 85
that it is time-consuming and, all too
often, somewhat frightening. So very
few interpreters do this.
2) What’s going on in the basement? In
the oil patch, we tend to assume a
homogeneous basement (containing
maybe the occasional intrabasement
body). When one studies basement
in great detail, you find that it isn’t
homogeneous as to density, suscepti-
bility, remanence, or anything else.
It’s often a real mess, especially mag-
netically. I suppose that means that
magnetic models using a homoge-
neous basement are potentially fairly
misleading, but I see them all the
time and frequently am required by
clients to produce them.
The only saving grace is that, as recently
shown, magnetic phenomena are fractal.
So they are self-similar and you can get
away with ignoring fine basement detail
when it is covered by thick sediment.
But if they’re fractal, they will show vari-
ability at all scales and you can’t ever
assume homogeneous basement.
All this is the sort of reason I like to use
natural enhancements and things such as
Euler deconvolution. Euler deconvolution
is particularly appealing because it self-
adjusts for magnetization intensity or den-
sity. It is criticized for this, but I consider
this a major advantage. So I prefer to
interpret the data using the enhancements
I listed below under General Philisophy 4
(with shaded imaging, of course), plus
things like Euler deconvolution, and then
check them with modeling.
Alan B. Reid
Geophysical Exploration Technology (GETECH)
Leeds, United Kingdom
One guiding principle that I find useful is
the following:
If a modeling project shows only one
model, it has not been finished yet.
In short, I strongly believe that all model-
ing projects should strive for a good fit
between the modeled and the observed
data. Unfortunately, 90% of the projects I
see stop there. The geophysicist receiving
a project should ask for alternative mod-
els which test the critical elements of the
model and demonstrate how wide a
range of variation can be tolerated.
Sometimes, several drastically different
models can fit the same data equally well.
I find that often the budget constraint is
what prevents serious investigation of
alternative models. In my view, one prac-
tices false economy if the modeling pro-
ject stops short.
As one anecdote that demonstrates the
point, I once was asked to model the
gravity over a proposed well on the edge
of a basin with a thick salt section, not
far removed from outcropping Pre-
cambrian rocks. There was a major
unconformity at 1200-m depth, and
below that the seismic reflections were
poorly defined. The well already had
been planned for a total depth of 4500
m, but someone had misgivings and
asked for a “second opinion” from the
nonseismic quarter. When I presented
the project, my opening comment was,
“I think you have a 30% chance of hit-
ting granite below 1200 m.” After the
manager had recovered from his initial
shock, his next question was, “Why only
30%?” I replied, “Because I am sure that
the average density below the unconfor-
mity is 2.65–2.70 g/cc, and I can come
up with three totally different models to
Geologic Applications of Gravity and Magnetics: Case Histories
give me that average density. Granite is
just one of those three.”
Several months later, after the well had
been abandoned at less than half the
prognosed depth, the chief geophysicist
asked me about the densities below the
unconformity. “Oh, they found shales of
density 2.67 all the way down,” I said.
“Heck,” he answered, “that’s not shale,
that’s slate!”
John Peirce
GEDCO
Calgary, Alberta, Canada
I completely agree with the statement
that if a modeling project shows only one
model, it has not yet been completed.
One of the important functions of model-
ing is to try to determine what new
information the potential-field data add
to a geologic (or engineering) problem.
By perturbing a “final” model con-
strained by available geologic, geophysi-
cal, and geochemical data, minimum and
maximum thicknesses of bodies often
can be determined. Rather than offering
“the” solution, the modeler can demon-
strate the contribution of the potential-
field data to a particular problem. This is
often the most difficult part of modeling.
Carol Finn
U. S. Geological Survey
Denver, Colorado, USA
Magnetic Anomalies Associated with the
North and South Morecambe Fields, U. K.
Richard Morgan
World Geoscience UK Ltd., Guildford, Surrey, U. K.
EDITOR’S NOTE: This paper breaks new ground in terms of the
information high-resolution aeromagnetic surveys offer on
the sedimentary section. The provocative interpretation corre-
lates real reservoir characteristics with subtle but significant
intrasedimentary magnetic effects.
Abstract
A close line-spaced aeromagnetic data set was acquired over the East Irish
Sea in 1993 by World Geoscience Corp. Ltd. Although the area was al-
ready a known hydrocarbon province containing one of the United
Kingdom’s largest gas accumulations (collectively, Morecambe North
and Morecambe South Fields), the aeromagnetic data reveal much about
the geology of the area which previously was unrecognized.
The continuation of faults known from seismic data or from surface
maps of the adjacent onshore areas is represented clearly by the magnetic
data, allowing the relationships between faults of different trends to be
elucidated. The geometries of major faults imaged in plan-view by their
magnetic expression reveal compartmentalization by secondary cross-cut-
ting fault suites which, although not clearly resolved by 2-D seismic data,
are relevant to sediment deposition and trap formation.
An unpredicted aspect of these data is the association of weak mag-
netic lows with known hydrocarbon accumulations. This phenomena is
expressed best over the Morecambe Fields themselves, where analysis of
the shape of the anomalies in relation to the structure of the fields sug-
gests their source to be associated directly with the structure of the clo-
sures rather than as alteration plumes above them. The cause of these
anomalies is considered to be secondary mineralization, believed to relate
to the complex diagenetic and multiphase charging history of these fields.
The detailed comparison of the seismically defined structure of the
fields with a corresponding magnetic structure would not be possible
without a close line-spaced, aeromagnetic survey configuration to define
clearly the margins of these intrasedimentary-sourced, nonlinear mag-
netic anomalies.
85
04 Plays-Field Studies 4/16/07 4:34 PM Page 86

Plays/Field Studies
86
Introduction resulted in geologic complexity—the preceding basin structure has been 4˚W 3˚W
reactivated extensively as transfer zones during Permo-Triassic rifting—
The East Irish Sea Basin forms a major element in a chain of irregularly
and economic complexity—the hydrocarbon source rocks are contained
shaped Permo-Triassic basins extending down the western coast of 6 7 8 9 10 6
within the Carboniferous basin structure (Hardman et al., 1993; 11 12 13 14 15 11
EXL 121

Scotland and northern England, then continuing onshore along the MUS

Trueblood et al., 1995), while the reservoir sequences are part of the 112
English/Welsh border (McClean, 1978). CUMBRIA
overlying Triassic basin (Bushell, 1986; Jackson and Mulholland, 1993). 12

Permian and Triassic subsidence in the East Irish Sea area was cen- 16 16 17

Moreover, the entire basin was uplifted during the Early Tertiary.
tered on faults trending northwest-southeast to north-south, accompanied
Estimates of the degree of unroofing vary from <1 km (Colter and Barr, 18
by some reactivation of northeast-southwest-trending Caledonoid faults 21 22 23
1975) to >3 km (Lewis et al., 1992). Also, it should be noted that the ISLE
(Figure 1). This pattern of subsidence contrasts with the development of OF
113
basin is presently below sea level, indicating a mild resumption in subsi- MAN
the underlying Carboniferous basin system, which was dominated by 28 29
dence during the Plio-Pleistocene. 26
CLY AMO EXL
38
faults trending east-west to east-northeast–west-southwest (Fraser and E
ULT

The complex basin history has resulted in a correspondingly compli- BG

27 S
S
EXL 127

Gawthorpe, 1990). These orthogonally superimposed basins have 30 O

PND
54˚N
cated petroleum system involving a number of trap-forming, hydrocarbon- 54˚N
4 5 1 2
BG
3 ORX
BG
45
RAN CON

charging, and breaching events (Bushell, 1986; Stuart and Cowan, 1991). MORECAMBE ORX

Tertiary igneous intrusions The Morecambe Field (Figure 2) was discovered in 1978. Although it 6 7 8 9 10 6 7 9
BG
10
PRE
EXL 135
Triassic Mercia Mudstone Group is represented by a clearly defined Triassic high (Figure 3), a complex bur- 110 HYD 8
BG EXL
EON

45
Triassic Sherwood Sandstone Group and Permian ial history is illustrated through an intricate diagenetic history (Bushell, 15 11 ENT
KLT HAMILTON
NORTH
13 14 15
CLY

11 12 13 14 BG HAM
Carboniferous 1986; Stuart and Cowan, 1991; Knipe et al., 1993). CON

Lower Palaeozoic and older 109 12

BG
HAMILTON

Twelve years followed before the discovery of the first commercial oil DOUGLAS HAM
FORMBY EXL
173
FIN
Gas field 19 20 16 17 18 19
accumulation in the basin, the Douglas Field. Within the same U. K. 16 17 18

license block, 110/13, two gas fields are also under production, Hamilton EXL
204
EVR
POINT OF

CUMBRIA and Hamilton North Fields (Figure 2). The proximity of the Douglas Field, 21 22 20 21 22 23 AYR
EXL 203
EVR

a low gas-oil-ratio accumulation (Yaliz, 1998), to gas fields within the ANGLESEY

same reservoir sequence and at similar depths testifies to the complexity 26

23
27 28
EXL 201
EVR EXL 186
ISLE TTN

OF of the petroleum system in this region.

MAN
4˚W 3˚W
Petroleum Geology
The Morecambe Field (Figure 3) originally was delineated as a single, 0 60 km Aeromagnetic survey

shallow-closure field (depth to crest is less than 730 m, water depth over
the field is approximately 60 m). Further drilling and seismic data acqui-
sition determined that the structure was divided by a narrow graben
trending east-northeast–west-southwest (Figure 4), separating Morecambe
North (gas-water-contact [gwc] 1204 m, or 3950 ft), from Morecambe
South, gwc 1143 m (3750 ft).
The field is a Triassic horst-block/tilted-block closure (Figure 4),
sealed by overlying mudstones and evaporites of the latest Scythian-
Anisian Mercia Mudstone Group (Figure 5). The reservoir lies within the
Scythian Ormskirk Sandstone Formation (Jackson and Mulholland,
ANGLESEY 1993), which is more than 240 m in thickness in the area of the field. The
maximum vertical gas column encountered exceeds 190 m (Stuart and
Cowan, 1991).
NORTH WALES A feature of the Morecambe Fields reservoir is an apparent paleo-gwc
Figure 1. The structure of the East Irish Sea Basin (after Arter and Fagin, 1993). which occurs in both structures (Bushell, 1986; Woodward and Curtis,
Figure 2. Oil and gas fields in the East Irish Sea area.
1987) and is represented by an absence of platy illite in the upper sections
of the reservoir. Both structures subsequently have been charged with fur-
ther gas, bringing the present gwc into the illite-affected zone. The platy
illite has a detrimental effect on permeability, although its effect on poros-
ity is minor (Stuart and Cowan, 1991).
The sequence of hydrocarbon charging and diagenesis is different in
the two structures (Stuart, 1993). Morecambe North displays a history of
sequential filling punctuated by the precipitation of platy illite (Figure 6).
Morecambe South appears to have been breached after the precipitation
of the illite (Stuart and Cowan, 1991), allowing a phase of quartz cemen-
tation within the reservoir zone unaffected by illite, in addition to that
observed in Morecambe North. Tilting and open folding of the top illite
04 Plays-Field Studies 4/16/07 4:34 PM Page 87

110/2a A 00 110/3b
Halite
40
3 36' W

Mercia Mudstone Group

BGE&P ORYX
gw
c Mudstone and shale

0
00

425
40 n
io
ect Sandstone
fs 4
e o gure

3750
n
35
00 Li F i Siltstone
110/3a

300
BGE&P Limestone

0
B B'

3250

55
00
0
300

Triassic
30
Top Ormskirk Sandstone

Sherwood Sandstone Group

00
00
40

3250
Top St. Bees Sandstone
35

C C'
00

3000

gw
c
Top Silicified Unit
350
32
50

0
300

300
0
0
3250

27
50 3250
325
0
32
50

Upper
Permian
D D'
53 50' N
110/2b
BGE&P Top St. Bees Evaporites
110/7a 110/8a
2700

BGE&P BGE&P Lwr

Hercynian Unconformity
50
West 0 - 150m
27 0
0
30
0
325
0
350

Carboniferous

Namurian
Contours in ft
600 - 1000m
A'

0 km 5

0 mile
Figure 3. Depth-to-top Ormskirk Sandstone Formation over the Morecambe Fields
(modified after Stuart, 1993, and Stuart and Cowan, 1991).
surface in both fields demonstrate continued development of the closures
between phases of hydrocarbon charging (Figure 6).
Magnetic Imaging of the Morecambe Fields
The survey
Aeromagnetic surveying of this area was conducted with a flight-line spac-
ing of 400 m and a tie-line spacing of 1200 m at a mean sea-level clear-
ance of 80 m. Flight lines are orientated 020–200 with orthogonal tie
lines. Navigation was by differential GPS satellite positioning and
measurement by a split-beam, cesium-vapour magnetometer with a reso-
3
3 36' W
Figure 4. Seismic profile and structural cross-section through South Morecambe Field
(after Knipe et al., 1993).
lution of 0.001 nanotesla (nT) and a cycle rate of 0.1 seconds, resulting
in a sample interval of approximately 7 m. The total noise of the acquisi-
tion system was retained beneath 0.1 nT.
Total magnetic intensity
A pseudocolored image of the total magnetic intensity, reduced to pole
(RTP) over the area of the fields, reveals little expression of the field struc-
tures (Figure 7). Merging it with a gray-scale image of the first vertical deriv-
ative (1vd) of total magnetic intensity renders short-wavelength, linear
magnetic structures visible. These are coincident with the position of faults
at Top Sherwood Sandstone level, which bound the fields to the west.
Top Dinantian
Din
Figure 5. East Irish Sea generalized stratigraphic column.
A basaltic sill of Paleogene age is clearly visible as an irregularly shaped
low in the northeast corner of the image. The sill lies at less than 1 km
depth within the Mercia Mudstone Group (Arter and Fagin, 1993). This
suite of mafic intrusives shows far greater magnetic susceptibilties than the
host sediments (Mussett et al., 1976). High susceptibility is coupled with
a high Q ratio (remanent/induced magnetization) where the remanent
component of magnetization is reversed or polarized southward. Because
there has been relatively little change in latitude or deformation of these
intrusives since emplacement, they appear as magnetic lows.

87
Geologic Applications of Gravity and Magnetics: Case Histories
04 Plays-Field Studies 4/16/07 4:34 PM Page 88

Plays/Field Studies
88
(Cowan et al., 1993), derived from 2-D seis-
Meters
mic data (Figure 4). The dividing structure
0 1000 2000
0
between the fields is imaged as a zone of fault-
Meters ing involving elements trending east-west and
500 Scale
vs : hs = 4 : 1 northwest-southeast. This zone has offset the
main westerly bounding fault with a left-lat-
1000
eral sense.
Morecambe North Morecambe South
North A A' South
The image also shows crosscutting mag-
netic structures in positions and orientations
500
Mercia Mudstone Group 500
not indicated by seismic mapping (Figure 8).
one
The action of these structures in crosscutting
ndst
Top Ormskirk Sa
1000 1000
and offsetting the seismically imaged fault
Depth
Meters boundaries of the field structures identifies
1500 Sherwood Sandstone Group 1500 them as part of the fault framework of the
fields.
2000 2000 The very close correlation between the
principal fault boundaries of the fields, as
Gas-water contact mapped seismically at top Ormskirk Sandstone
Top platy illite Formation (Figure 8), and the apparent mag-
Gas leg
netic expression of these faults indicate the
source of the fault-related anomalies to be
Figure 6. Simplified cross-section through the Morecambe Fields.
close to the top Ormskirk Sandstone surface.
Coincidence of the locations of fault-related,
Pseudodepth slice (PDS 1) linear magnetic anomalies and the seismically
mapped positions of footwall cutoffs at top Ormskirk Sandstone level is
The pseudodepth slice (PDS) filter operates in a similar way to the spec- observed basinwide.
tral shaping filter described by Cowan and Cowan (1993). In nonmathe- A magnetic susceptibility study of the Permo-Triassic section from on-
matical terms, it can be described as a frequency filter without hard shore boreholes around the East Irish Sea Basin (Cornwell and Raines,
edges, progressively reducing signal amplitude away from a desired fre- 1994) found the Ormskirk Sandstone to exhibit relatively uniform and Figure 7. Color image of total magnetic intensity over the Morecambe Fields with the
quency band width. This means leakage of frequencies occurs outside the consistantly lower susceptibility than the Mercia Mudstone. first vertical derivative as shade. The seismic-structure map of the fields is superim-
selected band width, and the filter cannot be used to isolate selected Consequently, the largest regional contrast in magnetization occurs at the posed, allowing comparison of fault delineation (fault structure of the Morecambe
anomalies to precise depths of origin. However, the advantage of this fil- Fields, after Cowan et al., 1993).
top of the Ormskirk Sandstone Formation. Thus, the most likely source of
ter in geologic interpretation is that the resolution over a desired fre- observable magnetic anomaly in the Permo-Triassic section is where the
quency range is sharpened while retaining a subdued expression of Ormskirk Sandstone is faulted and displaced vertically, in the absence of An Explanation for the Magnetic Anomalies over the
frequencies outside the range. This allows magnetic structures, particu- any overprinting, magnetizable, diagenetic mineral phase.
larly those from intrasedimentary sources, to be seen in context with
Morecambe Fields
The other notable feature in the PDS 1 image is the expression of the
deeper-sourced structures. The essential fractal nature of the data largely area of the fields as magnetic lows. The overlay of the field outlines shows Basement source
is preserved, and the resulting images have an intrinsic geologic appear- the shape of the North Morecambe gas accumulation to be imitated pre- North and South Morecambe are structural closures and, as such, any
ance (Figure 8). cisely by the shape of the magnetic anomaly, while the shape of the South associated magnetic anomaly may be sourced from layers above or below
The fault boundaries of the Morecambe North and South structures Morecambe accumulation differs significantly from the shape of the the reservoir, although retaining the approximate shape of the field out-
are clearly visible in the PDS 1 image (Figure 8). This image reveals the anomaly over its northern section (Figure 8). lines. However, the similarity between the shape of the accumulations
continuity and linkage of faults delineated by the top structure fault map and the associated anomalies suggests the source depth to be reasonably
proximal to the depth of the reservoir as mapped. Moreover, basement
04 Plays-Field Studies 4/16/07 4:34 PM Page 89

110/2a 110/3b
layers. Magnetic minerals occur on both sides of the reaction (principally
BGE&P 3 36' W ORYX magnetite and pyrrhotite), so the development of increased or decreased
395
0
susceptibility as a result of iron-oxide-to-sulphide reduction reactions in
the presence of hydrocarbons is not clear-cut. The identification of
110/3a
microbes which build magnetite in aerobic and anaerobic settings using
BGE&P
hydrocarbons as a nutrient source is becoming increasingly common also
(Farina et al., 1990; Hawthorne and McKenzie, 1993; Lovely et al.,
1987). Therefore, magnetite can be both destroyed and created in the
presence of a hydrocarbon plume.

00
30
On first inspection of the Morecambe Fields, good candidates for
hydrocarbon plume activity appear. The fields are relatively shallow, at a
depth of 730–1200 m, and substantial Early Tertiary uplift means numer-

00
30
ous faults cut from reservoir levels to seabed. However, the pattern of the
field-centred magnetic anomalies does not reflect the positions of faults

37
00
36 0
above the fields acting as conduits controlling magnetization contrasts, as

35 00
00
34 00
0
33 00
32
35
31
00 00
34
might be expected. In addition, the aerial form of near-surface-sourced
00

330
0 anomalies would not be expected to reflect accurately the form of the clo-
00
sures at reservoir level.
32

110/2b
53 50' N BGE&P
110/7a
BGE&P
Gas-water contact
Edge of field-
centered magnetic
anomaly
Contours in ft
0 km 5
0 3
mile
3 36' W
Figure 8. Color image of the high-frequency pseudodepth slice (PSD) over the Figure 9. Depth to source of the magnetic anomaly low over the Morecambe Fields,
Morecambe Fields with directional illumination from the northwest. The seismic- based on the intersection of the area of the anomaly with the top reservoir surface.
depth-structure map is superimposed, allowing comparison of closure geometries
with magnetic structure. The field-centred anomalies are visible as magnetic lows
within the outline of the fields. (Depth-to-top Ormskirk Sandstone Formation over the
Near-surface source
Morecambe Fields modified after Stuart, 1993, and Stuart and Cowan, 1991.)
Some oil accumulations show aerially coincident magnetic anomalies
influence may be discounted because the entire Permo-Triassic section is related to a chimney or plume of diagenetically altered strata above the
detached from the basement above the low-angle Keys Fault, which dips accumulation. The most well known of these is the Cement Oil Field in
to the northeast beneath the structures (Figure 4). It would be fortuitous Oklahoma (Donovan, 1972; Kirkland et al., 1995). The seepage of hydro-
indeed if the magnetic structure of the decoupled basement precisely imi- carbons is known to create a reducing environment, and the resulting
tated the position of the overlying fields. In addition, the magnetic-low reactions involving magnetic minerals are believed to be responsible for
character of these anomalies tends to invalidate the simple model of base- the magnetic anomalies seen. The most common of these reactions, mag-
ment raised beneath the field structures, unless a significant remanent netite or hematite to pyrite or pyrrhotite (Machel, 1995), can involve bac-
magnetization is involved. teria as a mediator, particularly, although not exclusively, in near-surface
3950
00
33
Intrastratal source
110/8a
34
00
BGE&P The close similarity between the shape of the North Morecambe accu-
mulation, as defined by the present gwc, and the margin of the attendent
35
00 magnetic anomaly suggests the source of the anomaly to be related closely
36
00 to the geometry of the field closure itself. Correspondingly, the edge of
37
00 the field-centred anomaly describes the intersection of a horizontal sur-
face with the top seal of the reservoir, coincident with the present gwc
(Figure 8).
The relationship between the South Morecambe Field structure and
its attendant anomaly is evidently more complex. The delineation of the
anomaly margin as an intersection of a surface with the top of the reser-
voir describes a very open fold in the northern third of the field, the axis
of the fold being nearly orthogonal to the axis of the field itself (Figure 9).
This surface dips gently to the southeast across the lower half of the field.
The tightening of the field closure along the axis is illustrated by the defor-
mation of the top illite surface (Figure 10), and the conceptual surface
defined by the magnetic anomaly may be interpreted to describe the later
development of the closure.
The filling-breaching-filling history of the field provides a possible
explanation for a physical surface with this geometry, that being a second,
younger paleo-gwc. Applying this model, the cause of the anomaly is
attributed most plausibly to mineral phase changes across this paleo-gwc.
This may have involved the precipitation, removal, or alteration of ferro-
magnetic minerals below the paleo-gwc.

89
Geologic Applications of Gravity and Magnetics: Case Histories
04 Plays-Field Studies 4/16/07 4:34 PM Page 90

Plays/Field Studies
90

North A
Present Day A' South
North A A' South

Mercia Mudstone Group Meters 500

Mercia Mudstone Group 500
Meters
500 500
0 1000 2000
0 1000 2000 e
dsto
ne 0 dston 0
Top Ormskirk San Top Ormskirk San
1000 1000 1000
1000 Depth Metres
Depth Meters
Metres Meters 500 Scale
500 Scale
vs : hs = 4 : 1 vs : hs = 4 : 1
1500 Sherwood Sandstone Group 1500 1500 Sherwood Sandstone Group 1500

1000 1000

2000 2000 2000 2000

Palaeogene ?
500 500

West B B' C C' D D' East

1000 1000

500 1500 1500

mag.
mag.
1000 ill. swc gwc.
gwc. mag. 2000 2000
ill. ill.
Gas-water contact
Metres ?
1500
? Late Jurassic Top platy illite
500 500
Surface delineated by
magnetic anomaly
2000 Gas leg
1000 1000
Gas-water-contact
Top platy illite
1500 1500
Surface delineated by
magnetic anomaly
2000 2000

Figure 10. Cross-sections through the Morecambe Fields, showing the positions of known and proposed paleo-gas-water contacts. Figure 11. Sequential restoration of the Morecambe Fields structure, using known and proposed paleo-gas-water contacts.

The simple, nearly planar, tilted geometry of the postulated second
gwc shows it to postdate the top illite-defined gwc, which is deformed as
well as tilted (Figure 10). Its position above the top of the illite in the cen-
tral parts of the field (Figure 10) may have resulted from a second breach-
ing event, the anomaly-related surface recording the second phase of
reservoir charging. Equally, this configuration may have resulted from
continued growth of the field closure without appreciable gas charging
over the period between gwc demarcation (Figure 11).
Using the paleo-gwc model, the break in the outline of the anomaly
at the northern end of the dip-closed section of South Morecambe devi-
ates significantly from the gently curved surface. This may represent the
fill point of the structure at the time of formation of this gwc, the same
mineral-phase change being preserved along the then active migration
pathway. This interpretation is supported by subsidence models for the
area which identify the Tynwald Basin, immediately east of the fields, to
be a source of fluids to the reservoir (Cowan and Bradney, 1998).
North Morecambe does not appear to record the same history of
breaching as South Morecambe (Stuart, 1993). The postulated second
gwc postdates the precipitation of platy illite and would appear to be rel-
atively late in the diagenetic and gas-charging history of the field, the clo-
sure being formed and filled to the present level.
Possible sources for the anomalies
An important late-stage diagenetic event occurred, resulting in the for-
mation of significant volumes of H2S (presently a problem in many of the
fields in the East Irish Sea area). This is believed to have been a basinwide
event (Hardman et al., 1993) resulting from either biogenic sulphate
reduction (BSR) or thermogenic sulphate reduction (TSR). Both mecha-
nisms require hydrocarbons and a source of sulphate in solution (Machel,
1987, 1995) to proceed.
These events are believed to have been triggered by either (1)
Paleogene uplift, enabling ingress of nonsterile meteoric waters resulting
in BSR, or (2) a pulse in heat flow associated with extensive Paleogene
mafic magmatism and resulting in the expulsion of hot, highly saline
brines from deeper parts of the basin, causing TSR (Hardman et al.,
1993).
Regardless of the mechanisms involved, the reactions will have taken
place where incoming waters and hydrocarbons mix, i.e., within the tran-
sition zone at the base of the accumulations and along the principal migra-
tion pathways. The H2S liberated by these reactions is the most likely
reactant with iron minerals such as magnetite, hematite, and maghemite
driving reduction to magnetic pyrrhotite and nonmagnetic pyrite. These
reactions will have been most aggressive where H2S was in the greatest
supply, i.e., local to its source in the transition zone.
If these anomalies result from a mineral phase distributed throughout
the hydrocarbon column, then the intensity of anomaly would be
expected to diminish as the column pinches out gradually against the
downdip closure. This is not observed, suggesting the source of the anom-
alies to be a surface or layer close to the present gwc in North Morecambe
and coincident with a postulated paleo-gwc in South Morecambe.
04 Plays-Field Studies 4/16/07 4:34 PM Page 91
These models fit well with the concept of the causative minerals hav-
ing been contained largely within a paleo-transition zone, but they also
require that no other appreciable magnetization contrast is present within
the Sherwood Sandstone.
These mineralogical models are yet to be tested through the appro-
priate analysis of well and core data.
Summary
The imaging of high-spatial resolution aeromagnetic data over the
Morecambe Fields demonstrates the clarity with which the magnetic
method can image subsurface fault structure within the sedimentary sec-
tion, and how this may be applied to better understand fault linkage when
interpreting seismic data. Conventional 2-D seismic coverage cannot con-
fidently delineate the orthogonal, crosscutting fault suites in this area
without considerable margin for error in fault linkage.
Solutions to this problem include the acquisition of seismic data over
progressively tighter grids, eventually arriving at 3-D seismic data (an
increasingly common choice), or integration of high-spatial resolution
aeromagnetic data with conventional 2-D seismic data (a far less expen-
sive option).
The correlation of areal variations in magnetic intensity with particu-
lar horizons or layers provides an indication of the sort of detailed, depth-
specific geologic data contained within the magnetic signal. However,
high spatial-resolution data is required to isolate such features and corre-
late them with known geology.
In this example, it is postulated that aeromagnetic data have imaged
the mineralogical product of a single diagenetic event, an event that was
localized and required hydrocarbons as a coreactant. The generation of a
magnetic anomaly from the mixing of migrating hydrocarbons with
groundwater is not unique; however, this process usually is identified with
the seepage of hydrocarbons above the level of the accumulation (Machel,
1995). In this example, it is proposed that the position of the resulting
magnetic anomalies relates to the final stage of secondary migration and is
the product of the transient influx of new basin fluids into the migration
pathway, causing localized instability in the magnetic mineral suite.
The implications for aeromagnetic data are considerable, serving to
reposition the method from one with solely regional exploration applica-
tions to one which can image distinct geologic features which bear on
both the exploration and development sides of petroleum exploitation.
The pursuit of improved resolution, allied with rock-magnetization data
and modeling, will provide explanations for this class of magnetic signa-
ture and potentially represents the “next step” in the evolution of aero-
magnetic surveying.
References
Arter, G., and Fagin, S. W., 1993, The Fleetwood Dyke and the Tynwald Fault Zone,
Block 113/27, East Irish Sea Basin, in Parker, J. R., Ed., Petroleum geology of north
west Europe: Proceedings of the 4th conference: Geol. Soc. London, 2, 835–844.
Bushell, T. P., 1986, Reservoir geology of the Morecambe Field, in Brooks, J., Geoff, J. C.,
and Van Hoorn, B., Habitat of Palaeozoic gas in N.W. Europe: Geol. Soc. London,
Special Publication, 23, 189–208.
Cornwell, J. D., and Raines, M. G., 1994, Magnetic susceptibility of Permo-Triassic
rocks from NW England: BGS Tech Rep. WK/94/7C.
Colter, V. S., and Barr, K. W., 1975, Recent developments in the geology of the Irish
Sea and Cheshire Basins, in Woodland, A. W., Ed., Petroleum and the continental
shelf of NW Europe, 1, Applied Science, 61–75.
Cowan, D. R., and Cowan, S., 1993, Separation filtering applied to aeromagnetic data.
Expl. Geophys., 24, 429–436.
Cowan, G., and Bradney, J., 1998, Regional diagenetic controls on reservoir properties
in the Millom accumulation: implication for field development, in Meadows, N. S.,
and Trueblood, S., Eds., The petroleum geology of the Irish Sea and adjacent
regions: Geol. Soc. London Special Publication, 124, 373–386.
Cowan, G., Ottsen, C., and Stuart, I. A., 1993, The use of dipmeter logs in the struc-
tural interpretation and palaeocurrent analysis of Morecambe Fields, East Irish Sea
Basin, in Parker, J. R., Ed., Petroleum geology of north west Europe: Proceedings
of the 4th conference: Geol. Soc. London, 2, 867–882.
Donovan, T. J., 1972, Petroleum microseepage at Cement, Oklahoma: Evidence and
mechanism: AAPG Bull., 58, 429–446.
Farina, M., Esquival, D. M. S., and Lins de Barros, H. G. P., 1990, Magnetic iron-sul-
phur crystals from a magnetotactic organism: Nature, 343, 256-–58.
Fraser, A. J., and Gawthorpe, R. L., 1990, Tectono-stratigraphic development and
hydrocarbon habitat of the Carboniferous in northern England, in Hardman, R. F.
P., and Brooks, J., Eds., Tectonic events responsible for Britain’s oil and gas
reserves: Geol. Soc. London, Special Publication, 55, 49–86.
Hardman, J., Buchanan, J., Herrington, P., and Carr, A., 1993, Geochemical modeling
of the Irish Sea Basin: Its influence on predicting hydrocarbon type and quality, in
Parker, J. R., Ed., Petroleum geology of north west Europe: Proceedings of the 4th
conference: Geol. Soc. London, 2, 809–821.
Hawthorne, T. B., and McKenzie, J. A., 1993, Biomagnetic magnetite: authigenesis
and diagenesis with changing redox conditions in Lake Greifen, Switzerland, in
Aissaoui, D. M., McNeill, D. F., and Hurley, D. F., Eds., Applications of palaeo-
magnetism to sedimentary geology: SEPM Soc. for Sed. Geol., Special Publication,
49, 13–15.
Jackson, D. I., and Mulholland, P., 1993, Tectonic and stratigraphic aspects of the East
Irish Sea Basin and adjacent areas: contrasts in their post-Carboniferous structural
styles, in Parker, J. R., Ed., Petroleum geology of north west Europe: Proceedings
of the 4th conference: Geol. Soc. London, 2, 791–808.
Kirkland, D. W., Denison, R. E., and Rooney, M. A., 1995, Diagenetic alteration of
Permian strata at oil fields of south central Oklahoma, USA: Marine and
Petroleum Geology, 12, 629–644.
Geologic Applications of Gravity and Magnetics: Case Histories
Knipe, R., Cowan, G., and Balendran, V. S., 1993, The tectonic history of the East
Irish Sea Basin with reference to the Morecambe Fields, in Parker, J. R., Ed.,
Petroleum geology of north west Europe: Proceedings of the 4th conference:
Geol. Soc. London, 2, 857–866.
Lewis, C. L., Green, P. F., Carter, A., and Hurford, A. J., 1992, Elevated KT tempera-
tures throughout northwest England: Three kilometers of Tertiary erosion?: Earth
and Plantetary Science Letters, 112, 131–145.
Lovely, D. R., Stoltz, J. F., Nord, G. L., and Phillips, E. J. P., 1987, Anaerobic produc-
tion of magnetite by a dissimilatory iron-reducing organism: Nature, 330,
252–254.
Machel, H. G., 1987, Some aspects of diagenetic sulphate-hydrocarbon redox reac-
tions, in Marshall, J. D., Ed., Diagenesis of Sedimentary Sequences: Geol. Soc.
London, Special Publication, 36, 15–28.
——— 1995, Magnetic mineral assemblages and magnetic contrasts in diagenetic
environments—with implications for studies of palaeomagnetism, hydrocarbon
migration and exploration, in Turner, P., and Turner, A., Eds., Palaeomagnetic
applications in hydrocarbon exploration and production: Geol. Soc. London,
Special Publication, 98, 9–32.
Mc Lean, A. C., 1978, Evolution of fault controlled ensialic basins in north western
Britain, in Bowes, D. R., and Leake, B. E., Eds., Crustal evolution in north west-
ern Britain and adjacent regions: Geol. J. Special Issue, 10, 325–346.
Mussett, A. E., Dagley, P., and Eckford, M., 1976, The British Tertiary Province:
Palaeomagnetism and age of dykes, Lundy Island, Bristol Channel: Geophys. J.
Roy. Astr. Soc., 505–513.
Stuart, I., 1993, The geology of the North Morecambe gas field, East Irish Sea Basin,
in Parker, J. R., Ed., Petroleum geology of north west Europe: Proceedings of the
4th conference: Geol. Soc. London, 2, 883–895.
Stuart, I. A., and Cowan, G., 1991, The South Morecambe Field, Blocks 110/2a,
110/3a, 110/8a, UK Irish Sea, in Abbots, I. L., Ed., United Kingdom oil and gas
fields: 25 years commemorative volume: Geol. Soc. London, Memoir, 14,
527–541.
Trueblood, S., Bryan, C., and Pickering, S., 1995, The Douglas oil field and its implica-
tions for exploration on the Irish continental shelf, in Croker, P. F., and Shannon,
P. M., Eds., The petroleum geology of Ireland’s offshore basins: Geol. Soc.
London, Special Publication, 93, 39–40.
Woodward, K., and Curtis, C. D., 1987, Predictive modelling of the distribution of
production constraining illites—Morecambe gas field, Irish Sea, Offshore UK, in
Brooks, J., and Glennie, K., Eds., Petroleum geology of northwest Europe:
Graham & Trotman Ltd., 205–215.
Yaliz, A., 1998, The Douglas oil field, in Meadows, N. S., and Trueblood, S., Eds., the
petroleum geology of the Irish Sea and adjacent regions. Geol. Soc. London,
Special Publication, 124, 399–416.
Additional Reference
Burley, S. D., 1984, Patterns of diagenesis in the Sherwood Sandstone Group
(Triassic), United Kingdom: Clay Minerals, 19, 403-–40.
91
04 Plays-Field Studies 4/16/07 4:34 PM Page 92
General Philosophy 4
edited by David A. Chapin
LaCoste & Romberg
Austin, Texas, USA
I offer these statements about the inter-
pretation process:
Interpret using “natural processes.”
A contractor’s job often consists of
enhancing potential-field data to make
them as interpretable as possible (by
either the contractor or the client). In
getting data ready for interpretation, I
like to confine the work to “natural
processes” and avoid arbitrary processes.
I see too many people abusing the facili-
ties available in commercial software
packages and applying polygonal fits of
the third, fourth, fifth, seventh, ninth
order and calculating residuals until
something strikes a chord, and then off
they go and interpret the results.
Alternatively, all too many people apply
arbitrary frequency-domain filters chosen
to pretty up the data without a lot of
thought about the consequence. Then
you no longer have a potential field.
By “natural processes,” I mean processes
that purport to do physically and/or geo-
logically meaningful things such as:
1) pole reduction
2) vertical continuation
3) pseudogravity or pseudomagnetics
4) vertical gradient
5) pseudodepth slicing
6) isostatic residual
These all do things to your data which
leave the data behaving like a potential
field or the gradient of one, so it can still
be modeled and understood. “But there
are assumptions in many of these
processes,” I hear people say. Sure there
are, but they are not too difficult to keep
in mind.
Whenever possible, avoid using second-
vertical derivatives.
I try to avoid second-vertical derivative.
It still shows up all too often in contract
specifications, largely for historical rea-
sons. If you remember that far back, it
can be generated in a reasonable time
using a hand-cranked Facit calculator,
and that is its main (only?) merit.
Avoid using variable-density Bouguer
corrections.
Some noted authorities (the late Jaques
Lakshmanan, to name one) like to use
laterally varying Bouguer densities based
on the observed local geology. Unfor-
tunately, some people who follow this
approach subsequently fail to record the
variable density used, so you don’t get
the chance to undo the process. The
other problem is that there is an implied
assumption that the density used goes
from the surface to sea level, which may
be OK in the Netherlands but is unlikely
in the Cordillera.
So I’ve always used fixed densities
(2.67), not because we think that’s bet-
ter, but because it’s a standard treatment
and it’s easily undone when you want to
do interpretation work.
Interpret with a good structural geolo-
gist, if you can.
I like to work with a good structural
geologist. They know all sorts of things
about the way the Earth behaves, which
Plays/Field Studies
aren’t that well known to the average
gravity and magnetic interpreter. On the
other hand, most structural geologists
don’t understand potential fields all that
well. Together, a great team emerges.
High-sensitivity, high-resolution aeromag-
netics applied to sedimentary basins
really took a major leap forward when
this concept was recognized. For too
long, our profession has been dogged by
“geophysicist” interpretations which are
often structurally naive.
Alan B. Reid
Geophysical Exploration Technology (GETECH)
Leeds, United Kingdom
My impression is that a large percentage
of gravity interpretation (>90%?) is for
determination of geologic structure (ide-
ally, in conjunction with other indepen-
dent data sets).
Don’t forget that gravity is a dynamic
force too!
Stresses caused by lateral density con-
trasts in the upper, brittle crust may
localize seismicity and may play a role in
other geologic processes. On a more
regional scale, the nature and extent of
isostatic balance may provide important
clues to tectonic processes. An elegant
example is the observation by Thompson
and McCarthy (1990) that the lack of
large Bouguer gravity anomalies associ-
ated with the highly extended core com-
plexes of the western Cordillera requires
deep emplacement of crustal-density ma-
terial in conjunction with that extension.
Rick Saltus
U. S. Geological Survey
Denver, Colorado, USA
Approach cultural effects in aeromag-
netic data with caution.
There are two ways to approach this.
One is to interpret “through” the cul-
ture, using whatever references are at
hand. This approach presents obvious
problems if high-frequency enhance-
ments are required. The more rigorous
method is to edit out cultural anomalies
from the profile data, preferably through
direct reference to the flight videos. In
such cases, I find it useful to refer to a
“difference” map (i.e., before and after
editing) during interpretation, so that the
residual cultural effects that remain in
the data are recognized easily.
Use of the “analytic-signal” method also
has physical meaning.
Another magnetic enhancement tech-
nique, with a physical meaning, is compu-
tation of the amplitude of the analytic
signal (Nabighian, 1972, 1974, 1984;
Roest et al., 1992). This enhancement is
particularly useful when dealing with sig-
nificant magnetic remanence and/or low
magnetic latitudes, because the analytic
signal is unaffected by the phase of the
anomaly (MacLeod et al., 1993; MacLeod
et al., 1993). After years of interpreting
complex or poorly represented anomalies,
it is amazing to see how the sources “coa-
lesce” on an analytic-signal map. I use it
as a matter of course in South America
and Africa, where the geology and inclina-
tion conspire to make life difficult for the
interpreter.
Don’t overlook physical rock properties.
It is fairly obvious that as much physical-
property information (e.g., density, sus-
ceptibility, remanence) should be incor-
porated into modeling and interpretation
92
as feasible. It often involves a lot more
work, not only in the collection of the
data, but in making sense of it at the
interpretation stage. It also can be ex-
tremely useful to revisit interpretations
during drilling, to incorporate new infor-
mation and perhaps significantly alter the
target selections for further drilling. In
areas of good well control, the density
and horizon depth information can be
used quite effectively to strip unwanted
gravity signatures and locate targets
whose responses are barely above the
noise level of the data. The advent of
high-resolution aeromagnetics in basin
areas in recent years has led to a plethora
of intrasedimentary anomaly sources.
Much work remains to be done to deter-
mine the physical processes behind these
anomalies, and ultimately their signifi-
cance to petroleum exploration.
Stephen W. Reford
Paterson, Grant & Watson Limited
Toronto, Ontario, Canada
Concerning aeromagnetic cultural effects,
we spend considerable effort on our high-
resolution aeromagnetic surveys to re-
move cultural influence. We have found
the only universally effective approach is
to hand edit each profile carefully in con-
junction with reviewing flight videos. To
do this, we follow these steps:
1) Assemble contractor’s field QC paper
printouts by flight number.
2) “Fly” each video on the VCR with
the paper printouts. Where there is
any unusual magnetic activity, note
any possible cultural sources. This
process takes about as many hours
on the VCR as it took to fly the origi-
nal survey, unless there is not much
human activity present.
04 Plays-Field Studies 4/16/07 4:34 PM Page 93
3) Plot a large-scale (e.g., 1:100,000)
flight map with digital representa-
tions of all known culture. In west-
ern Canada, we have pipelines and
wells easily available digitally, but still
have to look for culverts under oil-
company roads, grain silos, junk-
yards, etc. Transfer VCR edit infor-
mation onto the cultural flight map.
4) Edit each profile on a workstation. If
one is flying close to the ground and
draped, then one must build up an
understanding for the topographic
response in your area (differential
ground clearance, e.g., crossing steep
valleys).
It is no mean feat to accomplish all this
on a large survey, but the effects can be
dramatic on a spectral comparison. It is
essential for good depth estimates if one
is looking for intrasedimentary sources.
John Peirce
GEDCO
Calgary, Alberta, Canada
Use hypothesis testing as an approach
toward interpretation.
I like to approach interpretation from a
hypothesis-testing point of view. The
hypotheses are developed by gathering
all information about the problem or area
and discussing the possibilities with all
the geoscientists who are working or
have worked in the area. Usually there is
already a hypothesis or hypotheses pro-
posed by geologists. I like to include all
of these, even if it is obvious by inspec-
tion of the geophysical data that one
won’t work, to demonstrate the problem
clearly to the proponent of the hypothe-
sis. Other hypotheses arise from brain-
storming from the geophysical data in
conjunction with all the information and
talking to geologist colleagues. Normally,
the data allow for several hypotheses,
but some also get eliminated. I find it’s a
good way to focus follow-up investiga-
tions and to communicate the range of
possibilities to people. I also like using
ideal body method, etc. This is another
form of hypothesis testing.
V. J. S. (Tien) Grauch
U. S. Geological Survey
Denver, Colorado, USA
The term anomaly often is misunder-
stood.
Here is a quote from Norm Sleep:
“Most measurements of gravity during
the 1700s and 1800s were made by sci-
entists interested in planet-wide varia-
tions rather than surficial geology. Local
variations in gravity, which did not cor-
respond to a simple astronomical scale
model of the earth, were thus consid-
ered anomalous. This reasoning gave
rise to the currently used terms for local
and regional variations in gravity: gravity
anomalies or anomalous gravity field.
The term anomaly is not intended to
imply a mysterious origin, but rather a
deviation from idealized models of the
earth.”
Norm Sleep, Stanford University
Geologic Applications of Gravity and Magnetics: Case Histories
Rick Saltus
U. S. Geological Survey
Denver, Colorado, USA
References
MacLeod, I. N., Jones, K., and Dai, T. F., 1993,
3-D analytic signal in the interpretation of
total magnetic field data at low magnetic lati-
tudes: Exploration Geophysics, 24,
679–688.
MacLeod, I. N., Viera, S., and Chaves, A. C.,
1993, Analytic signal and reduction-to-the-
pole in the interpretation of total magnetic
field data at low magnetic latitudes, Proce-
edings of the Third International Congress
of the Brazilian Society of Geophysicists.
Nabighian, M. N., 1972, The analytic signal of
two-dimensional magnetic bodies with polyg-
onal cross-section: Its properties and use for
automated anomaly interpretation:
Geophysics, 37, 507–517.
——— 1974, Additional comments on the ana-
lytic signal of two-dimensional bodies with
polygonal cross-section: Geophysics, 39,
85–92.
——— 1984, Toward a three-dimensional auto-
matic interpretation of potential field data via
generalised Hilbert transforms: Fundamental
relations: Geophysics, 49, 780–786.
Roest, W. R., Verhoef, J., and Pilkington, M.,
1992, Magnetic interpretation using the 3-D
analytic signal: Geophysics, 57, 116–125.
Thompson, G. A., and McCarthy, J., 1990, A
gravity constraint on the origin of highly
extended terranes: Tectonophysics, 174,
197–206.
Additional Reference
Cordell, L., and McCafferty, A. E., 1989, A ter-
racing operator for physical property map-
ping with potential field data: Geophysics,
54, 621–634.
High-resolution Aeromagnetic
Interpretation over Sierra and Yoyo Reefs,
Northeastern British Columbia
John Peirce, Erwin Ebner, and Nathalie Marchand
Geophysical Exploration & Development
Corporation (GEDCO)
Calgary, Alberta, Canada
EDITOR’S NOTE: This is an excellent example of how magnetic sur-
vey parameters bear on the interpretation of subtle geologic
features that can directly affect the prospective sedimentary
section. Sedimentary sources for magnetic anomalies are
defined with confidence. Peirce et al. also show the kinds of
problems inherent in areas with cultural “noise,” and they
document processing difficulties that affect the interpretation
of high-resolution aeromagnetics.
Abstract
There has been a sudden resurgence of aeromagnetic data acquisition in
western Canada since 1994 because the new high-resolution aeromag-
netic (HRAM) surveys are now able to map intrasedimentary faults and
fractures on a regional scale, whereas previously, aeromagnetic data were
used only to delineate basement features. This increase in resolution has
been achieved primarily because of more precise navigation and position-
ing of data using the Global Positioning System (GPS), and also because
of flying in controlled drape mode close to the ground to enhance shallow
sources. Better instrumentation in the aircraft and improved software for
analysis and visualization also have helped.
In 1994, the first HRAM survey in western Canada was flown for
Focus Seismic Corp. by Questor Surveys, Ltd., over the Sierra and Yoyo
Reefs in northeastern British Columbia. Although these Devonian reefs
are well known and have been producing gas for years, the area is also
one of active exploration for smaller pinnacles and carbonate bank plays.
This paper discusses many of the concepts applicable to designing, fly-
ing, processing, and interpreting an HRAM survey, using the Sierra survey
to demonstrate the points. Particular emphasis is placed on editing the
data culturally, choosing a line spacing appropriate to the complexity of
the expected results, and the value of depth analysis on the profiles to
resolve subtle magnetic anomalies associated with faults and fractures.
93
04 Plays-Field Studies 4/16/07 4:34 PM Page 94

Such linear features can be mapped on adjacent profiles and traced for
many kilometers. The consistency of the fault interpretations from profile
to profile is an essential factor in sorting out geologically significant fea-
tures from mathematical artifacts.
We hypothesize that the cause of these fault-related intrasedimentary
anomalies is magnetization in the plane of the faults or fractures. Recent
work (Pierce et al., 1998) has demonstrated the importance of the sulfur
geochemistry in catalysing reactions involving iron-bearing minerals in
fractures in shales. Specifically, authigenic macroscopic pyrrhotite has
been observed in shales at shallow depth. Pyrrhotite is important mag-
netically because it has a similar susceptibility to that of magnetite. Pyrite,
which often is considered nonmagnetic but actually has a small suscepti-
bility, also may contribute significantly to fracture magnetization when
the surrounding sedimentary section is nonmagnetic.
This approach to magnetic interpretation allows one to map regional
faulting patterns over large areas at relatively low cost. The final product,
a magnetic structural-grain map, allows one to define exploration fairways
and to position seismic programs more efficiently, and it usually reveals
several new exploration leads.
Introduction
Plays/Field Studies
and a digital terrain model allows pilots to fly a preprogrammed drape
surface which produces much cleaner magnetic data with minimal
line-oriented noise.
3) Improved aircraft magnetic compensation systems and magnetome-
ters allow more frequent sampling (generally < 10 m on the ground)
with less aircraft-induced noise.
4) Better processing and visualization software allow large volumes of
data to be filtered quickly and displayed in shaded relief formats
which highlight subtle trends.
In this paper, we will discuss some general considerations regarding
survey design, processing, and interpretation of a high-resolution aero-
magnetic (HRAM) survey, using the Sierra HRAM survey to illustrate our
points. The Sierra survey (Figures 1 and 2) was acquired in 1994 by
Questor Surveys Limited for Focus Seismic Corp. as a nonexclusive sur-
vey. The purpose of the survey was to fly a test survey over large, well-
known Devonian reefs (Phipps, 1982; Collins and Lake, 1989) in an area
where current exploration interest still exists for smaller features.
94
Geologic Setting
The dolomitized Middle Devonian carbonates of northeastern British
Columbia (NEBC) host numerous gas fields. The gas accumulations are
trapped structurally and stratigraphically along the edge of carbonate
banks or in pinnacle reefs within the basinal facies (Figures 2 and 3). The
positions of the bank edges can vary significantly, from the time of Keg
River (Eifelian-Givetian: mid-Middle Devonian) to the later development
of the Slave Point carbonates banks (Givetian to late Middle Devonian).
Seals are provided by the surrounding Klua and Otter Park shales in the
basins, and by overlying Muskwa and Fort Simpson shales for the car-
bonate banks (Phipps, 1982).
The Yoyo, Sierra, and South Sierra reefs are relatively large pinnacle
reefs which have been producing gas for decades. Numerous smaller pin-
nacle reefs are in the area, but some are not productive because sufficient
dolomitization has not occurred to develop reservoir properties. The
smaller reefs are dolomitized bioherms of Pine Point dolomite (Givetian
age) 125–150 m high, and the larger compound reefs, such as Sierra,
include 200–250 m of Pine Point limestone overlying the dolomite
(Collins and Lake, 1989; Figure 3).

Why is there suddenly so much interest in aeromagnetics in western

Canada? Activity has increased from a negligible level in 1993 to about
300 000 km of new flying in 1995. For the oil industry, 1.8 million line
km of new HRAM data were acquired in 1994–1997 in western Canada.
In addition, approximately 500 000 line km of HRAM data were acquired
in Alberta for Diamond Exploration in 1996–1997. The simple answer is
that now it is possible to map intrasedimentary faults and fractures which
are magnetized, whereas previously, magnetics was used primarily to
decipher basement tectonics. The interpretive result is a structural-grain
map which complements seismic interpretations and covers a much
larger area, and the total cost is very modest by seismic standards.
What are the improvements in aeromagnetic surveying that allow us
to resolve intrasedimentary features? There are four main factors:
1) The Global Positioning Satellite (GPS) allows precise positioning of
the aircraft to a few meters’ accuracy horizontally. This improved
accuracy allows us to resolve spatially very small-scale magnetic
anomalies which previously were obscured by positioning errors.

2) Drape flying close to the surface (100- to 150-m ground clearance)

enhances the amplitude response of magnetic sources in the sedi-
mentary section. Using real-time differential GPS for vertical control Figure 2. Position of the Sierra Survey relative to mid-Devonian gas fields, bank
Figure 1. Location map for the Sierra Survey. edges, embayments, and pinnacle reefs.
04 Plays-Field Studies 4/16/07 4:34 PM Page 95
Figure 3. Generalized cross-section across a carbonate bank edge showing reef and porosity devel-
opment at the bank edge and a compound pinnacle reef in the basin (modified after Phipps, 1982).
The pinnacle reef is generalized from the Sierra Reef, which is formed by porous Pine Point dolomite
overlain by relatively nonporous Pine Point limestone (Collins and Lake, 1989).
The bank edges are dolomitized selectively and form another com-
mon target. In addition, within the Keg River embayments are small but
prolific targets of selective dolomitization in the carbonates, below the
overlying Klua shales.
Collins and Lake (1985) discuss the dolomitization of the Sierra Reef
in detail. They describe two separate dolomitization episodes evident in
the Pine Point dolomite, neither of which is related to subaerial exposure
of the reef. The first event caused pervasive dolomitization of the entire
reef. The second event was associated with brecciation, fracturing, and
the subsequent infilling of vertical fractures with normal, coarsely crys-
talline dolomite as well as saddle dolomite (Radke and Mathis, 1980;
Read, 1985). In particular, this latter dolomitization was associated
closely with pyrite mineralization and is believed to have been caused by
hydrothermal activity.
The Kotcho carbonate bank, to the east of the Sierra and Yoyo reefs,
has numerous productive fields (Figure 2) where dolomitization has
developed good reservoir, usually at the Slave Point level.
The primary objective of the HRAM survey was to determine what
structural control influenced the position of carbonate bank edges and the
nucleation of pinnacle reefs, if possible. Secondarily, it was hoped that
some hydrocarbon accumulations might have an asso-
ciated magnetic response from magnetization related to Sierra: Final Edited WRTP Total Magnetic Field
dolomitization or perhaps to the development of over-
lying geochemical chimneys. In this particular area, any
technique which could identify areas more likely to
undergo dolomitization could reduce exploration risk
significantly.
Survey Design
In general, the HRAM surveys in western Canada are
being flown with traverse line spacings of 300–1200 m,
with control lines at about three times the traverse line
spacing. The purpose of the control lines is primarily to
provide good control for leveling, and secondarily to
provide another line orientation if multiple geologic
trends are anticipated. If the survey is at high magnetic
latitude, the control lines usually are perpendicular to
the traverse lines, but not always.
The choice of traverse line spacing for an HRAM
survey depends on the anticipated complexity of the
intrasedimentary magnetic signal (generally less com-
plex as the sedimentary section gets thicker) and the
interpretation approach used. Interpreters who depend
Figure 4. Total-intensity aeromagnetic map over the area around the Sierra and Yoyo Reefs (north-
entirely on map images prefer the smallest possible pixel size for east British Columbia) after editing (see text). Red and yellow colors are highs; blue and magenta
display, and therefore generally choose tight line spacing. Our colors are lows. The Sierra aeromagnetic survey was acquired for Focus Seismic Corporation in
interpretation approach depends more on the integration of mag- 1994 by Questor Surveys Inc. The area shown is about 40 × 55 km.
netic depth solutions with the mapped data, and pixel size is not
a critical criterion for us. Therefore, we tend to choose relatively The resulting total field data (Figure 4) show a long-wavelength gra-
wider line spacings (600–1000 m seems to be most cost- effective in our dient caused by the juxtaposition of two structural terranes (Ross et al.,
experience) if the section is deep (> 2500 m) and the scope of the project 1991, 1994) with very different magnetic signatures.
is large.
In the Sierra survey, the line spacing chosen was relatively tight, at Processing
400 × 1200 m, because it was a test survey over known reefs, and we
were not sure what to expect. The survey was drape flown at nominal 1) In-field quality control (QC)
ground clearance of 100 m. GPS control was used for in-flight navigation, Proper data processing begins in the field. In-field processing is a cru-
and postprocessed differential GPS was used for final horizontal control cial component of quality control by the contractor. This QC will pro-
for data positioning. Barometric and radar altimeters were used for verti- vide an early indication of acquisition problems and allow the crew to
cal control. In more recent surveys, we have had excellent success flying remedy deficiencies before proceeding further with the survey.
in a controlled drape mode in which real-time differential GPS is used to Furthermore, the results from field processing (line plots of all para-
fly to a preprogrammed controlled drape flying surface. This strategy min- meters, initial mapping, filtering, and visualization) will allow the
imizes line-to-line vertical positioning mis-ties and reduces magnetic mis- client to have effective, independent quality control provided on-site
ties significantly. to ensure that survey specifications are being met. If surprises occur,
95
Geologic Applications of Gravity and Magnetics: Case Histories
04 Plays-Field Studies 4/16/07 4:34 PM Page 96
additional lines can be added easily while the plane is still on-site, to
detail unexpected features.
2) Cultural editing
Until the last few years, most aeromagnetic data have been acquired
at relatively high distances above the ground, using wider line spac-
ings, less precise navigation, and less sensitive acquisition systems
than are available today. Therefore, cultural interference effects were
less of a concern. Because HRAM surveys are flown at low ground
clearances, cultural noise has become a serious problem, although its
deleterious effects are not yet recognized universally.
The data sets delivered by acquisition contractors have been lev-
eled carefully to remove any mis-ties at line crossings, but editing of
“cultural features” (man-made magnetic sources at surface) is an
interpretive task. Cultural features such as wells, pipelines, bridges,
processing plants, etc., generate intense magnetic anomalies with
short wavelengths. A typical anomaly caused by a cased well is a spike
of 8–20 nT if the plane passes directly over the well, or a much more
subtle peak of < 1 nT if the flight path is 100 m away from the well-
head (Figure 5, upper panel). The detrimental effects of cultural noise
become very visible in residual and derivative data products, where
Figure 5. The upper panel shows an unedited profile, including the magnetic response
of three wells at fiducial locations 3050, 3200, and 4100. The lower panel shows the
same profile after cultural editing and the application of a Wiener filter. Some high fre-
quencies at fiducial locations 0–1000 and near 2100 also have been edited because of
culture seen on video review.
Plays/Field Studies
96
they obscure much of the subtle signal. Furthermore, because cultural data for the Sierra survey, calculated by subtracting the edited data
noise is spikelike, any digital filtering of the data set will cause ringing grid from the unedited data grid.
effects, enlarging the size of the anomaly and sometimes showing up 3) Gridding, filtering, and mapping
again in aliased harmonics. The solution is to edit the profiles as the
first step in interpretive processing (Figure 5, lower panel). Applying a low-amplitude Wiener filter (0.1–0.25 nT) to the releveled
profile data is a prudent first step, to remove residual low-amplitude
Efficient editing of cultural anomalies requires special software,
noise. We also apply such a filter after gridding, to suppress grid noise.
reliable knowledge of cultural surface sources, good in-flight video,
The choice of grid-cell size is very important. If one is using a stan-
and last, but not least, a competent interpreter who can remove the
dard rectangular gridding algorithm, we recommend that the grid-cell
cultural interference effect without changing the subsurface geologic
size be no smaller than a third of the traverse line spacing, to avoid
response. In most cases, isolated cultural anomalies are removed quite
excessive grid noise in the filtered products. A larger grid-cell size usu-
easily, while more complex anomalies caused by some pipelines,
ally is not suited for higher resolution image displays. This may be a
power lines traversed at acute angles, and overlapping cultural
problem for surveys in which wider line spacing is quite sufficient for
sources are much more difficult. River valleys and lakeshores often
geophysical reasons, but magnetic images (as contrasted with images
will generate a high-frequency magnetic signal, which can result from
together with profiles) are the only medium for presentation and
varying ground clearance, the effect of outcropping layers in the
interpretation. Regridding and antialias filters are a valid way around
scarps of the valleys, and varying sand lithologies related to modern
this dilemma (although the results may lack the resolution desired).
or relict stream channels. Deciding what is “cultural” and what is geo-
logic is always a challenge.
Our approach is to edit each profile on the workstation,
clipping out each cultural feature and interpolating across
Sierra: Difference Grid = Original TF - Edited TF
the gap in a manner which preserves the noncultural fre-
quency content. We estimate that perhaps 90% of the cul-
tural interference can be removed in this manner from a
typical data set from western Canada.
Another approach is to use special digital filters which
are custom designed for each cultural editing task and are
applied on a line-by-line basis. Deciding on the appropriate
parameters for each filter type requires many passes through
the data (Pearson, pers. comm., 1996). In our experience,
editing by hand on the workstation will produce the “clean-
est” data set (subjectively, the least cultural contamination).
The more automated filtering approach may be more cost-
effective in areas where the cultural contamination is lim-
ited to isolated cultural sources or in situations in which
suppression, rather then elimination, of the cultural signal is
the objective. Cultural editing is still in the early stages of
development; we expect to see significant improvements in
efficiency in the next few years. Hassan et al. (1998) com-
pare three methods of cultural editing.
As the density of cultural effects increases, it becomes
harder to remove only the culture without affecting the Figure 6. Difference map obtained by subtracting edited data (Figure 4) from unedited data (not
shown). The black lines are pipelines. The black crosses are wells. Note the large number of anom-
higher frequency components of the geologic signal also. alies associated with wells. The red linear feature in the southwest is interpreted as being a
Figure 6 shows the difference between unedited and edited Pleistocene stream channel with higher susceptibility gravels near the surface.
04 Plays-Field Studies 4/16/07 4:34 PM Page 97
The next step is the application of a reduction-to-the-pole (RTP)
filter. This filter applies a latitude-dependent phase shift to center an
anomaly peak over its source. This step greatly simplifies interpreta-
tion of maps and is done easily at high latitudes. However, in cases
where the magnetic inclination is low, special approaches may be
needed to accomplish the same objective.
Digital filtering allows one to analyze and enhance large data sets
quickly. Because the wavelength of the magnetic anomaly is strongly
dependent on the depth of the source, band-pass filters can be used to
separate deeper sources from shallow sources qualitatively. Vertical
derivative filters amplify the high-frequency components of the data,
and horizontal gradient filters produce an “edge” map on which high
data values are centered over the edges of the magnetic sources (e.g.,
to highlight faulting relationships). Strike filters can be used in special
situations to suppress or enhance trends, but they should be used
Sierra: Unedited Data, 1.5 - 7 km Band-pass
Yoyo
Sierra
S. Sierra
Junior
Embayment
Figure 7. Band-pass filter of 1.5- to 7-km wavelength of the unedited data in shaded relief format.
Illumination is from the northeast; red and yellow colors are highs; green and blue colors are lows.
Structural contours on mid-Devonian carbonate, taken from British Columbia government map, are
shown. Note that where the structural contours narrow in the south-central portion of the image, a
basement fault offsets the magnetic anomalies.
Geologic Applications of Gravity and Magnetics: Case Histories
with caution. Numerous other specialized filters are avail-
able for specific objectives or data problems. Sierra: Edited Data, 0-4 km Band-pass
Figure 7 shows an example of a band-pass filter
designed to highlight anomalies arising from the upper
basement and the lower sedimentary section. Because the
Precambrian surface is at about 2500-m depth (subsurface)
in this area, we can expect anomalies caused by sources
near the Precambrian surface to have dominant wave-
lengths of about 5000 m or longer. Applying a 1.5- to 7-km
band-pass filter to the edited total field data (Figure 4) sup-
presses the long-wavelength anomalies from the lower
crust and produces a dramatically different perspective.
For reference, structural contours on the mid-Devonian
carbonate are shown as an overlay.
Figure 8 is a higher frequency band-pass (0-4 km)
designed to minimize the effects of basement, and
it shows the intrasedimentary magnetic fabric.
Figure 9 is a combination of filters designed (by
empirical experimentation) to highlight an east-
west-striking feature which we have named the
Sierra lineament. In this case, a strike filter was
used to suppress the strong north-south Yoyo
Magnetic High. The importance of this lineament
as a structural feature has been confirmed by sub- Figure 8. Band-pass filter of 0- to 4-km wavelength, designed to highlight only sedimentary anomalies
sequent seismic and drilling. with minimal contribution from basement. Note that the Yoyo Magnetic High, which runs north-south
Display formats have changed over the years, from the Yoyo Reef, is still clearly visible, indicating that the intrasedimentary magnetic fabric is influ-
with ever more visualization software available. enced strongly by underlying basement structure. One of the lines over the Yoyo Reef was so contami-
nated by cultural effects (the Yoyo gas plant had a 300-nT anomaly in the original data) that we could
The actual display format, whether it has simple not reconstitute it totally. This is the origin of the sharp northeast-trending low on the east side of the
contours, color contours, shaded relief, or color- Yoyo Reef on this and some other images.
shaded relief, depends on the frequency content of
the data and what highlights critical features in the most effec- 10 shows the results for the same data with every second line dropped
tive manner. In this regard, manipulation of the dynamic (800- × 2400-m line spacing) at the 0- to 4-km frequency band of inter-
range of the colors can improve the visibility of certain fea- est, and Figure 11 shows the results for every third line dropped (1200-
tures dramatically. It is very important that an appropriate × 3600-m line spacing) at a band of the same frequency. Figures 12 and
choice of scale be made and that the map is annotated suffi- 13 show the difference grids. In our estimation, there is no significant
ciently for the user to cross-reference the results with other coherent signal in Figure 12. The primary differences seem to relate to
information easily. cultural effects. However, there clearly is some coherent signal in Figure
13. Our conclusion is that 800- × 2400-m line spacing would have been
Optimal Line Spacing optimal for a relatively small survey such as the Sierra survey. For a larger
Given that we have determined that 0- to 4-km band-pass survey covering several square degrees with more regional objectives, a
was the highest frequency filter we chose to use in this area traverse line spacing of 1000–1200 m probably would be most cost-effec-
tive and give essentially the same results.
(Figure 8), we now can examine the question, in hindsight,
of what is the most cost-effective line spacing to use. Figure
97
04 Plays-Field Studies 4/16/07 4:34 PM Page 98

Plays/Field Studies
98

Sierra: 1.7 - 4.1 km Band-pass + 180°/15° Sierra - Every 2nd Line, 0-4 km Band-pass of WRTP
Reject

Sierra Lin
eament

Figure 9. Band-pass filter of 1.7- to 4.1-km, cascaded with a 180°-strike suppression filter (15° wide), and applied to the edited data Figure 10. Sierra survey reconstructed using only every second line (800- × 2400-m line spacing), and filtered at 0- to 4-km
to highlight the Sierra lineament. Two other possible east-west lineations are indicated by the orange arrows. The black Xs indicate band pass.
the positions of wells.

Depth Interpretation Profiles 1953; Jain, 1976; Ku and Sharp, 1983; Rao, 1984) and 2-D Euler ters (circles and crosses on Figure 14) in a depth section with the profile
(Thompson, 1982) are our mainstays. Useful supplementary techniques (and its gradient or derivative) above. Even after clustering, a very large
It is somewhat ironic that less emphasis is placed now on interpreting the
are the various spectral and gradient analysis approaches and the Phillips number of solutions still are plotted. At this point, an experienced inter-
profile data for depth information than in previous years, when data accu-
method (Phillips, 1979), as well as the “old” manual methods such as preter must decide which solutions are significant and assign geologic
racy was lower. Grid products provide an excellent qualitative perspec-
Peters’ slope technique (Peters, 1949) and a series of mostly Russian man- meaning to them.
tive, but it is still necessary to interpret the profiles to extract quantitative
ual estimation approaches (implemented interactively on the screen). Figure 14 shows the results of magnetic depth interpretation using
depth information (except for 3-D Euler and 3-D inversion). Depth analy-
Another digital depth-estimation technique is 3-D Euler (Reid et al., MAGPROBE™ on a profile which crosses the Yoyo Reef. (The unedited
sis of the profiles is time-consuming and requires appropriate software,
1990), which is applied to gridded data. trace of this profile was relatively uncontaminated by cultural noise
but the results are definitely worth it. It makes the difference between
The Werner and Euler techniques calculate the position of point because of the producing wells and the gas plant.) The depth estimates
“seeing” something on a map and providing reasonably precise estimates
sources which could cause the anomalies included in various data win- are plotted as a depth section, with the x-dimension being distance (m)
of the depth distribution of magnetic sources.
dows along the profile. Both techniques calculate huge numbers of possi- along profile and the z-dimension being subsea depth (m). The position of
We recommend using a combination of depth-estimation techniques
ble sources. Clustering techniques are used to accept only those sources the Yoyo Reef has been drawn in schematically for reference. Note the
to obtain the best results. Some of the methods respond better than oth-
which are grouped tightly—that is, if several solutions are grouped tightly grouping of depth estimates highlighted by yellow lines. The line above
ers to certain amplitude/frequency combinations, profile length, and fre-
in space, they are more likely to be valid. The software plots these clus- the word reef is interpreted as a magnetized east-dipping, listric fault
quency content. Well-documented methods such as Werner (Werner,
04 Plays-Field Studies 4/16/07 4:34 PM Page 99
Sierra - Every 3rd Line, 0-4 km Band-pass of WRTP
Figure 11. Sierra survey reconstructed using only every third line (1200- × 3600-m line spacing), and filtered at 0- to
4-km band pass.
plane which appears to pass through the reef. This intrasedimentary fault
or fracture is associated spatially with the eastern side of the basement
ridge marked by the Yoyo Magnetic High (Figure 15). A less well defined
fault or fracture is interpreted on the southwest side of the reef, associ-
ated with the west side of the same basement ridge.
The concept of magnetized faults or fractures has been confirmed for
us in numerous projects, both in western Canada and elsewhere.
Although we have seen several examples of seismic/magnetic correla-
tions (e.g., Peirce et al., 1998), it is clear that not all faults seen seismi-
cally are detectable using magnetic techniques. Conversely, many
magnetic “faults” may in fact be fractures with little vertical offset, and
therefore invisible seismically. Our sense is that these fault-related mag-
netic anomalies arise from magnetization along the fault or fracture plane,
in part because we often can see relationships between intrabasement
faults and subsequent intrasedimentary faults (often with opposite dips).
When depth interpretations are run on adjacent profiles, faults which
Sierra: 0-4 km Band-pass Filter; Original - Every 2nd Line
Figure 12. Difference grid made by subtracting Figure 10 from Figure 8. Note that very little coherent information is
lost by doubling the line spacing. The primary differences seem to be in the vicinity of cultural features around the
Sierra and Yoyo Reefs.
are magnetized consistently can be traced from profile to profile, and the Conclusions
interpreter can have some confidence that the interpreted faults or frac-
High-resolution aeromagnetic surveys now are producing new perspec-
tures are real geology and not some mathematical artifact. If the results of
tives of sedimentary structures. The increase in resolution has been
the depth interpretation are plotted at the scale of the filtered maps, then
achieved in large part because of the increased navigational accuracy pro-
the positions of these fault interpretations can be correlated in map format.
vided by the Global Positioning System (GPS), operated in differential
The result is a structural-grain map (Figure 15) which in some ways resem-
mode. Better magnetometers and drape flying close to the ground are
bles a satellite image interpretation, but one can code the results to indi-
used to sample higher frequencies, and the computer power of worksta-
cate which faults are intrasedimentary and which arise in the basement.
tions is used to analyze the data rapidly and to visualize it in ways that
Figure 15 shows a completed interpretation superimposed on the fil-
simply were not feasible a few years ago. The results are new and often
tered magnetic data (Figure 7). In fact, this interpretation is a portion of a
startling. As with any technique which opens up new vistas of interpre-
later, much larger project which happened to overfly the Sierra area with
tation, interpretive methods are evolving rapidly, and novel ways to apply
1200-m line spacing. This latter interpretation is considered more reliable
the new information are close behind.
than the original Sierra interpretation because it was made using longer
In this particular survey, the primary objective of understanding
lines, which give more stable depth solutions for relatively long wave-
regional structural control was met. Some participants’ hopes of being
lengths, and because the entire interpretation covered a much larger region.
able to detect small pinnacle reefs directly were not met. However, the
99
Geologic Applications of Gravity and Magnetics: Case Histories
04 Plays-Field Studies 4/16/07 4:34 PM Page 100

Plays/Field Studies
100

Sierra: 0-4 km Band-pass Filter; Original - Every 3rd Line Yoyo Reef - MagProbe Depth Analysis

NE

20nT

2km

1km

Yoyo Reef

Figure 13. Difference grid made by subtracting Figure 11 from Figure 8. Note that a small amount of coherent information is lost Figure 14. Results of magnetic depth analysis (using MAGPROBE™) on a profile passing over the Yoyo Reef. Depth-solution clus-
by tripling the line spacing. Our conclusion is that the range for most cost-effective spacing of traverse lines is 800-1200 m, ters are plotted for Werner (circles) and 2-D Euler (crosses) depth solutions, with yellow lines highlighting two interpreted faults
depending in part on the level of detail expected in the survey objectives. or fractures. The position of the reef is shown schematically for reference. We speculate that the line of solutions dipping south-
west over the word Yoyo may represent a facies boundary within the reef.

improved structural understanding provided valuable new perspective to Acknowledgments References

the interpretation of existing seismic data. Collins, J. G., and Lake, J. H., 1989, Sierra reef complex, Middle Devonian,
The authors thank Focus Seismic Corp. for permission to use the Sierra
The discovery that many intrasedimentary faults are magnetized suf- Northeastern British Columbia, in Geldsetzer, H. H. J., James, N. P., Tebbutt, G.
data for this case history, an anonymous company for a permission to
ficiently to be mappable means regional faulting can be mapped at very E., Eds., Reefs: Canada and adjacent areas, Calgary, Alberta: Can. Soc. Petr. Geol.
extract the Sierra portion of a larger interpretation, the Canadian Society Mem. 13, 414–421.
modest cost, allowing seismic programs to be positioned more effectively
of Exploration Geophysicists for permission to use some figures previously
and oriented appropriately. This benefit alone is enough to justify the Ebner, E., Peirce, J., and Marchand, N., 1995, Interpretation of aeromagnetic data:
published (Ebner et al., 1995), and Debi Walker for producing the figures CSEG Recorder, 20, 8–11.
expense of a high-resolution aeromagnetic survey. In addition, the mag-
for this paper.
netic structural-grain map allows one to define exploration fairways and Hassan, H. H., Peirce, J. W., Pearson, W. C., and Pearson, M. J., 1998, Cultural edit-
to identify new exploration leads. ing of HRAM data —comparison of techniques: J. Expl. Geophys. (Scheduled for
Dec.)
Ku, C. C., and Sharp, J. A., 1983, Werner deconvolution for automated magnetic
interpretation and its refinement using Marquardt’s inverse modelling:
Geophysics, 48, 754–774.
MAGPROBE™ is a software product of LCT Software Inc.
04 Plays-Field Studies 4/16/07 4:34 PM Page 101

Jain, S., 1976, An automatic method of depth interpretation of magnetic profiles: General Philosophy 5
Geophysics, 41, 531–545.
edited by David A. Chapin
Peirce, J. W., Abercrombie, H. J., Charters, R. A., De Paoli, G. R., and Goosser, S. A., LaCoste & Romberg
1998, Intrasedimentary magnetization by vertical fluid flow and exotic geochem- Austin, Texas, USA
istry: The Leading Edge, 17, 89–92.
Peters, L. J., 1949, The direct approach to magnetic interpretation and its practical Concerning Nonuniqueness
application: Geophysics, 14, 290–320.
Our interpretations never can be unique,
Phillips, J. D., 1979, ADEPT: A program to estimate depth to magnetic basement from but they can be unambiguous.
sampled magnetic profiles: Open-File Report 79–367, U.S. Geol. Surv.
Phipps, G., 1982, Exploring for dolomitized Slave Point carbonates in northeastern There are models we might like to
British Columbia: J. Can. Soc. Expl. Geophys., 18, 7–13. accept that just don’t fit gravity, mag-
Radke, B. M., and Mathis, R. L., 1980, On the formation and occurrence of saddle netic, or electrical data. This benefit of
dolomite: J. Sedimentary Petrology, 50, 1149–68. modeling is important. It forces the inter-
Rao, V. B., 1984, Discussion on paper by Ku and Sharp, 1983: Geophysics, 49, 1119. preter to prove that an interpretation is
possible, and it eliminates impossible
Read, J. F., 1985, Carbonate platform fault models, AAPG Bull., 69, 1–21.
models—even seismic models. A simple
Reid, A. B., Allsop, J. M., Granser, H., Millett, A. J., and Somerton, I. W., 1990, depth estimate may be all the model one
Magnetic interpretation in three dimensions using Euler deconvolution: needs to eliminate an idea. Other times,
Geophysics, 55, 80–91.
we need a careful and detailed model to
Ross, G. M., Broome, J. and Miles, W., 1994, Potential fields and basement structure, help us understand what’s possible—
in Geological atlas of the western Canada sedimentary basin, G.D. Mossop and I. and unambiguously impossible.
Shetsen, Comps., Can. Soc. Petr. Geol. and Alberta Research Council, 41–48.
Ross, G. M., Parrish, R. R., Villeneuve, M. E., and Bowring, S. A., 1991, Geophysics Alan T. Herring
and the geochronology of the crystalline basement of the Alberta Basin, western EDCON
Canada: Canadian Journal of Earth Sciences, 28, 512–522. Denver, Colorado, USA
Thompson, D. T., 1982, EUDELPH: a new technique for making computer-assisted
depth estimates from magnetic data: Geophysics, 47, 31–37.
Werner, S., 1953, Interpretation of magnetic anomalies as sheetlike bodies, Sver. Geol.
Undersok., Serv. C, Arsbok 43, Vol. 6. It is wearisome to keep hearing the
“nonuniqueness” issue raised as an
excuse not to use potential fields. We
need to get the message out that other
Figure 15. Interpreted structural-grain map overlaid on 1.5- to 7.0-km band pass of data or assumptions provide sufficient
magnetic data (same image as Figure 7). This particular interpretation was based on a constraints on solution space to eliminate
larger and more recent survey with east-west-oriented flight lines. The Xs on each
“nonuniqueness.”
fault indicate locations in which a fault was interpreted on the magnetic interpretation.
Note the Yoyo Magnetic High, which runs south from the Yoyo Reef. The listric fault
shown in Figure 14 is the intrasedimentary fault shown on this interpretation on the E. K. Biegert
east side of the Yoyo magnetic high. This correlates with a poorly defined basement International Center for Remote Sensing
high on which the Yoyo Reef apparently nucleated. and Potential Field Studies
Shell
Houston, Texas, USA

101
Geologic Applications of Gravity and Magnetics: Case Histories
04 Plays-Field Studies 4/16/07 4:34 PM Page 102
This criticism fits almost every seismic
modeling project I see because most geo-
physicists seem to forget that seismic
response is also nonunique. Potential-
field practitioners are generally more
aware of the pitfalls of nonuniqueness,
and therefore, they commonly, but not
always, offer multiple models.
John Peirce
Geophysical Exploration & Development
Corporation
Calgary, Alberta, Canada
I see the nonuniqueness issue as not
being an interpretation question at all,
but rather an important technical issue in
writing inverse algorithms. Without
knowing that you are inverting in some
class of models in which the inversion is
unique, you may find yourself wandering
around in parameter space at the mercy
of round-off error. Nonuniqueness is not
peculiar to potential-fields data. Ask the
people who try to invert offset-versus-
amplitude data for porosity.
Richard Hansen
Pearson, deRidder, & Johnson
Lakewood, Colorado, USA
We are all familiar with the application
of Gauss’s law to estimate the causative
mass of a given gravity “anomaly” (of
course, in practice it is often better to
model a feature rather than compute a
surface integral to make this estimate).
This mass estimate can have direct
importance in answering geologic ques-
tions such as the volume of an ore body
or the size of an intrusion. These kinds
of gravity solutions to geologic problems
are not plagued by the same degree of
“nonuniqueness” that affects some mod-
els of causative body shape.
Rick Saltus
U. S. Geological Survey
Denver, Colorado, USA
I would like to add a few bullet com-
ments we can use to explain to our
clients that the “potential-field solutions
are nonunique” academic statement they
wrote down and memorized in their for-
mative years is the least of their worries.
Upon being told that your model is a
nonunique solution, you might say:
“Well, so is the seismic.”
You probably don’t want to say this
because it will only make clients mad,
either because they won’t understand it
or they will, and won’t want to be
reminded of it.
“Well, so is the (insert the name of their
favorite well log interpretation.)”
Same warning as above.
“Well, I think that what we must con-
sider here is that any model (geologic,
financial, or whatever) is really just a
best fit of all the available information.”
That is, when we use all the bits of data
we have as cross-checks of the unknown
parameters we are solving for, we con-
strain the number of possible solutions to
a point where there may be a single solu-
tion. This is not to say that errors in data
measurements do not permit a range for
that solution. For example, uncertainty
Plays/Field Studies
in velocity limits the accuracy in conver-
sion from depth to time. Errors in near
borehole density readings limit the accu-
racy of shape and size. And of course,
accuracy in measuring the potential field
itself, although this is becoming less of a
problem, will limit the model’s validity.
And the list goes on. Still, even with the
limitations from these sources of error,
there can be only one best fit, and this
is it.
Same warning as above, except go ahead
and say it.
David R. Oxley
Mobil Exploration & Production
Technology Center
Dallas, Texas, USA
It is true that an infinite set of sources
can produce an observed anomaly—the-
oretically. When we put our feet on the
ground and consider the real world, the
possible solutions are certainly finite.
That is, even in a rank unexplored area
with no information except for our
beloved potential-fields data, it’s easy to
constrain these possibilities. We know if
our data were acquired over a foreland
basin or along a passive margin or in a
rift setting, etc.
Our data are not nonunique. We always
have a feeling for what we’re looking for,
and we can interpret them.
Dale Bird
Bird Geophysical
Houston, Texas, USA
102
It is often the case that key end-member I feel that some explorationists mix up
structural styles are not supported by nonuniqueness and resolving power. If I
the data, but that a region of “geologic use magnetics to predict a structure at
variation” space admits interpretational 550 m depth and it turns out to be at
ambiguity. Some may call this a model 350 m, well, that is a poor depth esti-
parameter space, and in physics we call mate. But the structure is there. I did
it a phase space. not miss the structure! Resolution can be
a function of data spacing, data quality,
A key element of any interpretation is to anomaly interference, lack of geologic
identify which structural styles are control, lack of good software, any num-
allowed by the data and which are ber of things.
denied.
Nonuniqueness would apply if I pre-
The region of interpretational ambiguity dicted a structure and it wasn’t there,
(phase space) can be determined for i.e., there was an anomaly but no
most interpretations. Common tools structure, and that anomaly could be
include sensitivity analysis, Monte Carlo interpreted as a structure or not. This
simulations, and ideal body analysis. The almost never happens.
multidimensional space usually is con-
strained by some very simple assump- Pat Millegan
Marathon Oil Company
tions about minimum and maximum
Houston, Texas, USA
densities, minimum and maximum
depths, lateral extent of bodies, etc.
I often have found that the largest source
of interpretational ambiguity is not the
uncertainty associated with the potential-
field data, but the encasing geology, usu-
ally assumed to be known!
E. K. Biegert
International Center for Remote Sensing
and Potential Field Studies
Shell
Houston, Texas, USA
05 Mining and Environmental 4/16/07 4:46 PM Page 103

Gravity and Magnetic Study of the Southern Chocolate Mountains Region,

Southeastern California: Possible Extension of the Mesquite Gold Mine . . . . . . . . . . 105
K. L. Mickus, Department of Geosciences
Southwestern Missouri State University
Acid Mine Drainage in Northern Ontario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
T. Jeffrey Gamey, Aerodat, Inc., Canada
Sulfur Exploration with Core-hole and Surface Gravity . . . . . . . . . . . . . . . . . . . . . . . . . . .113
M. Alexander, Integrated Geophysics Corporation
K. O. Heintz, Electro Mechanical Specialists
An Airborne Multisensor Characterization of an Active Nuclear Waste Site . . . . . . . . . 120
T. Jeffrey Gamey, Aerodat, Inc., Canada
J. Scott Holladay, Vanguard Geophysics
Johnathan Nyquist, William Doll, Oak Ridge National Laboratory

5 Mining
and
Environmental
05 Mining and Environmental 4/16/07 4:46 PM Page 104
General Philosophy 6:
A Different Approach
Gravity and Magnetics
Community
edited by P. S. Millegan, Marathon Oil, USA
Mining and Environmental
Approach
Although gravity and magnetic surveys
for hydrocarbon exploration typically are
designed with relatively wide data spac-
ings, to define relatively deep-seated fea-
tures, environmental and mining projects
are concerned with information from
much nearer the surface. One of the
biggest differences in these two kinds of
surveys is, therefore, line or data spacing,
which is much closer for shallow-target
work. Ground magnetometry commonly
is used instead of airborne magnetics, so
that the sensor can be closer to the small
target anomalies.
Magnetic data still find their greatest
application in mapping buried metallic
objects such as waste drums, but excel-
lent magnetic surveys have been used
archaeologically to define tepee rings,
hearths, and other structures. Similar
approaches may contribute to under-
standing soil and channel distribution in
aquifers and other environmentally inter-
esting situations. An important use has
been finding abandoned steel-cased well
bores, which have serious impact on sub-
surface fluid flow. Gravity data can help
locate buried void spaces and can assist
in the delineation of subsurface features
such as channels. A high-tech supercon-
ducting gravity meter can discern be-
tween a “filled” aquifer and a mostly
empty one.
Mining studies are concerned with min-
eral distributions, and their distinctive
gravity and magnetic signatures can
assist with such interpretation. Tradition-
ally, mining applications (especially using
magnetics) have ranged from frontier
exploration to ore-body delineation.
Tectonic studies also bear on mining-
exploration problems, and for diamond
exploration, kimberlite pipes are often
almost directly detectable by their char-
acteristic bull’s-eye aeromagnetic anom-
alies. In recent years, more magnetic
flying in North America has been in sup-
port of diamond exploration than any
other application.
An excellent reference for geophysical
work of all types in environmental appli-
cations is the SEG monograph Geo-
technical and Environmental Geophysics,
1990 (Investigations in Geophysics No.
5, edited by Stanley H. Ward).
Richard I. Gibson
Gibson Consulting
Golden, Colorado, USA
My own experience is primarily with
precious-metal and copper exploration,
tempered with training in economic geol-
ogy of a wide variety of deposits and in a
variety of geophysical exploration tech-
niques. My consulting business is
focused primarily on use of potential
fields for mineral exploration, but I
highly recommend combining this infor-
mation with Landsat, induced polariza-
tion data, geochemistry, etc., and with
prior experience with remote sensing
and a variety of electrical (and electro-
magnetic) techniques.
Interpretation of gravity and magnetics
data for mineral exploration is indeed
very similar to petroleum exploration.
The primary differences are due to the
Mining and Environmental
size of the final target, the depth that is
economically of interest to the prospector
and mining company, and the level of
detail necessary for a final target before
choosing drill sites that have a high prob-
ability of success or outlining an exten-
sion of an existing deposit. Regional
surveys are of interest for choosing
prospect areas or claim blocks, but suc-
cessful drill sites should not be based on
regional surveys alone.
Regional surveys may be similar to a typi-
cal petroleum exploration survey or a bit
more detailed. Interpretations at the
regional level usually involve:
1) delineating significant structures that
tie to known mineralization or favor-
able combinations of rock types
2) estimating course depth, to rule out
(sterilize) basins that are too deep to
access economically
3) searching for igneous rocks for cer-
tain deposit types that rely on a heat
source to drive hydrothermal fluids
that collect, concentrate, and deposit
minerals
4) visualizing regional tectonic features
controlling geology
It is essential to work with the geologist
and to be as familiar as possible with the
deposit type(s) targeted. I highly recom-
mend touring pits of similar deposits in
the area, if available.
Most deposit types are not economic
today if they can’t be mined via an open
pit. This means that depths less than 91–
152 m (300–500 ft) are of most interest,
and depths below 305 m (1000 ft) are
not worthy of further study, except when
adjacent to shallower features and
needed as background information (the
second half of a profile’s curve).
Underground mines require more
extreme concentrations in smaller total
areas and generally become tiny targets
masked by near-surface variations and
with anomalies that are stretched out by
the distance between the recording
instrument and the target. For these rea-
sons, successful interpretation is tied inti-
mately to tightening the survey spacing
over areas of interest, prospects, and
potential drill sites. On the other hand,
accurate depth estimates require collect-
ing stations off the target that help
describe the gradients and curves associ-
ated with changing density or magnetic
susceptibility adequately. Ties to bedrock
with multiple stations perpendicular to
the structure are necessary for gravity
surveys. Drill information, if available,
helps reduce uncertainty about basin
depths, rock types involved, etc.
Attention to geology and geomorphology
in the field can aid with survey design on
the fly, ensuring that necessary data are
available later for interpretation.
A gravity survey with 91-m (300-ft) spac-
ing between stations, 30 m (100 ft)
between stations at targeted structures or
drill sites, is reasonable for a local
prospect survey. The 152- and 305-m
(500- and 1000-ft) surveys commonly
used can alias a target and provide
annoying one-point anomalies. Quarter-
mile aeromagnetic surveys that tighten to
1
⁄8-mile spacing over prospect areas pro-
vide the level of detail needed for a typi-
cal mineral exploration survey. The extra
money spent on this level of detail for
the geophysics is recovered by more
accurate targeting of drill sites. An explo-
ration geologist commonly has funds for
three holes per new prospect each sea-
son. If those holes are successful, the
104
budget may be increased. If the three ini-
tial holes don’t show adequate mineral-
ization, then funding for exploration the
following season may be nonexistent or,
at the very least, hard to defend. Good
mapping and solid geophysical and geo-
chemical data sets are the mineral explo-
rationist’s best bet toward success with
each prospect.
Variations in target models for different
deposit types are too numerous to list
one by one, but in general, the approach
outlined is how a mineral exploration
geophysicist needs to operate. The best
“final” interpretations combine all exist-
ing information about the prospect. The
goal is to assist the geologist in targeting
drill sites that prove economic mineral-
ization. Related goals may include map-
ping bedrock for nonmineralized areas
adjacent to existing mines, “sterilizing”
these areas for mine facilities, dumps
(dirt piles), etc., or assisting with engi-
neering problems, such as delineating
faults that change rock or sediment engi-
neering properties relating to strength
and stability. A statement that “No, this
area is not economically mineralized”
can be valuable information for an exist-
ing mine with a land crunch.
Phyllis Fett Halvorson
Consulting Geophysicist
Boulder Creek, California, USA
I spent a fair bit of time on the mining
end before getting involved in the petro-
leum business. The main differences that
struck me were:
1) In the mining game, we tended to
spend more time on the fabric of the
magnetic data and less on detailed
modeling of individual anomalies.
05 Mining and Environmental 4/16/07 4:46 PM Page 105
This may be partly because they
were too complex to model profitably
and partly because we knew that
they tended to be aliased, especially
because we didn’t often have the
benefit of data flown with 100-m or
200-m line spacing. We did the occa-
sional excess mass calculation, dike
dip check, or depth to top. If we
could get a rough feel for dip and
depth to top of a body, that was a tri-
umph and we drilled it.
2) The mining mag is a lot more
detailed much of the time because
the sources are often right there, not
buried under 5–10 km of sediment.
It has taken a bit of an effort to come
to terms with gravity anomalies being
sourced shallower than magnetic
anomalies in petroleum work, instead
of the other way around much of the
time in mining work.
3) The status of potential-fields geo-
physics is quite different. In mining,
it’s a vital tool. Most geologists rely
on it and certainly know about it.
The magnetic map is the bread-and-
butter tool of minerals exploration.
4) When you are walking basement out-
crop, the fractal nature of magnetic
fields is much more apparent. You
know if you do a ground profile that
the detail will increase as you
shorten the pole. You also know any
susceptibility measurements you
make are likely to vary chaotically.
So you avoid models that assume
large volumes of constant magnetiza-
tion (which means that you distrust
models).
5) Models remain useful for environ- Gravity and Magnetic Study of the
mental work if you’re looking for
drums and piping. Southern Chocolate Mountains Region,
Southeastern California: Possible
6) You tend to have a more diverse geo- Extension of the Mesquite Gold Mine
physical tool bag in mining. You are
looking at conductivity or resistivity
(EM and contact resistivity), geo-
K. L. Mickus
chemical polarizability (IP), radioac- Department of Geosciences
tivity and geochemistry (gamma Southwest Missouri State University
rays), and not just playing with sus- Springfield, Missouri, USA
ceptibility, remanence, and density.
So you often get several extra (and
all different, but complementary) EDITOR’S NOTE: This paper provides good examples of both the
views of the geology. You get good at detail possible in modeling and the limitations inherent in it.
integrating all these views in discus- Mickus shows excellent integration of multiple geophysical
sion with the geologists. techniques with geology and tectonics. Without the geologic
attention to detail, confident conclusions would not have
7) EM can be incredibly diagnostic,
especially my favorite, time-domain
been possible.
EM with a fixed and large transmit
loop. We got the geologists trained so
that they knew they would hit con- Abstract
ductors where we said they were. Gravity and magnetic data are used to determine the possible location of
And they did. Of course, the conduc- detachment fault-related upper-plate tilt blocks that may contain ore bod-
tors weren’t always rocks you could ies in Jurassic gneisses similar to or associated with the Mesquite gold
sell, but it was a start. mine of southeastern California. Residual gravity and magnetic anomaly
maps show the possible location of several of these tilt blocks buried
8) The perspective was a longer one
beneath Tertiary and Quaternary sedimentary rocks north of the
and the exploration budget was
Chocolate Mountains. Gravity models constructed along profiles north-
mostly a lot tighter. From the deci-
sion to look in an area to a working east of the Mesquite Mine, with detailed gravity data and constrained by
mine often could be more than ten surface density measurements and nearby seismic reflection profiles, indi-
years, and then you might have ten cate that the proposed tilt blocks containing the possible ore bodies in
to twenty years of mine life (or often Jurassic gneiss lie between 0.05 and 0.20 km beneath the earth’s surface.
longer).
Introduction
Dr. Alan B. Reid
Geophysical Exploration Technology (GETECH) Exploration for base- and precious-metal mineral deposits in western
Leeds, U. K. Arizona and southeastern California has exploded since the discovery of
numerous mineral deposits associated with detachment faults in the
highly extended Colorado River extensional corridor (Spencer and Welty,
1986). The mineralization seems to occur during the formation of detach-
ment faults and to be concentrated in the high-angle normal faulted
blocks found on the upper plates of the detachment faults (Spencer and
Welty, 1986). One of the largest gold mines in the region (Mesquite
105
Geologic Applications of Gravity and Magnetics: Case Histories
05 Mining and Environmental 4/16/07 4:46 PM Page 106

Mining and Environmental

106
Mine) lies along the southwestern edge of the highly extended region rocks. The major rock units important to this study will be described (Dillon, 1975). Intruding the Jurassic gneisses are Tertiary plutonic rocks
(Willis and Tosdal, 1992). The complicated geologic and tectonic history briefly and are shown in Figures 1 and 2. of which only the Mount Barrow granodiorite is exposed within the study
of the Mesquite Mine and Chocolate Mountains region has led to various The Mesozoic Orocopia Schist is exposed in the core of the southeast- area (southern end of profile 1 on Figure 2). These plutons have been
geologic interpretations of the area (e.g., Dillon, 1975; Dillon et al., 1990; trending Chocolate Mountains anticlinorium. The Orocopia Schist is con- described as laccoliths or sills (Morris, 1990) and are important in pro-
Duke, 1991) and of the origin of the ore bodies within the Mesquite sidered to represent the deepest basement exposed in southern California viding mineralizing fluids for the region’s ore deposits (Frost, 1990).
Mine. The complicated tectonic history includes convergent tectonic (Dillon et al., 1990) and may be part of a schist terrane (including the Overlying the Jurassic gneisses in and around the Mesquite Mine are
regimes producing the Chocolate Mountain thrust fault (Dillon et al., Pelona, Rand, and Catalina Schists) that underlies most of southern
1990), detachment faulting associated with the formation of metamor- California (from seismic reflection data and field relationships by Jacobson
phic core complexes (Duke, 1991), and strike-slip faulting related to for- et al., 1988). Jurassic gneisses are defined as fine-grained quartzo-felds-
mation of the San Andreas transform fault (Crowell, 1979). The main pathic rocks that are interbedded with a more mafic metaigneous suite
tectonic/geologic questions important to this study are: (1) is there
detachment faulting in and southwest of the Chocolate Mountains, and
(2) what are the structural controls on the formation of the Mesquite
Mine ore bodies?
There are two main ideas as to the structural origin of the ore bodies:
(1) they formed during detachment faulting and the ore bodies lie on the
upper plate of a detachment fault (Frost et al., 1986; Frost and Watowich,
1987), or (2) they formed during strike-slip faulting associated with the
formation of the San Andreas transform fault (Willis, 1988; Willis and
Tosdal, 1992). Detailed structural evidence seems to favor the strike-slip
mode of formation, but the influence of detachment faulting cannot be
ruled out. If the ore bodies have been influenced by detachment faulting
either during or after ore-body formation, then part of the ore body or
additional ore bodies may lie to the northeast of the Mesquite Mine.
Structural information on faults within the Chocolate Mountains indi-
cates that any fault movement (thrust or detachment) has moved upper-
plate rocks toward the northeast (Simpson, 1990). To investigate whether
potential mineralized upper-plate blocks lie to the northeast of the
Mesquite Mine, gravity and magnetic data were analyzed, and
2.5-D gravity models were constructed to try to locate buried blocks
within the alluvium northeast of the Mesquite Mine.

Geology of the Chocolate Mountains Area

and the Mesquite Mine
There have been several tectonic models to explain the existence of the
wide range of structures and rock types found in southeastern California
and southwestern Arizona. These include convergence (e.g., Dickinson,
1981), accretion of microcontinents (e.g., Haxel et al., 1985), and highly
extended detachment faulted regimes (e.g., Duke, 1991). In the study
area, several of the features that led to these tectonic models are dis-
played, including the Chocolate Mountains thrust fault and several
detachment faults. Within the study area, the major lithologic units Figure 1. Location map and generalized geologic map of southeastern California and Figure 2. Detailed geologic map of the southern Chocolate Mountains showing the
include schists, gneisses, granite intrusions, volcanic, and sedimentary southwestern Arizona. Adapted from Sherrod and Tosdal (1991). location of Mesquite Mine and the two profiles to be modeled.
05 Mining and Environmental 4/16/07 4:46 PM Page 107
Miocene to Pliocene clastic sedimentary rocks interbedded with various
basalt layers. More widespread than the Miocene to Pliocene clastic sed-
imentary rocks are Miocene to Oligocene volcanic and volcanoclastic
units that crop out north of the Mesquite Mine (Figures 1 and 2). These
rocks include basaltic to rhyolitic lava flows, rhyodacitic to rhyolitic
domes and tuffs, and interbedded volcanoclastic units (Crowe, 1973).
Overlying all other units is a series of Tertiary to recent sedimentary units,
including breccias, the Bear Canyon Fanglomerate, and recent alluvium.
Structurally, this area is dominated by a series of thrust faults
(Chocolate Mountains thrust fault), high-angle normal faults, detachment
faults, strike-slip faults, and antiforms (including the Chocolate
Mountains anticlinorium). The origin of the thrust faults and the
antiforms was believed to be related to the convergent tectonic setting
that existed from the Mesozoic to Oligocene (?) (Engebretson et al.,
1985). However, in the late Oligocene to early Miocene, large-scale
regional extension occurred to the northeast of the Chocolate Mountains
in the Colorado River extensional corridor (Lister and Davis, 1989), and
it may have extended within and south of the Chocolate Mountains
(Simpson, 1990; Duke, 1991). This regional extensional event formed a
number of metamorphic core complexes where middle-crustal ductile
mylonites were brought near or to the Earth’s surface along low-angle
normal faults (detachment faults). These mylonites represent the lower
plates of the detachment faults, and the upper plates consist of a series of
high-angle normal-faulted blocks (Lister and Davis, 1989). Within the
study area, the Chocolate Mountains thrust fault has been interpreted to
be a regionally important feature where presumably older gneisses are
thrust over the younger Orocopia Schist (Dillon, 1975; Dillon et al.,
1990). However, recent studies by Simpson (1990) and Duke (1991)
have questioned this relationship and have suggested that the thrust fault
in some areas actually may be a detachment fault. If this is the case, the
area of extension may extend southwest at least to the Mesquite Mine
area. Also, interpretation of seismic reflection data in the alluvial basins
north of the Chocolate Mountains and surface geologic data in the
Midway Mountains (Morris and Frost, 1985; Morris, 1990; Duke, 1991)
indicate the presence of detachment faults with upper-plate tilted blocks
(Figure 2), implying that the region of extension definitely exists on the
north side of the Chocolate Mountains. In addition to the normal and
thrust faults, inactive strike-slip faults are found in the Chocolate
Mountains and the Mesquite Mine area. These faults are associated with
the formation of the San Andreas transform fault that formed in Middle
to Late Miocene time (Crowell, 1979). The existence of these faults in
the Chocolate Mountains suggests that these mountains comprise the
eastern edge of the Salton Trough (Willis and Tosdal, 1992).
The Mesquite Mine is located on the southwestern edge of the reduction density of 2.67 gm/cc and a datum plane of mean sea level.
Chocolate Mountains (Figures 1 and 2) in Jurassic quartzo-feldspathic and Terrain corrections were applied to all the newly acquired data and were
mafic gneisses that were intruded by Tertiary granitic rocks (Willis, 1988; available for most of the older data. The merged data then were gridded
Tosdal et al., 1991; Willis and Tosdal, 1992). The ore bodies are not at an interval of 1.0 km and contoured to produce a Bouguer gravity
exposed at the surface but are covered by a thin veneer of Late Tertiary anomaly map (Figure 3).
and Quaternary alluvium. The mineralized regions consist of two subpar- Aeromagnetic data were obtained from the USGS (Mariano and
allel ore bodies that are Oligocene in age and consist mainly of dissemi- Grauch, 1988) for California and from the University of Arizona (Sauck,
nated gold concentrated in northwest- to north-striking faults which have 1972) for Arizona to assist in the interpretation of the gravity data. The
been cut subsequently by a series of northeast-striking low-angle oblique-
dipping faults (Frost and Watowich, 1987; Willis and Holm, 1987; Willis
and Tosdal, 1992). The ore bodies were formed at epithermal conditions
(210°–230°C) at shallow depths (300 m or less) by the circulation of
meteoric water within the fractured gneisses (Frost and Watowich,
1987), where the source of the ore-forming fluids may have been Tertiary
granitic rocks (Frost, 1990). A major argument exists on the type of struc-
tural control on the mineralization. Frost et al. (1986) and Frost and
Watowich (1987) argue that the northwest-striking faults are normal
faults that lie on the upper plate of a detachment fault with the ore-form-
ing fluids being concentrated in the detachment fault. Both authors favor
a detachment-fault model similar to ore deposits to the north in Arizona
and eastern California (e.g., Spencer and Welty, 1986). Recent structural
data by Willis (1988) and Willis and Tosdal (1992) show that the north-
west-striking faults are more likely strike-slip faults related to the San
Andreas system and ore bodies formed during dextral strike-slip faulting.
Even if the strike-slip faulting model proves to be correct, Willis and
Tosdal (1992) show that mineralized, normal faults can be found within
the mining district and these ore bodies could have some relationship to
extensional tectonics. If the ore bodies were formed before or during
detachment fault movement, part of the ore body could have been trans-
ported toward the northeast, as structural work by Simpson (1990)
showed that northeastward movement occured on the fault systems
(thrust or detachment) in the Chocolate Mountains.
Gravity and Magnetic Data
To study the Earth’s gravity field in the vicinity of the Chocolate
Mountains, gravity data were obtained from the Defense Mapping
Agency and the United States Geological Survey (USGS) (Mariano et al.,
1986). To supplement the above data, I collected approximately 300 data
points at a station spacing of approximately 0.7 km, mainly along four
seismic reflection profiles (Duke, 1991), to aid in mapping the detach-
ment faults and any other structures that might be related to mineraliza- Figure 3. Bouguer gravity anomaly map of the Chocolate Mountains area. The contour
tion (Figure 3). All the data were merged and reduced to Bouguer gravity interval is 2 mGal. Dots represent the station locations. Hachured contours represent
anomalies using the 1971 international gravity formula (Morelli, 1976), a relative gravity lows. The star represents the location of the Mesquite Mine.
107
Geologic Applications of Gravity and Magnetics: Case Histories
05 Mining and Environmental 4/16/07 4:46 PM Page 108

Mining and Environmental

108
USGS data were collected in east-west lines every 1.0 km, with north-
south tie lines every 8.0 km at a mean elevation of approximately 0.3 km.
To remove the effect of the magnetic field due to the Earth’s core, the
1981 International Geomagnetic Reference Field (IGRF) was removed
from the data. The University of Arizona’s data were collected in north-
south flight lines every 5 km, with east-west tie lines every degree of lat-
itude at a mean elevation of approximately 0.3 km. The 1966 IGRF was
removed from the data. Because the data sets had different IGRFs
removed from them, they could not be merged unless a constant or lin-
ear transformation was applied to one of the data sets. I determined that
a constant value of –410 nanoteslas (gammas) added to the University of
Arizona data adequately merged the two data sets. The data then were
gridded at an interval of 1.0 km and contoured to produce a total-field
magnetic intensity map (Figure 4).

Gravity and Magnetic Anomalies

A casual glance at the Bouguer gravity and total-field magnetic anomaly
maps (Figures 3 and 4) shows relatively low values over the alluvial basins
and relatively high values over the mountain ranges. A closer examina-
tion of both maps reveals anomalies opposite to the above scenario (e.g.,
the gravity maximum over the alluvium next to the Mesquite Mine). To
obtain a better understanding of the anomalies caused by near-surface fea-
tures, residual anomaly maps can be constructed to remove the regional
anomalies due to deep-seated or long-wavelength features. There are
many techniques (e.g., wavelength filtering, polynomial-trend surface fit-
ting, isostatic gravity residuals) available to remove the regional anom-
alies, but in general, the residual anomalies generated by these techniques
cannot be modeled but only discussed qualitatively because they are
mathematical manipulations that do not exactly remove the regional
anomaly field (Ulyrch, 1968). After trying several different polynomial
Figure 4. Total-field magnetic intensity anomaly map of the Chocolate Mountains area.
trend surfaces and wavelength filters, I determined that a second-order The contour interval is 25 nT. Hachured contours represent relative magnetic lows.
polynomial trend surface produced the best residual gravity and magnetic The star represents the location of the Mesquite Mine.
anomaly maps (Figures 5 and 6, respectively) which reflected the upper- Figure 5. Residual gravity anomaly map of the Chocolate Mountains area obtained by
crustal density and magnetic susceptibility changes. Removing a second- subtracting a second-order polynomial surface. The contour interval is 2 mGal.
The residual gravity and magnetic anomalies (Figures 5 and 6) reflect Hachured contours represent relative gravity lows. Numbers refer to anomalies dis-
order trend surface is justified because the regional anomaly fields are cussed in the text. The star represents the location of the Mesquite Mine.
relatively uncomplicated. Regional gravity and magnetic modeling the tectonic history of the area (Mesozoic thrust faulting, Tertiary volcan-
(Mickus, 1989; Mickus and James, 1991) and seismic crustal models in ism and plutonism, detachment faulting, normal faulting, and strike-slip
southeastern California (Hearn and Clayton, 1986; McCarthy and faulting). An understanding of the interrelationships of these events is nonmafic material which could be buried granites and gneisses similar to
Thompson, 1988; McCarthy et al., 1992) show that the depth to the essential in determining the location of mineral deposits within the study those found beneath surface sedimentary material in the Mesquite Mine
lower crust and upper mantle is generally constant or slightly deepening area. Anomaly 1 (Figure 5) is a roughly circular gravity maximum which area.
from east to west, and most of the lateral geologic changes are in the is one of a series of gravity maxima superimposed on a larger-scale grav- On the residual gravity anomaly map, there is a strong northwest-
upper crust. ity minimum that occurs within the Salton Trough. There is no corre- trending gradient (anomaly 2) that parallels the Chocolate Mountains and
sponding magnetic anomaly (Figure 6) which indicates a near-surface, probably represents the density contrast between the crystalline rocks in
05 Mining and Environmental 4/16/07 4:46 PM Page 109
Figure 6. Residual magnetic intensity anomaly map of the Chocolate Mountains area
obtained by subtracting a second-order polynomial surface. The contour interval is
25 nT. Hachured contours represent relative magnetic lows. Numbers refer to anom-
alies discussed in the text. The star represents the location of the Mesquite Mine.
the Chocolate Mountains and the unconsolidated sediments within the
Salton Trough. There is no corresponding magnetic gradient (Figure 6),
and this suggests that the Tertiary volcanic material, Jurassic gneisses, and
the Orocopia Schist are relatively nonmagnetic. Magnetic susceptibility
measurements of surface samples (R. Simpson, written communication,
1987) support this observation.
Northeast of the Chocolate Mountains are a number of gravity and
magnetic maxima and minima. Anomaly 3 (Figures 5 and 6) is a mini-
mum value probably caused by the combination of unconsolidated sedi-
ments and granitic intrusions (Mount Barrow granodiorite). Granitic
intrusions are a common cause for gravity minima in southeastern
California (Mickus and James, 1991). Anomaly 4 occurs over alluvium,
and based on gravity models shown below, probably represents a tilted
block on the upper plate of a detachment fault. In fact, anomalies 6, 7,
and 9 also are interpreted to be associated with similar tilted blocks, and
their locations agree with blocks imaged on seismic reflection profiles
(Duke, 1991). The lack of definite positive anomalies near anomaly 7
(Figures 5 and 6) is probably due to insufficient data (Figure 3). Before the
addition of new gravity data, anomalies 4 and 9 were not observed. The
lack of corresponding magnetic anomalies to some of these gravity anom-
alies suggests that the Tertiary volcanic material in this region is relatively
nonmagnetic. Above 33°15´N are a number of high-amplitude, north-
west-trending magnetic anomalies, where most of the anomalies are asso-
ciated with more magnetic Tertiary volcanic material and magnetite in
skarn deposits related to Mesozoic granitic intrusions (Mickus, 1989).
Also associated with the tilt blocks are gravity and magnetic minima
caused by alluvial basins between the tilt blocks, as at anomaly 8 (Figures
5 and 6).
Anomaly 5 (Figure 5 and 6) is a gravity and magnetic low which
occurs mainly over an alluvial basin but also over the southern Chocolate
Mountains. This suggests that part of the anomaly may be associated with
granitic intrusions underneath the surface Tertiary volcanic material.
Modeling and Discussion
The above qualitative discussion illustrates the general relationships
between the surface geologic features and the observed gravity and mag-
netic anomalies in the vicinity of the Chocolate Mountains. To better
understand the geometries and the densities of the causative bodies,
2.5-D models were constructed using the technique of Lai (1984). The
modeling program calculates the gravitational attraction at each observa-
tion point due to polygonally shaped bodies, with each body having a spe-
cific density value.
Profiles 1 and 2 (Figure 2) were selected to illustrate the subsurface
location of proposed detachment faults and subsurface tilt blocks lying on
the upper plate of the detachment faults. The models were generalized
into four bodies: (1) Quaternary-Holocene alluvial sediments; (2) Tertiary
volcanic material; (3) Jurassic gneiss; (4) Orocopia Schist; and (5) ductile
middle crust (?) or mylonite. These initial body geometries were based on
the mapped geology (e.g., Dillon, 1975; Frost et al., 1986; Frost and
Watowich, 1987) and nearby seismic reflection profiles (Morris and Frost,
1985; Duke, 1991). The subsurface geology is probably more compli-
Geologic Applications of Gravity and Magnetics: Case Histories
cated than shown in the models, but gravity or magnetic data with this
distribution cannot be used to model small-scale geologic features. The
simplified geologic models can explain adequately the observed anom-
alies. The initial densities were determined using surface rock samples
from the Chocolate Mountains vicinity (R. Simpson, written communica-
tion, 1987; Oppenheimer, 1986) (Table 1). A final model was obtained by
modifying the initial geometries and physical properties until the pre-
dicted gravity anomaly matched the observed gravity anomaly. The final
models are shown in Figures 7 and 8.
The use of a 2.5-D model is justified because the modeled geologic
units extend more than 100% of the depth distance perpendicular to the
profile direction (strike direction), so 3-D effects will be minimal. The
only exceptions could be alluvium and Tertiary volcanic material, because
both these units thin in both strike directions away from the profiles, but
the thickness of these units is small and the 3-D effect is minimal.
Table 1. Densities of geologic units used in models 1 and 2. The values
were obtained by averaging densities measured from surface rock sam-
ples (Simpson, written communication; Oppenheimer, 1986) in the vicin-
ity of the Chocolate Mountains.
Rock Type No. of Samples Average Density (g/cm3)
Tertiary volcanic rocks 16 2.38
Jurassic gneiss 28 2.62
Orocopia Schist 49 2.68
Alluvium 323 2.12
Model 1 (Figure 7) shows one detachment fault that transported
Tertiary volcanic material, Jurassic gneiss, and the Orocopia Schist toward
the northeast. The lower-plate mylonite which is exposed in the
Chocolate Mountains (Berg et al., 1982) represents possible ductile mid-
dle-crustal material that may have been brought to the surface during
Tertiary extension. This mylonite is similar to mylonites exposed to the
north in the Whipple and Rawhide Mountains (Lister and Davis, 1989).
Gravity data cannot confirm that the proposed lower-plate rocks are
mylonite because other lithologies such as the Mount Barrow granodior-
ite may compose part of the body (Figure 7). Seismic reflection data north
of the Chocolate Mountains do not image reflections in the lower-plate
rocks, which may suggest the presence of mylonitic rocks similar to those
indicated by seismic reflection data in the Whipple Mountains area
(Pridmore and Frost, 1992). So without drilling, the exact nature of the
109
05 Mining and Environmental 4/16/07 4:46 PM Page 110
Figure 7. Two-and-one-half-dimensional gravity model along profile 1. The number
underneath the rock name is its density in gm/cc. There is no vertical exaggeration. The
bold number 1 refers to the location of upper-plate tilt blocks discussed in the text.
lower-plate rocks probably cannot be determined in this region from seis-
mic, gravity, or magnetic data. The northeast end of the profile shows the
tilted block that corresponds to anomaly 4 on Figure 5. The Jurassic
gneiss which is the ore host rock in the Mesquite Mine is modeled to be
0.05 km below the surface and is certainly within a depth to be drilled
economically.
Model 2 (Figure 8) shows a model similar to model 1 (Figure 7), but
its greater length shows at least one other detachment fault (?) and pos-
sibly two other tilted blocks. The tilt block on model 1 is the same tilt
block shown on Figure 8, and this block accounts for the 8-mGal gravity
maximum. A second detachment that starts at approximately 9 km on
Figure 8 corresponds roughly to the proposed location of a subsurface
detachment on Figure 2. There is not a seismic reflection profile along this
Mining and Environmental
Figure 8. Two-and-one-half-dimensional gravity model along profile 2.The number
underneath the rock name is its density in gm/cc. The middle figure has no vertical
exaggeration. The bold numbers 1 and 2 refer to the locations of upper-plate tilt blocks
discussed in the text. TV = Tertiary volcanic material.
model, but a profile does intersect model 2 at 14 km. The depth to
Tertiary volcanic material, Jurassic gneiss, and the Orocopia Schist deter-
mined from the gravity modeling agrees with the seismic reflection data
(Duke, 1991). Based on gravity modeling alone, the second detachment
could not have been inferred, and uplift 2 (Figure 8) could be a tilt block
formed by high-angle normal faulting. On most metamorphic core com-
plexes, the upper plate consists of a series of small normal fault blocks
(Lister and Davis, 1989), but the seismic reflection data show a fault
extending at least 10 km into the earth, which implies a detachment
fault.
110
The depth to the first tilt block is 0.15 km on model 2 and 0.05 km
on model 1. The longer profile with the inclusion of additional bodies on
model 2 made the modeling more difficult. To fit any individual maxi-
mum or minimum, the geometry or density (or both) of a number of
given bodies had to be changed. The smaller number of bodies on model
1 made it relatively easy to achieve a good fit between the calculated and
observed gravity values. However, by modeling profile 2, one could
obtain a better estimation of the range of geometries and densities that
could be used in obtaining a good fit. The geometries and depths to bod-
ies shown in Figure 8 could be changed by 20% and the same predicted
anomaly curve could be obtained. In fact, the 0.05-km depth to the top
of the tilt block shown in model 1 could not be obtained in model 2. The
0.05-km depth then must be assumed to be a minimum depth to the top
of the block, with 0.20 km being a maximum depth.
Tilt block 1 is thought to have formed during mid-Tertiary extension
that created the detachment faults in and north of the Chocolate
Mountains. If the extension occurred as far southwest as the Mesquite
Mine, then part of the ore bodies containing gold may have transported
to the northeast. The ore bodies associated with the Jurassic gneisses may
be encountered at depths between 0.05 and 0.20 km below the earth’s
surface. These models do not distinguish between the ore’s mode of ori-
gin (strike-slip or detachment), but suggest that if detachment faulting
occurred in the Mesquite Mine region, part of the ore body or similar
types of ore bodies may be encountered in upper-plate blocks to the north-
east.
Conclusions
Gravity and magnetic data in the form of residual anomaly maps indicate
the general location of possible tilt blocks that reside on the upper plates
of detachment faults formed during mid-Tertiary extension in southeast
California and western Arizona. These blocks, especially those to the
northeast of the Mesquite Mine, may contain similar ore deposits or part
of the Mesquite ore body (assuming that detachment faulting occurred in
the Mesquite Mine region). Structural data indicate that transport is
toward the northeast along detachment faults (?) in the Chocolate
Mountains, but no surface evidence has been found for detachment fault-
ing in the Mesquite Mine area. Gravity modeling of detailed data along
two profiles trending northeast from the Mesquite Mine area, constrained
by surface density measurements and seismic reflection profiles, indicates
that a tilt block lies between 0.05 and 0.20 km beneath the Earth’s sur-
face approximately 10 km northeast of the Mesquite Mine. This tilt block
contains Jurassic gneiss, which is the ore host rock in the Mesquite Mine,
and may contain similar ore if the proposed tectonic regime is correct.
05 Mining and Environmental 4/16/07 4:46 PM Page 111
Acknowledgments
I would like to thank Eric Frost, who suggested this project and provided
the initial ideas on possible models. I also would like to thank Pat
Millegan and Dick Gibson for their encouragement throughout the
review of the paper.
References
Berg, L., Leveille, G., and Geis, P., 1982, Mid-Tertiary faulting and manganese mineral-
ization in the Midway Mountains, Imperial County, California, in Frost, E., and
Martin, D., Eds., Mesozoic-Cenozoic tectonic evolution of the Colorado River
region, California, Arizona and Nevada: Cordilleran Publ., 299–315.
Crowe, B. M., 1973, Cenozoic volcanic geology of the southeastern Chocolate
Mountains, California: Ph.D. diss., Univ. of Calif. at Santa Barbara.
Crowell, J., 1979, The San Andreas fault system through time: Geol. Soc. London J.,
136, 293–302.
Dickinson, W., 1981, Plate tectonic evolution of the southern Cordillera, in Dickinson,
W., and Payne, W., Eds., Relations of tectonics to ore deposits in the southern
Cordillera: Ariz. Geol. Soc. Digest, 14, 113–135.
Dillon, J. T., 1975, Geology of the Chocolate and Cargo Muchacho Mountains, south-
eastern California: Ph.D. diss., Univ. of Calif. at Santa Barbara.
Dillon, J. T., Haxel, G. B., and Tosdal, R. M, 1990, Structural evidence for northwest-
ward movement on the Chocolate Mountains thrust, southeasternmost California:
J. Geophy. Res., 95, 1953–1971.
Duke, M., 1991, Styles of extensional deformation and associated detachment faults in
the Chocolate-Orocopia Mountains, and their relationship to the formation of tilted
crustal slabs in southern California: M.S. thesis, San Diego State Univ.
Engebretson, D. C., Cox., A., and Gordon, R., 1985, Relative motions between oceanic
and continental plates in the Pacific basin: Geol. Soc. Am. Spec. Paper 206.
Frost, D., 1990, Lead isotope information contained within gold and other ore minerals
from the Mesquite gold deposit, southeastern California: Geol. Soc. Am. Abstr. with
Prog., 21, 81.
Frost, E. G., Drobeck, P., and Hillemeyer, B., 1986, Geologic setting of gold and silver
mineralization in southeastern California and southwestern Arizona: Geol. Soc.
Am. Guideb., 70–119.
Frost, E. G., and Watowich, S. N., 1987, The Mesquite and Picacho Gold Mines:
Epithermal mineralization localized within Tertiary extensional deformation, in
Davis, G.H., and VandenDolder, E. M., Eds., Geologic diversity of Arizona and its
margins: Excursions to choice areas: Ariz. Bur. of Geol. and Mining Tech. Spec.
Paper 5, 324–336.
Haxel, G., Tosdal., R., and Dillon, J., 1985, Tectonic setting and lithology of the
Winterhaven Formation, a new Mesozoic stratigraphic unit in southeasternmost
California: U.S. Geol. Sur. Bull. 1599.
Hearn, T. M., and Clayton, R. W., 1986, Lateral velocity variations in southern
California, II, Results for the lower crust from Pn waves: Bull. Seis. Soc. Am., 76,
510–520.
Jacobson, C. E., Dawson, M. R., and Prosethwaite, C. E., 1988, Structure, metamor- Spencer, J., and Welty, J., 1986, Possible controls of base- and precious-metal mineral-
phism and tectonic significance of the Pelona, Orocopia and Rand Schists, southern ization associated with Tertiary detachment faults in the lower Colorado River
California, in Ernst, W. G., Ed., Metamorphism and crustal evolution of the west- trough, Arizona and California: Geology, 14, 195–198.
ern United States (Rubey Vol. 7), Prentice Hall, Inc., 976–997.
Tosdal, R. M, Willis, G. F., Manske, S. L., Lang, D., and Lusk, M., 1991, Mesquite min-
Lai, S. F, 1984, Generalized linear inversion of two and one-half dimensional gravity and ing district, southeastern California, in M. McKibben, Ed., The diversity of mineral
magnetic anomalies: Ph.D. diss., Univ. of Texas at Dallas. and energy resources of southern California: Soc. Econ. Geol. Guideb. Series 12,
87–103.
Lister, G., and Davis, G., 1989, The origin of metamorphic core complexes and detach-
ment faults formed during Tertiary continental extension in the northern Colorado Ulrych, T. J., 1968, Effects of wavelength filtering on the shape of the residual anomaly:
River region, USA: J. Structural Geol., 11, 65–94. Geophysics, 33, 1015–1018.
Mariano, John, and Grauch, V. J. S., 1988, Aeromagnetic maps of the Colorado River Willis, G. F., 1988, Geology of the Big Chief orebody, Mesquite district, Imperial County,
region, including the Kingman, Needles, Salton Sea and El Centro quadrangles, California: Soc. Mining Eng. Preprint 88–16.
California, Arizona and Nevada: U.S. Geol. Survey Map MF-2023.
Willis, G. F., and Holm, V. T., 1987, Geology and mineralization of the Mesquite open-
Mariano, John, Helferty, M. G., and Gage, T. B., 1986, Bouguer and isostatic residual pit gold mine, in Johnson, K., Ed., Bulk mineable, Guidebook for field trips: Geol.
gravity maps of the Colorado River region, including the Kingman, Needles, Salton Soc. Nevada, 52–56.
Sea and El Centro quadrangles: U.S. Geol. Survey Open-File Rept. 86–347.
Willis, G. F., and Tosdal, R. M., 1992, Formation of gold veins and breccias during dex-
McCarthy, J., Criley, E., Benz, H., and Ammon, C., 1992, Crustal structure of the south- tral strike-slip faulting in the Mesquite mining district, southeastern California:
ern Basin and Range Province: Preliminary results form the PACE 1992 refraction Econ. Geol., 87, 2002–2022.
experiment in southeastern California: EOS, Trans. Am. Geophys. Union, 73, 370.
McCarthy, J., and Thompson, G. A., 1988, Seismic imaging of extended crust with
emphasis on the western United States: Geol. Soc. Am. Bull., 100, 1361–1374.
Mickus, K. L., 1989, Backus and Gilbert inversion of two and one-half dimensional grav-
ity and magnetic anomalies and crustal structure studies in western Arizona and the
eastern Mojave Desert, California: Ph.D. thesis, Univ. of Texas at El Paso.
Mickus, K. L., and James, W. C., 1991, Regional gravity studies in southeastern California,
western Arizona, and southern Nevada: J. Geophys. Res., 96, 12333–12350.
Morelli, C., 1976, Modern standards for gravity surveys: Geophysics, 41, 1051.
Morris, R. S., 1990, Crustal structure of southeasternmost California: Multiple defor-
mational events evident in industry seismic reflection profiles from the Milpitas
Wash area, Chocolate Mountains: M.S. thesis, San Diego State Univ.
Morris, R. S., and Frost, E. G., 1985, Geometry of mid-Tertiary deformation overprint of
Chocolate Mountains thrust system from seismic reflection profiles in the Milpitas
Wash area of southeastern California: EOS, Trans. Am. Geophys. Union, 66, 978.
Oppenheimer, J. M., 1980, Gravity modelling of alluvial basins, southern Arizona: M.S.
thesis, Univ. of Arizona.
Pridmore, C., and Frost, E., 1992, Detachment faults: California’s extended past: Calif.
Geol., 45, 3–17.
Sauck, W. A., 1972, Compilation and preliminary interpretation of Arizona aeromag-
netic map: Ph.D. diss., Univ. of Arizona.
Sherrod, D., and Tosdal, R. M., 1991, Geologic setting and Tertiary structural evolution
of southwestern Arizona and southeastern California: J. Geophys. Res., 96,
12407–12423.
Simpson, C., 1990, Microstructural evidence for northeastward movement on the
Chocolate Mountains Fault zone, southeastern California: J. Geophys. Res., 95,
529–537.
111
Geologic Applications of Gravity and Magnetics: Case Histories
05 Mining and Environmental 4/16/07 4:46 PM Page 112

Mining and Environmental

112
Acid Mine Drainage in Northern Ontario
T. Jeffrey Gamey
Aerodat, Inc., Canada
Mississauga, Ontario, Canada

EDITOR’S NOTE: Gamey’s paper is an instructional discussion of

an approach to the “largest single environmental problem
facing the Canadian mining industry.” Geophysical surveys
prove to be powerful noninvasive tools for identifying prob-
lems associated with acid mine drainage.

The effects of mine tailings on local groundwater have been recognized

for decades. The oxidation of newly exposed mine tailings, particularly
sulphide minerals such as pyrite, releases acids and heavy metals into the
environment. The rate of acid production is far higher than that caused
by natural leaching and erosion of solid rock, and is beyond the environ-
mentís capacity to absorb. As a result, acid mine drainage has been iden-
tified as the largest single environmental problem facing the Canadian
mining industry (Filion et al., 1990).
Recent mining practices have reduced the amount of waste rock in
tailings ponds through activities such as backfill and reprocessing, and
improved site engineering has reduced the flow of contaminants beyond
the confinement area. Federal restrictions on groundwater acidity and
heavy-metal concentration levels regulate conditions in the immediate
vicinity of the mine, and Ontario mines have a respectable track record
for compliance (Environment Canada, 1988). Abandoned mines, how-
ever, generally date to an era before environmental regulation, and thus
may pose a considerable risk.
The high acid and heavy-metal concentration in the leachate make
the groundwater extremely conductive and an excellent electromagnetic
geophysical target. The example shown has been extracted from a larger
mineral exploration survey conducted for the Ontario Geological Survey.
The survey was flown with helicopter electromagnetics (HEM)/magne-
tometer/VLF-EM sensors at a nominal terrain clearance of 30 m (HEM)
and 45 m (magnetometer and VLF-EM) and a 200-m line spacing.
Observations were positioned by radar ranging techniques, and results
were presented originally as EM profiles at 1:20,000. The actual location
and orientation of the data have been disguised.
The area covers an abandoned mine site with open pit, waste rock
piles, mill, and tailings pond. During the milling process, specific metals
Figure 1. Apparent resistivity in ohm-m.
were extracted from the sulphide minerals in the host rock. Grinding the
rock is an essential step in the process but has numerous undesirable side
effects. The first is that the volume of debris will more than triple. The
second is that it exposes the residual sulphide minerals to accelerated
weathering conditions, producing large quantities of acidic tailings. These
tailings are treated and deposited into a man-made lake or tailings pond
to settle out.
Although every precaution is taken to engineer an adequate disposal
site, continuous monitoring of groundwater seepage is required. Wells are
drilled at strategic locations around the tailings, and samples are taken at
intervals to monitor compliance with federal and provincial regulations.
Knowledge of local hydrographic conditions is essential to ensure ade-
quate containment. The depth to the water table and direction of ground-
water travel are especially important, as most acid mine drainage occurs
in a band between 3 and 10 m in depth and can be traced for hundreds
or even thousands of meters (King and Sartorelli, 1991).
Groundwater mapping by drilling can be an expensive prospect, and
although the direction of flow usually conforms to the surface topography,
Figure 2. Apparent resistivity in ohm-m.
this is not always the case. Airborne electromagnetic surveys of the type
used for minerals exploration can be used to map groundwater contami-
nants very effectively. Figures 1 through 3 show the apparent resistivity
in the vicinity of the tailings pond at three distinct frequencies. Lower fre-
quencies obtain readings from successively greater depths. Typically, the
frequencies of 935, 4175, and 32 000 Hz shown here would penetrate
to 100, 50, and 20 m respectively. Shallow features still contribute to low-
frequency readings, but generally are masked by the larger volumes at
depth. The plume from the tailings pond in the center of the figures is vis-
ible on all three frequencies. This is not because it has great depth extent,
but because it is so conductive that it successfully masks the nature of the
rock beneath it by absorbing all of the transmitted energy.
The most conductive areas represent the highest concentrations of
contamination. In this case, it lies directly over the tailings pond, where
solid tailings contribute the largest part of the signal. The conductive halo
around the boundary of the pond represents the extent of the groundwa-
ter contamination. The majority of the plume extends to the southwest,
where conductivity eventually drops off. At the lowest frequency, the con-
05 Mining and Environmental 4/16/07 4:46 PM Page 113
Each EM reading tends to average data over a larger area than a drill
hole, and so is less susceptible to random errors caused by small-scale geo-
logic variability or to short-term fluctuations brought on by changes in
precipitation. The second advantage is the ability to expand the area of
coverage. For example, periodic mapping of the area would show if the
contaminant plume is in danger of advancing on the open lake some
500 m farther on.
Multifrequency EM systems also can produce depth measurements,
although readings below tailings are masked by strongly conductive
groundwater. In many cases, the exploration database which led to the
original establishment of the mine includes an airborne geophysical sur-
vey of the kind shown here. If such a survey exists, it can be used as an
excellent baseline study of the pristine area, and subsequent airborne sur-
veys can be used to map changes during the life of the site. With the
exception of towers, headframes, and similar buildings which restrict air-
space, airborne surveys are also independent of changes in ground access
or usage, making them practical long-term monitoring tools.
References
Filion, M. P., Sirois, L. L., and Ferguson, K., 1990, Acid mine drainage in Canada: CIM
Bull., December.
Environment Canada—Mining, Mineral and Metallurgical Processes Division, 1988,
Status report on water pollution control in the Canadian metal mining industry
Figure 3. Apparent resistivity in ohm-m. (1986).
King, A., and Sartorelli, A. N., 1991, Mapping acidified groundwater using surface geo-
physical methods: Presentation to second international conference on the abate-
trast between the resistive bedrock and the conductive plume is very dis- ment of acidic drainage in Montreal.
tinct. The light overburden cover can be seen as intermediate conductors
Paterson, N., and Stanton-Gray, R., CANMET, 1995, Applications of remote sensing and
on the higher frequency data. The extent of the plume also appears larger geophysics to the detection and monitoring of acid mine drainage, Proceedings of
at higher frequencies because of the predominance of the plume at shal- Sudbury ’95—Mining and Environment, 13, 955–967.
lower depths of investigation.
The stark linear features are rail and power lines. A small isolated high
can be seen northeast of the tailings pond. This corresponds to a pond
immediately downstream from the tailings dam where heavy metals accu-
mulate in the sediments, presumably in acidic waters. Surface water from
the tailings therefore drains to the northeast, while groundwater less than
10 m below the surface appears to flow “uphill,” with respect to surface
topography, toward the southwest.
Geophysical surveys are recognized as powerful, noninvasive tools for
identifying potential leachate problems around mine sites for inventory
purposes, and for mapping the extent of plumes to guide drilling and fol-
low-up (Patterson and Stanton-Gray, 1995). They cannot replace ground
samples for quantitative measurements of concentrations, but they do
have numerous advantages. The first is the quantity of data collected.
Geologic Applications of Gravity and Magnetics: Case Histories
Sulfur Exploration with Core-hole and
Surface Gravity
M. Alexander, consultant
Integrated Geophysics Corporation
(formerly with Exxon Co. USA)
Houston, Texas, USA
K. O. Heintz*
Electro Mechanical Specialists
(formerly with Exxon Production Research Company)
Houston, Texas, USA
* Deceased
EDITOR’S NOTE: No matter what the target, this paper has some
excellent examples of the effective application of gravity data.
The study uses real, measured subsurface densities and gives
a fair assessment of the advantages and disadvantages of
that information.
Abstract
Announcement of a major discovery near Orla, Texas, in 1968 set off an
extensive sulfur exploration play in the Permian Basin. The terrane, geologic
setting, and mode of sulfur deposition were favorable for use of the gravity
exploration method, and it became the most widely used and cost-effective
geophysical technique for the play. Prospects commonly were located
through use of reconnaissance coverage, then detailed gravity surveys were
conducted to delineate the most favorable sites for initial core drilling.
An important procedure in the exploration program designed by
Exxon Co. USA was the analysis of prospective gravity anomalies through
model studies, using measured subsurface densities for control. Exxon
Production Research Company developed a modified borehole gravimeter
which successfully metered twenty-eight core holes. Densities calculated
from the core-hole gravity were judged to be more accurate and consis-
tent than those derived from other sources.
The core-hole gravimeter was a technical success. Whether continued
use might have led to a commercial discovery is uncertain. An abrupt
plunge in sulfur prices on the world market ended Exxon’s exploration
play before several promising anomalies were evaluated fully.
113
05 Mining and Environmental 4/16/07 4:46 PM Page 114

Mining and Environmental

Introduction
An extensive sulfur exploration play in the Permian Basin was touched off
in 1968 when Duval Corporation announced a major sulfur discovery
near Orla in Culberson County, Texas. Shallow sulfur occurrences had
been noted in west Texas as early as 1854, and several open-pit mines
operated intermittently during the early 1900s. By the late 1960s, rising
world demand and improved production technology made sulfur explo-
ration commercially attractive. However, the economic potential of west
Texas–type deposits was not appreciated fully until the magnitude of the
Orla deposit finally was recognized: announced reserves of more than
80 million long tons, with local concentrations in excess of 40 000 tons
per acre.
Industry employed a variety of remote sensing or surface prospecting
methods in advance of the core drill. The more commonly used were
photogeology, surface geology, geochemical soil analysis, gravity, surface
resistivity, and induced polarization. Of these, a combination of photoge-
ologic and gravity methods appears to have been adopted widely.
Selection of these two methods may have been influenced by features of
the Orla site, which is characterized by both a prominent photogeologic
drainage anomaly and a local Bouguer gravity minimum (see Figures 1
and 2).
All significant west Texas sulfur deposits discovered in the late 1960s
and early 1970s are identified closely with gravity minima. The associa-
tion is theoretically sound on the basis of the geologic assumptions made.
This paper briefly describes gravity methods used by Exxon Co. USA dur-
ing the active phase of its sulfur exploration program.
3) Castile—characteristically a dense anhydrite containing limestone
“castiles”
Figure 4 illustrates a generalized concept of the geochemistry of
native sulfur generation and accumulation within a near-surface fracture
zone. Oxidation of hydrogen sulfide gases and/or reduction of sulfates
from anhydrites or gypsum are believed to be vital to the process.
Whether the fracture zones were generated by deep-seated tectonic
movement or by collapse due to local salt solution in the Salado, the
porosity zones normally are quite restricted laterally. Sulfur prospects are
measured areally in terms of a few hundred acres. Vertical extent and
richness of the ore body are therefore critical economic factors.
Use of the Gravity Method
No geophysical method is known to detect directly the presence of native
sulfur deposits. However, induced polarization, surface resistivity, mag-
netic, gravity, and seismic surveys each could be used to some extent in
the various phases of prospecting. Gravity became accepted as the pri-
mary, most cost-effective method for the west Texas play; the following
114
factors were particularly favorable for its use:
1) Good lateral density contrasts. The country rock is normally dense,
ranging from 2.50 gm/cc (limestones) to 3.00 gm/cc (anhydrites).
Pure sulfur has a density of approximately 2.00 gm/cc; even if fractures
in the country rock were filled solidly with sulfur, the fracture zone
would exhibit a density significantly lower than that of the host rock.
2) Measurable gravity anomalies. The prospective ore bodies are often
less than 457 m (1500 ft) deep and are restricted areally. Density con-
trasts range up to 0.50 gm/cc (see above). Gravity modeling demon-
strates that such bodies generate sharp, short-wavelength minima of
sufficient amplitude to be detectable with conventional survey instru-
ments and techniques.
3) Anomaly recognition. If the survey coverage were sufficiently dense,
the intense, short-wavelength minimum caused by a typical ore body
would be recognizable and have characteristics suitable for various
anomaly separation and enhancement techniques.
4) Logistics and economy. Vegetation is sparse and the road network lim-
ited in this semiarid plains area. Gravity survey operations there are

Local Stratigraphy and Sulfur Occurrence

Some background information regarding local stratigraphy and modes of
sulfur occurrence serves to provide a better appreciation of the suitability
of gravity as a sulfur-prospecting method.
The Upper Permian beds exposed in Culberson and Reeves Counties,
Texas, form a northeasterly-dipping homocline with occasional gentle
folds and faults of small vertical displacement (Figure 3). Sediments of the
Ochoa series were deposited in a restricted sea environment and are pre-
dominantly carbonate, sulfate, and salt formations. Sulfur deposition is
related primarily to, but not restricted to, the following members of the
Ochoa series:
1) Rustler—usually limestone or dolomite, with some sands and/or
gypsum
2) Salado—usually salt or limestone, sometimes anhydrite or gypsum Figure 1. Landsat image of west Texas–New Mexico area with approximate outline of Figure 2. Bouguer gravity at the Orla sulfur deposite site. Survey is reconnaissance
Permian (Delaware) Basin. Orla site indicated by arrow. quality; note relatively sparse station locations.
05 Mining and Environmental 4/16/07 4:46 PM Page 115
Figure 3. Generalized geologic cross-section, Delaware Basin, Texas (after Yarborough
et al., unpublished Exxon report).
fast and highly mobile, permitting rapid, detailed coverage of small,
widely scattered tracts. Personnel and equipment requirements are
modest, and environmental impact is minimal.
5) Empiricism. All significant Permian Basin sulfur deposits reported during
the 1968–1970 play are associated with recognizable gravity minima.
Geologic conditions other than those directly related to sulfur accu-
mulation also generate gravity anomalies with similar amplitude and
wavelength character. Attractive-looking, shallow-sourced anomalies
could be generated by barren fracture zones or by compact deposits of
alluvium, gravel, or gypsum. Even at Orla, less than half the gravity anom-
aly is directly attributable to the known ore body itself. Therefore, proper
interpretation of a prospective anomaly requires that the range of reason-
able solutions be constrained by use of all available structural and forma-
tion density data.
Figure 4. Diagram of near-surface Upper Permian formations and generalized concept Figure 5. Bouguer gravity mapped from reconnaissance survey at Prospect A.
of sulfur generation and accumulation. Note density contrasts (after McCreight et al.,
unpublished Exxon report).
Surface Gravity Coverage fur ore body to the east. Even with the coarse station spacing, the
larger anomaly appeared attractive, but the smaller anomaly could
General play areas were established on the basis of surface geology anom-
have been overlooked readily and the ore body missed.
alies, reported sulfur shows, and suspected mineralization trends. A two-
stage gravity survey or data acquisition program provided Exxon with an 2) Once the most prospective anomalies were located, they were
efficient approach to rapid discovery and development of local prospects: defined more closely by detailed surveys having a station density of
25–50 stations per square mile, normally on a 268-m (880-ft) grid.
1) First, regional trends and individual prospect outlines were defined by
Gravity measurements, including terrain corrections out to the
existing coverage or new reconnaissance gravity surveys. Station den-
Hammer “D” ring (Hammer, 1939), were specified to have an accu-
sity of at least one or two stations per square mile and gravity accu-
racy of 0.1 mGal or better.
racy to within 0.5 mGal were basic requirements. Although a
Figure 6a illustrates the Bouguer gravity mapped from a detailed
compilation of surveys with uneven data quality was acceptable at
survey over the same Prospect A discussed above. The additional con-
this stage, it was recognized that the higher quality surveys would
trol permitted mapping of individual minimum closures and several
provide valuable base networks for future detailed work.
lobes in the broad minimum area to the west. Note that the minimum
Figure 5 illustrates the Bouguer gravity mapped from reconnais-
identified with the sulfur body is defined more clearly now.
sance coverage over Prospect A. Note the contrast in amplitude and
areal size between the large minimum nose in the western portion of Zones of steep contour gradient also are defined more sharply by the
the map and the subtle minimum nose associated with a known sul- additional coverage. Steep gravity gradient zones often are regarded as
115
Geologic Applications of Gravity and Magnetics: Case Histories
05 Mining and Environmental 4/16/07 4:46 PM Page 116
Figure 6a. Bouguer gravity mapped from detailed survey at Prospect A.
significant indicators of contacts between bodies of laterally varying den-
sity contrast (e.g., faults zones, ore-body edges, etc.). Steep gradient zones
in the prospect areas were interpreted as edges of the porosity or alter-
ation zones and were used to estimate the effective areal limits of miner-
alization. In some graben-type deposits where porosity and mineralization
tended to intensify toward the bounding faults, definition of the gravity
contour gradient zone was especially important.
Bouguer anomalies and gradient zones can be enhanced by various
analytic methods such as subjective regional-residual separation, high-
pass filtering, grid residual calculation, or derivative approximations.
Figure 6b is an example of the latter, using Elkins’ Formula 13 (Elkins,
1951) to calculate a derivative approximation of the detailed data at
Prospect A. Because the minima and gradient zones were of primary
interest, only the negative derivative values were contoured, in addition
to the zero contours defining centers of the steep gradient zones.
Mining and Environmental
Figure 6b. Gravity minima at Prospect A, based on detailed survey. Anomalies
enhanced by second-derivative approximation using Elkins’ Formula 13.
Subsurface Density Data
Once adequate surface gravity data were acquired by Exxon and the ini-
tial core tests were drilled, subsurface density control was needed to
interpret the prospective anomalies more fully. Densities were obtained
from several sources, primarily:
1) Density measurement of drill cuttings and cores
Advantages—relatively inexpensive if done on a selective basis.
Cuttings were routinely available, even if logs could not be run.
Density values were laboratory measured, not estimated.
Disadvantages—density data was not vertically continuous because
not all rock samples were recovered and/or tested. Although density
measurements from cores were considered reliable, those from cut-
tings were found to be less consistent. The method could not predict
lateral density variations away from the hole. Moisture content of
near-surface material could vary unpredictably and affect bulk densi-
ties significantly.
116
Figure 7. Plot of formation bulk density versus “typical” near-surface lithology in the
study area. Dots represent measured densities from core-hole cuttings and cores.
Vertical bars represent averaged values from core-hole meter surveys.
2) Estimates from gamma ray–neutron logs and lithologs
Advantages—these logs normally were run for formation identifica-
tion and correlation. They provided continuous vertical data and
could give acceptably accurate density values for massive anhydrite
and distinctive shale or gypsum reference formations. Some hole
deviation could be tolerated, and the gamma ray–neutron logs could
be run in cased holes.
Disadvantages—accuracy of gamma ray–neutron density values was
dependent on calibration to standardized densities of reference for-
mations by means of lithologs. When hole conditions were poor, cor-
relations between the density estimates and measured densities from
cuttings or cores often proved inconsistent from hole to hole, or even
within the same hole.
3) Measurements from gamma-gamma (density) logs
Advantages—the log normally could be calibrated properly to provide
05 Mining and Environmental 4/16/07 4:46 PM Page 117
acceptably accurate, continuous, hole-compensated density measure-
ments.
Disadvantages—an open hole was required for logging, but many of
the core holes were cased because of surface gravels and/or intensive
near-surface formation fracturing. The sondes in use had 43⁄8-inch
diameters, thus requiring more expensive 6-inch-diameter holes rather
than the standard 47⁄8-inch core holes. Availability of the logging units
on short notice was limited and logging costs were relatively high.
4) Measurements from core-hole gravity meter
Advantages—because of its 35⁄8-inch diameter, Exxon’s core-hole grav-
ity meter could be run in standard 47⁄8-inch core holes, either cased or
open. Calculated formation bulk densities gave consistent agreement
to within 0.2 gm/cc of the most reliable values derived from other
sources, and were the most dependable for the very shallow forma-
tions. Some zones of lateral density variation could be predicted by
the remote sensing aspects of the survey. Survey costs were moderate,
on the order of $2100 per hole (1969 dollars).
Figure 8. Bouguer gravity mapped from detailed survey at Prospect B.
Disadvantages—the meter could not be operated in a hole with ver- Gravity Prospect Analysis
tical deviation greater than 4°. Because the meter required discrete
Once promising gravity anomalies had been identified by reconnaissance
downhole stations, continuous density readings could not be
and the detailed surface gravity surveying had been completed, the gravity
obtained. A previously run gamma ray–neutron log usually was
and available geologic data were integrated to determine optimum loca-
needed to plan optimum borehole station depths.
tions for a core test program. Prior to drilling each core test, 2-D gravity
modeling was used to evaluate the most probable subsurface density con-
In all, Exxon ran twenty-eight core-hole surveys, one as deep as 884 m figurations producing the surface gravity anomalies. Drilling normally was
(2900 ft), in eight prospects. Figure 7 shows calculated and averaged planned to test minima along the central axis and to test along one or more
meter-derived densities from the surveys with comparable downhole lithol- steep contour gradient zone. If no encouraging traces of sulfur were
ogy. Results from the core-hole meter surveys were used in gravity prospect encountered, the gravity interpretation was reviewed to determine
analysis, as illustrated by two examples below. whether further drilling was warranted. Models were updated with the
additional density and lithology control obtained from the core tests. Then
if the most geologically reasonable interpretations could satisfy the gravity
data without producing additional leads, the prospect was abandoned.
Prospect B
The prospect appeared to be a
complex near-surface fracture or
alteration zone. Three core holes
were drilled normal to the axis of
a northwest-southeast Bouguer
minimum (Figure 8). Formation
densities for all three holes were
measured from cuttings and
cores and also were estimated
from gamma ray–neutron logs.
Core-hole gravity meter surveys
subsequently were run in the
three holes. Overall densities cal-
culated from the meter survey in
core hole No. 2 were as much as
0.6 gm/cc higher than those
estimated from the log and cut-
tings (Figure 9). In hole No. 3,
the meter-derived densities were
within 0.15 gm/cc of those esti-
mated from the gamma ray–neu-
tron log as calibrated by cutting
Figure 9. Plot of formation bulk densities versus depth at Figure 10. Plot of formation bulk densities versus depth at core hole densities (Figure 10). Density
core hole No. 2, Prospect B. Dashed line represents values No. 3, Prospect B. Thin solid line is gamma ray–neutron log trace. variations, both lateral and verti-
estimated from cuttings and gamma ray–neutron log. Solid Dashed line represents averaged values estimated from gamma cal, among the three holes drilled
line represents values calculated from core-hole gravity ray–neutron log. Thick solid line represents values calculated from
core-hole gravity meter.
across Prospect B were so great
meter survey.
117
Geologic Applications of Gravity and Magnetics: Case Histories
05 Mining and Environmental 4/16/07 4:46 PM Page 118

Mining and Environmental

118
only the center hole located inside the
bounding faults. Core-hole meter surveys
were run in all three holes; however, the
meter could not be run to total depth in
the center hole (No. 2) because of extreme
hole deviation below 244 m (800 ft).
Densities for the deep graben section
therefore were derived from litholog and
gamma ray–neutron log data. These data
indicated a formation density of 2.80
gm/cc for a 70-m (200-ft)-section (note
shaded area in model) immediately below
the 244-m (800-ft) meter station in hole
No. 2.
Initial 2-D gravity models for Prospect
C using core-hole meter densities plus the
deep graben log–derived densities pro-
duced mismatches of up to 1.1 mGal be-
tween calculated and observed profiles
(Figure 15). The models suggested the
presence of an unresolved low-density
Figure 12. Initial 2-D gravity model for Prospect B, using formation densities (gm/cc)
zone within the graben. From prior experi-
estimated from cuttings and gamma ray–neutron logs. ence at Prospect B and other areas, meter-
calculated densities gave more consistent
Figure 11. Plot of formation bulk density versus depth for three core holes at Prospect B. Density values were calculated from
core-hole gravity meter surveys. and reliable formation bulk density infor-
mation than did lithologs or gamma–ray
(Figure 11) that even the data from the core-hole meter surveys seemed densities to total depth of the core holes. These estimated density values neutron logs. Hence, the most probable
suspect initially. ranged between those from the meter surveys and those from cuttings cause for the residual gravity minimum was judged to be an anomalous
Two-dimensional gravity models, using all available geologic data to and logs. Figure 13 illustrates the final best-fit model obtained. zone in the deep graben where logged densities might not be representa-
establish geometry of the known or predicted lithologic units, were con- A general conclusion reached from this and other model studies was tive of the surrounding section. This concept was tested in models by
structed to test the validity of density values obtained from cuttings, logs, that the core-hole meter surveys provided more accurate and reliable for- replacing the 2.80 gm/cc section with a lighter density section. A model
and core-hole meter surveys. Suspicions that densities estimated from cut- mation bulk densities than could be obtained readily from other methods. using a section density of 2.30 gm/cc gave the best fit (Figure 15).
tings and logs were generally too low were confirmed by the models. A A specific conclusion reached for Prospect B was that the final best-fit Results of the meter surveys and model studies indicated a promising
typical example is illustrated by Figure 12. The country rock lithology, for- model was compatible with the geology, the subsurface densities, and the new lead at Prospect C. The core test in the graben may have narrowly
mation thickness, and density are standard for the surrounding area. surface gravity anomaly. The core tests did not reveal any significant sul- bypassed a relatively light-density, high-porosity, potentially mineralized
Within the bounds of the prospect, the initial model used densities esti- fur accumulation, and no unresolved gravity anomalies remained at the zone 244–335 m (800–1100 ft) deep. On the basis of the gravity analy-
mated from logs and cuttings for each of the three core holes (see Figures site. The prospect therefore was deemed as adequately tested and was sis, two additional lateral offset core holes within the graben were rec-
9 and 10). The calculated gravity minimum is approximately 2.0 mGal abandoned. ommended to evaluate Prospect C more fully. They were not drilled
larger than the observed minimum from the surface gravity survey. The because of cancellation of the entire exploration program.
misfit cannot be reduced significantly without unrealistic major revisions Prospect C
The Core-hole Gravity Meter
to the geometry of the model. Prospect C was a small, near-surface, graben-type feature. Three core tests
Subsequent models with generally similar geometry used meter- were drilled normal to the axis of the Bouguer minimum (Figure 14), with The core-hole gravity meter was developed by Exxon Production
derived densities to total depth of the surveys, then a suite of estimated Research Company expressly for use in Exxon’s sulfur exploration pro-
05 Mining and Environmental 4/16/07 4:46 PM Page 119
Figure 13. Best-fit 2-D gravity model for Prospect B, using formation densities
(gm/cc) calculated from core-hole gravity meter surveys and deeper densities esti-
mated from cuttings and gamma ray–neutron logs.
gram. The instrument was a special modification of the Exxon vibrating
string meter which previously has been described in detail (Howell et al.,
1966). Both the new sensitive element and the electronic section were
essentially identical with those of the older meters, but use in the low-
temperature environment of the average core hole permitted a reduction
in the thermostat controls for meter temperature. In addition, expectation
of low core-hole pressures allowed use of thin-wall instrument casing;
thus the entire assembly could be contained in a 35⁄8-inch OD case.
The entire well string, consisting of a nose, meter section, electronics
section, connector case, and Schlumberger head, measured 1103⁄4 inches
in length (Figure 16). Because of the danger of instrument loss from core-
hole bridging or casing collapse, two complete instruments plus one spare
sensitive element were fabricated to provide backup capability. Total cost
was approximately $30 000 (1969 dollars).
An electronics technician operated the meter and performed any nec-
essary field maintenance. Common practice was for the meter survey to
be run after the gamma ray–neutron log. Thus, a cable truck was imme-
diately available for the meter survey, as well as additional personnel to
Figure 14. Bouguer gravity mapped from detailed survey at Prospect C.
help the meter operator attach the instrument to the cable head. The
Exxon geologist present during the logging would interpret formation
breaks from the gamma ray–neutron log to preselect meter station loca-
tions for the survey. Because each station reading required about twenty
minutes, economy, hole conditions, and survey scheduling usually
dictated that the total number of stations per hole be kept to a minimum.
Station spacing greater than 61 m (200 ft) was seldom necessary,
however.
Summary of Results
No commercial discoveries by Exxon were attributed directly to Exxon’s
use of surface or core-hole gravity surveys, although surface surveys con-
ducted over known sulfur deposits detected anomalies which confirmed
validity of the method. The core-hole meter proved to be a technical suc-
cess. Whether continued core drilling, metering, and prospect analysis
would have led to a commercial discovery is uncertain. At the height of
the west Texas sulfur play, increased sulfur extraction from sour gas and
reduced world demand created an oversupply of the commodity, followed
Geologic Applications of Gravity and Magnetics: Case Histories
Figure 15. Comparison of initial and final versions of 2-D gravity model for Prospect
C. Formation bulk densities (gm/cc) calculated from core-hole gravity meter surveys
and/or estimated from cuttings and gamma ray–neutron logs.
by abrupt plunge in world sulfur prices. Exxon’s exploration effort was
halted before several remaining surface and subsurface prospect leads
were evaluated fully.
However, a positive by-product of the program was the recognition
that explorationists and data normally oriented toward hydrocarbon
exploration could play key roles in certain nonhydrocarbon ventures. The
program also demonstrated that data which might have declined in
importance for a given play could become a major asset to some new
unexpected venture, if properly archived.
Acknowledgments
Release of the gravity data, models, logging results, and meter description
by Exxon Exploration Company is acknowledged gratefully. Special
thanks are given to L. J. Srnka of Exxon for assisting in the release.
Appreciation also is extended to Corine Prieto and Steve Stephens of
Integrated Geophysics Corporation for their support in preparation of the
final text and illustrations for this paper.
119
05 Mining and Environmental 4/16/07 4:46 PM Page 120

Mining and Environmental

120
References An Airborne Multisensor Characterization
Elkins, T. A.,1951, The second vertical derivative method of gravity interpretation:
Geophysics, 16, 39–56.
of an Active Nuclear Waste Site
Hammer, S.,1939, Terrain corrections for gravity stations: Geophysics, 4, 184–194. T. Jeffrey Gamey
Howell, L. G, Heintz, K. O., and Barry, A.,1966, The development and use of a high-
precision downhole gravity meter: Geophysics, 31, 764–782. Aerodat, Inc.,Canada
Mississauga, Ontario, Canada
Additional References
Butler, S. W., and Dixon, L. P., 1968, Sulfur occurence in the Rustler Springs sulfur dis- J. Scott Holladay
trict, Culberson and Reeves County, Texas: Symposium and Guidebook, 1968 Field
Trip, Permian Basin Section: Soc. Econ. Paleont. Mineral., Publication 68–11, Vanguard Geophysics
91–93. Toronto, Ontario, Canada
Goodell, R. R., and Fay, C. H., 1964, Borehole gravity meter and its application:
Geophysics, 29, 774–782. Jonathan Nyquist
Haye, E. F.,1969, Application of gravity to sulfur exploration—Delaware Basin:
PhotoGravity Co., Inc., brochure. Oak Ridge National Laboratory
McCulloch, T. H., Kandle, J. R., and Schoellhamer, J. E.,1968, Application of gravity
Oak Ridge, Tennessee, USA
measurements in wells to problems of reservoir evaluation: Ninth Annual Logging
Symposium, Soc. Prof. Well Log Analysts. William Doll
Nettleton, L. L., 1954, Regionals, residuals, and structures: Geophysics, 19, 10–22. Oak Ridge National Laboratory
Smith, N. J., 1950, The case for gravity data from boreholes: Geophysics, 15, 605–635. Oak Ridge, Tennessee, USA
Zimmerman, J. B., and Thomas, E., 1969, Sulfur in West Texas: its geology and eco-
nomics: Bur. Econ. Geol. Cir. 69-2.
EDITOR’S NOTE: Integrated methods help constrain a worrisome
environmental problem—locating known and unknown waste
burial sites.

From November 1992 through January 1993, an airborne geophysical

survey was flown to characterize the Department of Energy’s Oak Ridge
Figure 16. Diagram of Exxon core-hole gravity meter in assembled well string. Reservation in Tennessee. This was the first phase of an extended pro-
gram involving multiple geophysical sensors operating from a helicopter
platform. These sensors included multifrequency helicopter electromag-
netics (HEM), total-field and vertical gradient magnetics, VLF-EM (very
low communications frequency), and multichannel gamma-ray radiomet-
ric spectrometry.
Survey lines were flown in a north-south direction, at a constant
draped altitude and a 50-meter line spacing. The EM sensors were
mounted in a 7-meter bird and flown at 30-m ground clearance. Magnetic
gradiometer and VLF-EM sensors were flown at 45-m ground clearance,
and the radiometrics package was mounted in the helicopter at 75-m
ground clearance. Navigation and data positioning were provided by a
real-time differential Global Positioning System (GPS).
05 Mining and Environmental 4/16/07 4:46 PM Page 121
The first objective of the survey was to verify and refine the underly-
ing geology. The reservation is situated in the classic fold-thrust belts of
the Appalachian Valley. Sedimentary layers of interbedded shale and lime-
stone dip to the southeast at roughly 45°.
Apparent resistivity (Figure 1) derived from a 1-D inversion of the 4-
kHz coaxial EM is presented here with the geologic divisions overlaid in
white. Resistivities of geologic units range from 100 to 1000 ohm-m, with
surficial cultural features such as buildings, power lines, or fences appear-
ing as conductors to less than 10 ohm-m in places. The larger holes in the
data are buildings in the Oak Ridge Reservation complex. FAA regulations
prohibit helicopter flights with sling loads over buildings.
The geophysical maps correlated exceptionally well with the geology
and confirmed structures that previously had been interpreted below
overburden. They also were able to suggest possible modifications to
existing contacts. Of particular interest was the possibility of detection of
karst features. These underground fractures and caves are known to exist
throughout much of the southern portion of the reservation, but cannot
be mapped easily. Many are dry, but some are fluid filled and represent a
Figure 1. Apparent resistivity in ohm-m.
significant transport mechanism. As such, they are a major concern for
those responsible for monitoring potential leachate and contaminant flow
off-site.
The second objective of this reconnaissance phase was the detection
of known and possibly unknown waste burial sites. This reservation has
been used for the disposal of various kinds of waste since the mid-1940s.
Several records have been lost over the years, and perceived danger
existed from undetected and therefore unremediated dumps.
The vertical magnetic gradient (Figure 2) clearly depicts the major
burial grounds as well as numerous other surface features. The limits of
the known and controlled waste sites have been presented here as black
overlays, with power lines shown in white. Although some waste group-
ings appear to extend somewhat farther than indicated in the records, no
new or previously unknown waste sites were detected. A large section in
the north of the reservation was declared safe and turned over to the city
of Oak Ridge after follow-up of the magnetic anomalies detected here.
Much of the waste generated and disposed of at this site is radioactive
in nature. The high-level gamma-ray survey measured responses in energy
Figure 2. Magnetic gradient in nT/m.
Geologic Applications of Gravity and Magnetics: Case Histories
bands corresponding to the common natural emissions of potassium, ura-
nium, and thorium, as well as those related to the man-made radionu-
clides of cesium, cobalt, and protactinium.
The potassium counts were highly responsive to the natural geology
of the area; thorium and uranium showed weaker correlation. The prox-
imity of the potassium window to those of cobalt, cesium, and protac-
tinium meant that the effects of Compton scattering were visible. In
addition to downward scattering of energy from potassium, the energy
band associated with the man-made isotopes in areas of strong contami-
nation was so wide as to affect the potassium data.
The potassium and total count data were used to strip out geologic
correlation from the counts of the man-made isotopes. The made it pos-
sible to create a detailed map tracing the contaminants alone. Most of
these anomalous paths were known from previous airborne radiometric
surveys and contained a mixture of isotopes. It is interesting to note, how-
ever, that isolated pockets exist where a single isotope predominates.
The data in Figure 3 show a ternary-style presentation of potassium,
cesium, and cobalt. Each element was assigned a color, the intensity of
121
05 Mining and Environmental 4/16/07 4:46 PM Page 122
Figure 3. Gamma-ray radiometric ternary presentation
which was determined by the magnitude of the response. The potassium
response in yellow reflects the background geologic effects. The red and
blue represent the cesium and cobalt counts, respectively. This format
allows clear definition of the extent and character of the radioactive con-
tamination. Single isotope sources have been highlighted for interest and
are subject to future investigation.
Detailed airborne follow-up began in February 1994. It included mul-
tifrequency EM, and total-field and vertical gradient magnetics flown
from a helicopter platform. The system consisted of a single bird contain-
ing both the EM and magnetics sensors, plus the navigation and attitude
instruments. This lightweight system permitted safer handling at low alti-
tudes and enabled the magnetometers to be flown at the same time and
height as the EM. This lower effective magnetic survey height repre-
sented an increase of approximately 300% in signal response from near-
surface objects.
Mining and Environmental
122
Figure 4. Aerial photograph of main waste area, viewed from southwest to northeast.
Survey sites were set out with three distinct objectives: finer geologic In the eastern end of the reservation, the karst features which showed
definition, research into karst feature responses, and higher resolution anomalous EM responses in the reconnaissance phase of the survey were
over waste-area groupings. flown at extremely low altitudes and close spacing. Because these areas
The Melton Valley and Bear Creek Valley were flown with east-west were clear of trees or obstructions, it was possible to fly as low as 5–10
lines at treetop level over the known waste areas to better define waste- m over large areas of ground. Similar coverage was possible over selected
pit boundaries and to eliminate the possibility of smaller unknown waste groupings in the Melton Valley area. Follow-up of these anomalies
dumps. This area also included a site which has not yet been made oper- unearthed previously unknown man-made objects at various locations
ational. It is expected that the data collected here will be an extremely about the reservation. These included a narrow-gauge underground
useful baseline study of the area before development. Figure 4 shows an pipeline and assorted construction debris.
aerial photograph of the Melton Valley waste area groups as seen from the
southwest.
The area to the east of the K-25 plant has an extremely complex geol-
ogy and a considerable amount of interference from surficial features. This
area was flown in two directions at treetop level to better define the geo-
logic contacts. A smaller area within this, known as the Whitewing scrap-
yard, also was flown at a tighter line spacing to better define its character.
06 State of the Art 4/16/07 4:57 PM Page 123

6
Airborne Altrex™ Electromagnetic/Magnetic Definition
of Hydrocarbon Allteration Plumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
J. D. Rowe, R. S. Smith, D. D. Beattie, Geoterrex
R. K. Warren, Consultant
Sources of Magnetic Anomalies over a Sedimentary Basin: Preliminary Results
from the Coastal Plain of the Arctic National Wildlife Refuge, Alaska . . . . . . . . . . . . 130
Jeffrey D. Phillips, Richard W. Saltus, and Richard L. Reynolds
U. S. Geological Survey
The Rise and Fall of Early Oil-field Technology: The Torsion
Balance Gradiometer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Robin E. Bell, Lamont-Doherty Earth Observatory
R. O. Hansen, Pearson, deRidder and Johnson, Inc.
Results from a High-resolution, 3-D Marine Gravity Gradiometry Survey
over a Buried Salt Structure, Mississippi Canyon Area, Gulf of Mexico . . . . . . . . . . . 137
State of the Art
L. F. Pratson, R. E. Bell, R. N. Anderson, Lamont-Doherty Earth Observatory
C. Affleck, D. Dosch, A. Grierson, J. White, Bell Aerospace Co.
B. E. Korn, R. L. Phair, Texaco Exploration and Production Inc.
E. K. Biegert, Shell Exploration and Technology Co.
P. E. Gale, Shell Offshore Inc.
The Role of Vertical Magnetic Gradiometry in Hydrocarbon Exploration . . . . . . . . . . . 146
H. O. Seigel and T. J. McConnell, Scintrex Limited
06 State of the Art 4/16/07 4:57 PM Page 124
Airborne ALTREX™
Electromagnetic/Magnetic Definition
of Hydrocarbon Alteration Plumes
J. D. Rowe, R. S. Smith, and D. D. Beattie
Geoterrex
Ottawa, Ontario, Canada
R. K. Warren
Consultant, Houston, Texas, USA
EDITOR’S NOTE: This paper documents provocative correlations
between electromagnetic/magnetic anomalies and production
in several oil fields in the western United States. Its strength
lies in the integration of methods to help understand the
near-surface geology.
Summary
ALTREX™ (ALTeRation EXploration) combined airborne electromagnetic
and magnetic surveys were flown over a number of oil fields in the south-
western United States with the intention of testing the feasibility of map-
ping hydrocarbon alteration plumes from the air. In some of the areas,
there is an anomalous magnetic response, and in many of the areas, there
is an excellent correlation between an enhanced conductivity anomaly
and the oil fields. The approach indicates that alteration plumes may be
detected and defined with greater confidence from a survey employing a
multisensor system than from a survey using only a single-sensor system
(e.g., magnetics only).
Introduction
In recent years, there has been an increased interest in identifying oil
deposits by detecting possible alteration plumes in the rocks above an oil
field, e.g., Tedesco (1995). This enables geophysical techniques with pen-
etration depths of a few hundred meters to detect anomalies associated
with deeply buried oil fields. However, it requires that the alteration be
near to the surface and that the alteration has measurably changed the
quantities being mapped.
State of the Art
124
In the case of magnetic methods, anomalous magnetic susceptibilities
above an oil reservoir have been documented by Reynolds et al. (1991).
The conditions required to obtain anomalous susceptibilities have been
described by Machel and Burton (1991) and by Goldhaber and Reynolds
(1991). Not all oil fields will show alteration. The effect of man-made cul-
ture (wells, pipes) also can be great, as in the Cement oil field (Reynolds
et al., 1991).
Tests of the radiometric method have been described by Saunders et
al. (1994), who found that the potassium and uranium count rates, when
normalized by the thorium count rate, are sensitive to hydrocarbon
microseepage alteration.
Electrical and IP measurements also have been tried (Sternberg,
1991), with the strongest anomalies occurring where the near-surface
rocks are porous and iron-rich. Again, culture (fences, pipes, well casings)
can be significant (Holladay and West, 1984). Electrical methods are sen-
sitive to conductivity, as are electromagnetic (EM) methods.
The advantage of airborne techniques is that large areas can be cov-
ered, increasing the possibility of traversing an oil field with an alteration
Figure 1.
signature. Magnetic and radiometric measurements commonly are col-
lected on an airborne platform. Electrical and IP measurements are not,
because they require contact with the ground, but an electromagnetic sys- Geophysical Target
tem can be mounted on an aircraft.
Typical alteration plumes, as shown in Figure 2, have a carbonate-silica
Geoterrex, a division of CGG Canada Ltd., has performed airborne
cap (0–30 m) at the surface. Below this conductivity low is a much more
transient electromagnetic (GEOTEM®) surveys over oil fields with alter-
electrically conductive unit. The low conductivity at the surface is caused
ation signatures (Figure 1). The surveys were to test the hypothesis that
by carbonate and silica deposited in the pore spaces, and the high con-
airborne EM methods can detect alteration in cases when the conductiv-
ductivity underneath is caused by clay alteration within the plume.
ity of the altered material varies from background. The probability of
Diagenetic magnetite also may be formed in the alteration zone. To be
detecting alteration was increased by mounting a magnetometer on the
detected, the respective anomalous responses must be greater than back-
aircraft to test for alteration of magnetic minerals.
ground.
Geochemical Model
The ALTREX™ Airborne Geophysical System
Hydrocarbons dissolved in water are moved continuously to the surface
through faults and microfractures. Temperature gradients, compaction, The ALTREX™ airborne geophysical system consists of a cesium-vapor
and changes in atmospheric pressure help to drive this upward migration. magnetometer and a GEOTEM® TDEM (time-domain electromagnetic)
The water in the vicinity of a hydrocarbon reservoir is reducing (oxida- system (Annan and Lockwood, 1991; Smith et al., 1996), installed in a
tion-reduction potential) relative to other waters in the basin. This water CASA C-212 STOL aircraft. The system, as shown in Figure 3, incorpo-
reacts with the surrounding rock to create alteration features that can be rates a large vertical-axis three-turn transmitter coil generating a dipole
mapped by geophysical methods. Rock types near the surface control the moment of approximately 450 000 Am2. The digital EM receiver is con-
intensity of the alteration. Following microseep reaction, the shallow sec- nected via a tow cable to multiaxis induction coils which are towed
tion can become more conductive (or resistive) and produce an obvious below and behind the aircraft. Survey grids of 500 m × 4000 m were
geophysical target. flown, with the lines oriented parallel to the larger dimension, for all of
the oil fields. The surveys were sufficiently large to sample the back-
ground in the region surrounding the fields.
06 State of the Art 4/16/07 4:57 PM Page 125

Railroad Valley, Nevada

Railroad Valley is an asymmetrical graben in the Basin and
Range Province. The complex geology and topography,
presence of shallow basalt flows and extremely complex
faulting (Vreeland et al., 1979) create a difficult environ-
ment for data acquisition by conventional geophysical
exploration methods. The mountain ranges, Grant Range
on the east and the Pancake Range on the west, are com-
posed of Paleozoic sediments, Cretaceous and Tertiary
sediments, Oligocene ignimbrites, and occasional overly-
ing Tertiary lavas. Some Tertiary intrusives cut the sec-
tion. The valley floor is composed of Quaternary sands,
gravels, and playa lake deposits. The sands and gravels are
alluvial fan deposits and are composed of rock fragments
from the nearby mountains (Duey, 1979). The Trap
Spring Field is located on the west side of Railroad Valley
and produces from a depth of 1400 m from the Pritchards
Station ignimbrite (French and Freeman, 1979). Many of
the Railroad Valley fields which produce from the ign-
Figure 2. imbrite formations have been charged from leaking of
deeper hydrocarbons from Devonian and Mississippian
marine and nonmarine formations.
Cultural editing of the geophysical data was
employed to remove obvious cultural spikes, and these
data have been presented as aeromagnetic derivative
maps and EM decay constant (tau) maps. These have
been overlaid on a base map which shows productive
wells as black dots. The Trap Spring Field is kidney
shaped and is located centrally on the survey grid.
The horizontal gradient map, as shown in Figure 4,
shows a micromagnetic signal which corresponds closely
to the field. In this presentation, the north-northeast-
trending axes of the magnetic high and low delineate the
edges of the alteration plume.
The vertical gradient map, as shown in Figure 5, Figure 4.
identifies contacts (yellow) related to horst and graben
fault blocks in Devonian rocks associated with the hydrocarbon reservoir. suggested to the north and south-southwest of the Trap Spring Field. The
The Precambrian basement rocks also contribute to the magnetic alteration plume mapped by the system is approximately 2 km × 6 km.
response. The Eagle Springs Field (the original discovery in Railroad Valley) and
The tau map, as shown in Figure 6, shows a conductivity high which the Kate Spring Field are located approximately 10 km east of the Trap
Figure 3.
corresponds to an alteration zone related to the hydrocarbon plume. In Spring Field. The Eagle Springs Field, a fault wedge structural and strati-
addition to an excellent correspondence with the field, two new leads are graphic trap, is near the center of the survey grid. The Kate Spring Field
is located in the south-central portion of the survey grid.

125
Geologic Applications of Gravity and Magnetics: Case Histories
06 State of the Art 4/16/07 4:57 PM Page 126
Figure 5.
The vertical gradient of the magnetics (Figure 7) shows basement con-
tacts (yellow) in the vicinity of the Eagle Springs Field which could indi-
cate basement zones of weakness and a possible relationship to the fault
blocks within the section. The Kate Spring Field shows a very weak ver-
tical gradient anomaly that may be related to basement tectonics. The tau
image (Figure 8) is dominated by a linear conductivity high along the
State of the Art
126
economic trade-off in line spacing versus
historic production for the area is an
important planning step for any airborne
survey. With good Global Positioning
System (GPS) location control, a closer
line spacing is always an option.
The Grant Canyon and Bacon Flat
Fields are structural traps located ap-
proximately 15 km south of the Eagle
Springs Field. The Grant Canyon and
Bacon Flat Fields produce at 1800 m
from the Devonian, Simonson, and
Guilmette formations (Duey, 1983; Suek
and Krazan, 1983). The magnetic verti-
cal gradient (Figure 9) identifies base-
ment contacts, possible faults or zones of
weakness (yellow) in the basement
which could relate to block-fault control
of the fields. The tau image is dominated
by a conductivity high along the western
edge of the survey grid. This conductivity
pattern (Figure 10) reflects a shallow geo-
logic unit that outcrops to the west. A
broad conductivity high corresponds
with the Grant Canyon and Bacon Flat
Fields alteration plume. The conductivity
high continues to the east and may indi-
cate an area of hydrocarbon potential.
Paradox Basin
The Paradox Basin fields produce from
Mississipian and Pennsylvanian carbon-
ates bounded by anhydrite (seal) and
shale and porous mound carbonate
zones. The fields in our study area are
Figure 6.
near the southwest shelf of the Paradox
Basin and frequently have traps related
southeastern edge of the survey. The conductivity in this probable fault to stratigraphy; structure is not a controlling factor (Ott and Roylance,
zone likely was caused by mineralized waters related to the geothermal 1983; Kopp et al., 1983; Crist et al., 1983). Some of the fields are related
springs in the area. The Eagle Springs Field is defined clearly by another to subtle basement structures (highs) where reef development may have
conductivity high, as are several leads to the north. The Kate Spring Field, been present. Basement faulting and fracturing may have produced a
however, does not have an obvious conductivity high. The aerial extent migration path for the hydrocarbons. The fields in this area produce from
of alteration is small compared with the flight-line spacing of 500 m. The approximately 1800 to 1900 m.
06 State of the Art 4/16/07 4:57 PM Page 127
Figure 7.
The tau map (Figure 11) shows several conductivity highs which
appear to correspond with topographic features, the most predominant
being Tin Cup Mesa. This reflects a conductive unit within the mesa for-
mation. The low-conductivity background is enhanced in the vicinity of
the Little Nancy Field, in the north-central portion of the survey area,
which may reflect the conductivity signature of an alteration plume.
It is interesting to note that the
Pollen Field was predicted after a series
of near-surface prospecting method tests
(Groth, 1994). Dark fossil pollen grains
were recovered from shallow bore-
holes. The preservation of the pollen
color was related to the oxidation-
reduction conditions. Hydrocarbon
microseepage retards or prevents deep
oxidative weathering over the oil reser-
voir. Induced polarization, thermal gra-
dient, and elemental soil analysis also
were used. Groth states, “The geo-
chemical significance of the Pollen
Field discovery is that hydrocarbons
migrated upward, without deviation, to
near the surface, where they produced
a distinct signature.”
The tau image (Figure 12) identi-
fies a series of approximately north-
south-trending conductivity highs that
correspond extremely well with oil
production as indicated by the well-
control information. It is very interest-
ing to note the extension of the
conductivity high north of the Pollen
Field, a strong indication that the
alteration plume extends in this direc-
tion. Other excellent associations are
noted, corresponding with existing
well control, and other leads are also
visible in the data.
Figure 8.
Conclusions
Airborne TDEM data can map surface
Denver Julesburg Basin conductivity contrasts effectively, and high-resolution aeromagnetic data
The Pollen Field, as described by Ziegler (1983), is located in the south- can identify anomalous patterns which may be related to hydrocarbon
central portion of the Denver Julesburg Basin, approximately 3 km south alteration plumes. The advantage of airborne methods is that large areas
of Strasburg in Arapahoe County, Colorado. Production is from a J sand- can be covered relatively easily, and alteration can be identified where the
stone marine-bar complex, which is quite extensive in the area, at approx- conductivity and/or magnetic susceptibility are anomalous. It must be
imately 2500 m. The same complex produces at the Quill and Fairway noted that other near-surface geologic and cultural features also will be
Fields. mapped and it will be the responsibility of the interpreter to determine
which anomalies, if any, relate to an alteration plume.
127
Geologic Applications of Gravity and Magnetics: Case Histories
06 State of the Art 4/16/07 4:57 PM Page 128

State of the Art

Figure 9.
A combined airborne TDEM/aeromagnetic program will minimize
the need for land access at the early stages of a regional evaluation pro-
gram. Leads may be identified for subsequent upgrading to prospect status
by surface geophysical, geologic, and other studies. The application
should be considered for further study in these areas and in similar geo-
logic settings.
128
French, D. E., and Freeman, K. J., 1979,
Tertiary volcanic stratigraphy and reser-
voir characteristics of Trap Spring Field,
Nye County, Nevada: Rocky Mountain
Assn. of Geologists and Utah Geol. Assn.,
1979 Basin and Range Symposium,
487–502.
Goldhaber, M. B., and Reynolds, R. L., 1991,
Relations among hydrocarbon reservoirs,
epigenetic sulfidization and rock magnet-
ism: Examples from the south Texas
coastal plain: Geophysics, 56, 748–757.
Groth, P. K. H., 1994, Pollen Field, Colorado:
Evidence for vertical hydrocarbon migra-
tion and geochemistry economic utility:
AAPG Annual Meeting Abstracts, Am.
Assn. Petr. Geol. and SEPM Soc. for Sed.
Geol., 159.
Holladay, J. S., and West, G. F., 1984, Effect
of well casings on surface electrical sur-
veys: Geophysics, 49, 177–188.
Kopp, R. A., Crist, R. P., and Walters, J., 1983,
Little Nancy, in Fassett, J.E., Ed., Oil and
gas fields of the Four Corners area, III:
Four Corners Geological Society,
1092–1093.
Machel, H. G., and Burton, E. A., 1991,
Chemical and microbial processes caus-
ing anomalous magnetization in environ-
ments affected by hydrocarbon seepage:
Geophysics, 56, 598–605.
Ott, V. D., and Roylance, M. H., 1983, Tin
Cup Mesa, in Fassett, J.E., Ed., Oil and
gas fields of the Four Corners area, III:
Four Corners Geological Society,
1134–1135.
Figure 10.
Reynolds, R. L., Fishman, N. S., and Hudson,
M. R., 1991, Sources of aeromagnetic
anomalies over Cement oil field
(Oklahoma), Simpson oil field (Alaska)
References and the Wyoming-Idaho-Utah thrust belt:
Annan, A. P., and Lockwood, R., 1991, An application of airborne GEOTEM® in Geophysics, 56, 606–617.
Australian conditions: Exploration Geophysics, 22, 5–12. Saunders, D. F., Branch, J. F., and Thompson, C. K., 1994, Tests of Australian aerial
Duey, H. D., 1979, Trap Spring Field, Nye County, Nevada: Rocky Mountain Association radiometric data for use in petroleum exploration: Geophysics, 59, 411–419.
of Geologists and Utah Geological Association, 1979 Basin and Range Symposium, Smith, R. S., Annan, A. P., Lemieux, J., and Pedersen, R. N., 1996, Application of a mod-
469–476. ified GEOTEM® system to reconnaissance exploration for kimberlites in the Point
Duey, H. D., 1983, Bacon Flat, in Fassett, J.E., Ed., Oil and gas fields of the Four Corners Lake area, NWT, Canada: Geophysics, 61, 82-92.
area, III: Four Corners Geological Society, 909–910.
06 State of the Art 4/16/07 4:57 PM Page 129

Figure 11. Figure 12.

Sternberg, B. K., 1991, A review of some experience with the induced-
polarization/resistivity method for hydrocarbon surveys: Successes and limitations:
Geophysics, 56, 1522–1532.
Suek, D. M., and Krazan, C. G., 1983, Stratigraphy and oil and gas production of
Nevada, in Fassett, J. E., Ed., Oil and gas fields of the Four Corners area, III: Four
Corners Geological Society, 846–848.
Tedesco, S. A., 1995, Surface geochemistry in petroleum exploration: Chapman & Hall.
Vreeland, J. H., and Berrong, B. H., 1979, Seismic exploration in Railroad Valley,
Nevada: Rock Mountain Assn. of Geologists and Utah Geol. Assn., 1979 Basin and
Range Symposium, 557–569.
Ziegler, J. R., 1983, Pollen Field: Oil and gas fields of Colorado and Nebraska and adja-
cent areas 1982: Rocky Mountain Association of Geologists, 396–399.

129
Geologic Applications of Gravity and Magnetics: Case Histories
06 State of the Art 4/16/07 4:57 PM Page 130
Sources of Magnetic Anomalies over
a Sedimentary Basin: Preliminary Results
from the Coastal Plain of the Arctic
National Wildlife Refuge, Alaska*
Jeffrey D. Phillips, Richard W. Saltus,
Richard L. Reynolds
U. S. Geological Survey
Denver, Colorado, USA
EDITOR’S NOTE: This paper describes remarkable magnetic sedi-
mentary rocks, showing how—in some special cases—mag-
netic data can provide a wealth of information about the
sedimentary section. It also gives us lessons about the role of
remanent magnetization in analysis.
Abstract
As part of tectonic studies by the Energy Program of the U.S. Geological
Survey, we have modeled aeromagnetic anomalies over the coastal plain
of the Arctic National Wildlife Refuge (ANWR), Alaska. Preliminary mod-
els indicate that the lineated, moderate-intensity anomalies produced by
shallow sources within the coastal plain are best fit by a series of strati-
graphic layers with both normal and reversed remanent magnetization.
The layers follow seismically determined stratigraphic and structural
boundaries from near the surface to depths of 1 to 2 km. The modeled
total magnetic intensities range up to .115 A/m for the reversely magne-
tized units and up to .069 A/m for the normally magnetized units. Based
on these models, we suspect that the magnetic anomalies are primarily
the result of detrital remanent magnetization that formed as the sedi-
ments were deposited. Another plausible explanation involves chemical
remanence, acquired rapidly with respect to geomagnetic polarity rever-
sals, as the marine turbidite sediments accumulated, thus producing a
stratigraphically ordered polarity sequence. The high total magnetizations
and reversed polarities leave open the additional possibility that thick
sequences of originally reversed magnetization were overprinted by nor-
mal remanence through some stratigraphically controlled mechanism.
*Prepared by an agency of the U. S. government.
State of the Art
Introduction
Magnetic anomalies and sedimentary rocks
Sedimentary rocks are not normally considered to be an important source
of aeromagnetic anomalies. However, with modern magnetometers and
satellite navigation systems, it is sometimes possible to map the low-
amplitude, short-wavelength (localized) anomalies produced by shallow
sedimentary rocks having favorable magnetization contrasts.
Certain conditions may enhance our ability to detect aeromagnetic
anomalies produced by sedimentary rocks. These include (1) a lack of
magnetic basement rocks within the upper few kilometers, (2) an absence
of young volcanic units overlying the sedimentary rocks, (3) favorable
magnetization contrasts within the sedimentary section, (4) folding or
faulting of the sedimentary section resulting in outcropping (or subcrop-
ping) contacts, (5) magnetic survey specifications appropriate for measur-
ing low-amplitude, short-wavelength anomalies, and (6) a relative
absence of cultural (ie., man-made) magnetic sources with pervasive
anomalies capable of masking the magnetic signatures of the sedimentary
units.
Basement rocks can have large contrasts in magnetic properties,
resulting in high-amplitude magnetic anomalies. When the basement
rocks are shallow, the strong, short-wavelength anomalies they produce
can obscure any weaker anomalies of similar wavelengths produced by
sedimentary rocks. When the basement rocks are deep, their anomalies
become broader, and therefore can be separated from the more localized
anomalies of the shallow sedimentary rocks by appropriate wavelength fil-
tering. Young volcanic units tend to produce strong magnetic anomalies
at all wavelengths, which obscure anomalies from underlying sedimen-
tary units.
Many sedimentary rocks do not contain sufficient amounts of mag-
netic minerals to generate measurable magnetic anomalies. Even if ample
amounts of magnetic minerals are present in the sedimentary section, no
anomalies will be produced unless sedimentary units with contrasting
magnetic properties are juxtaposed near the ground surface as a result of
folding, faulting, facies changes, and/or erosion.
Magnetic anomalies produced by outcropping sedimentary units can
be expected to have peak-to-trough widths equal to the widths of the out-
cropping units. Therefore, the average flight altitude (terrain clearance) of
an aeromagnetic survey designed to detect sedimentary rocks should be
no greater than the average width of the target units. Flight direction
should be perpendicular to the average strike, and flight-line spacing
should be no more than twice the flight altitude and preferably equal to
the flight altitude (Reid, 1980). Typical parameters used for high-resolu-
130
tion aeromagnetic surveys of sedimentary basins are terrain clearances of
30 m (in a helicopter) to 150 m (100 to 500 ft), and flight-line spacings
of 100 to 400 m (330 to 1300 ft). Optically pumped cesium-vapor mag-
netometers with sensitivities of 0.01 nanotesla (nT) or better are the
instruments of choice for high-resolution surveys. Use of differential
Global Positioning System (GPS) navigation is also standard practice.
Cultural sources of magnetic anomalies include railroad tracks, trains,
power transmission towers, metal buildings, drill rigs, pumps, storage
tanks, steel well casings, pipelines, parking lots, and other objects con-
taining significant amounts of iron or steel. Many of these sources pro-
duce isolated dipolar or narrow linear magnetic anomalies that can be
edited out of the data set or ignored during interpretation. However, large
concentrations of cultural sources or cultural sources with particularly
strong and pervasive magnetic fields (such as cathodically protected
pipelines) can seriously mask the geologic information contained in aero-
magnetic survey data.
Aeromagnetic anomalies of the ANWR 1002 Area
For examples of aeromagnetic anomalies produced by sedimentary rocks,
we will examine profile data from the northern part of the Arctic National
Wildlife Refuge (ANWR), northeastern Alaska (Figure 1). This part of
ANWR is named the 1002 Area after legislation passed by the U.S.
Congress in 1981 which permitted preliminary oil and gas exploration
activities in the area (Bird and Magoon, 1987). The 1002 Area consists of
a sedimentary basin containing undeformed to highly deformed
sequences of clastic rocks. Exposures of the rocks are rare within the
basin because of pervasive tundra cover and surficial deposits (Bader and
Bird, 1986). The material in the upper part of the basin was derived from
the Brooks Range to the south, and is referred to as the Brookian strati-
graphic sequence. The material in the lower part of the basin was derived
from the north, and has been designated the Ellesmerian stratigraphic
sequence. Basement rocks underlying the basin at depths of 3 to 6 km
include argillite, phyllite, quartzite, volcanic rocks, carbonate rocks, and
granite (Bird and Molenaar, 1987). Both the Brookian and Ellesmerian
sections contain potential petroleum source and reservoir rocks (Bird and
Molenaar, 1987; Bird et al., 1987; Magoon et al., 1987). Structural and
stratigraphic traps are abundant (Bruns et al., 1987; Kelly and Foland,
1987).
Because of the poor surface exposures, most of our knowledge about
basin structure and stratigraphy in the 1002 Area comes from seismic
reflection profiles collected by a consortium of oil companies in 1984 and
1985, combined with data for wells drilled on Alaska state and tribal
lands adjacent to ANWR. Much of the seismic reflection data remains pro-
06 State of the Art 4/16/07 4:57 PM Page 131

70˚30´
147˚ 144˚ 141˚
70˚30´
1 1986, 1988; McCabe and Sassen, 1986; McCabe et al., 1987; Elmore et
70˚ 70˚
A B
MC
A reactions are described by Reynolds et al. (1993) and by Goldhaber and
69˚30´ 69˚30´
147˚ 144˚ 141˚
0 100 KM
Figure 1. Location map showing the 1002 Area of the Arctic National Wildlife Refuge,
northeastern Alaska. Symbols indicate the Marsh Creek anticline (MCA) and locations
of magnetic profiles shown in other figures (1) (Figure 3), A (Figure 4), and B
(Figure 5).
prietary, but some of the data were published and interpreted in a USGS
Bulletin (Bird and Magoon, 1987). In addition to seismic and well data,
members of the consortium collected gravity data and ground magnetic
profile data, and analyzed shothole lithology and paleontology. Pro-
prietary aeromagnetic surveys were flown in 1968 (2-mile line spacing)
and 1981 (infill to 1-mile line spacing) by Aero Service (Geonex), and in
1991 and 1992 (1⁄2-mile line spacing) by BGM Airborne Surveys, Inc.
(World Geoscience). The USGS collected aeromagnetic data (1-mile line
spacing) in 1981 (see discussion below).
The 1002 Area of ANWR is a nearly ideal test area for the study of
sedimentary aeromagnetic anomalies. It has many of the conditions nec-
essary for detecting sedimentary magnetic sources: It is completely free of
cultural magnetic sources; the sedimentary units are folded and faulted,
resulting in potential magnetic contrasts near the ground surface; and the
crystalline basement rocks are deep enough so that their magnetic effects
can be removed easily from the data. In addition, a large amount of pro-
prietary geophysical information exists for the area, including seismic
reflection profiles and a detailed, although not strictly high-resolution,
aeromagnetic survey (the 1⁄2-mile BGM survey).
Magnetic Properties of Sedimentary Rocks
Although regarded as magnetically transparent by early practitioners of
the magnetic method for oil exploration, sedimentary rocks have become
the target of many modern surveys. Depending on the genesis of magne-
tization in sedimentary rocks, these modern surveys may yield detailed
information about structure, fluid pathways, or geochemical reactions.
Among the important geochemical reactions are those postulated to form
certain species of iron oxides over oil fields (Donovan et al., 1979, 1984,
al., 1987; Henderson et al., 1984). Reactions that produce magnetically
important iron sulfide minerals and the geologic settings that favor such
Reynolds (1991).
Magnetization in sedimentary rocks may be divided into two main
classes: primary, or detrital (produced when particles of magnetic miner-
als are deposited with other sediment), and secondary (formed after depo-
sition). In general, processes leading to detrital magnetization are
understood better than the myriad possible reactions that can lead to sec-
ondary magnetization. Detrital magnetization is controlled by variations
in the type and abundance of magnetic grains from the sediment source
area and by the depositional environment itself. Secondary magnetiza-
tions may arise from many processes, but herein, we consider only mag-
netizations associated with chemical reactions that produce new
magnetic minerals and/or that destroy detrital magnetic particles.
Magnetic anomalies result from the total magnetization of the rock
body, which is the vector sum of the induced and remanent magnetiza-
tions. Induced magnetization arises from the interaction of magnetic min-
erals with the Earth’s magnetic field; induced magnetization acts in the
direction of the Earth’s magnetic field and thus produces positive anom-
alies at high latitudes. The intensity of induced magnetization is the prod-
uct of the effective magnetic susceptibility (k) and the local strength of the
Earth’s magnetic field (typical value 50 000 nT). Susceptibilities for typi-
cal sedimentary rocks can vary from 36 to 3600 × 10–6 in SI units (3 to
300 × 10–6 cgs), with values as high as 20 920 × 10–6 SI (1665 × 10–6 cgs)
reported (all values from Carmichael, 1989).
Remanent magnetization exists independently of the present Earth’s
magnetic field. The intensity and orientation of remanent magnetization
depend on the rock’s magnetic mineralogy and orientation relative to the
Earth’s magnetic field at the time the magnetization was formed. If rema-
nent magnetization is sufficiently strong and acts in a direction opposite
to the present Earth’s field, it can produce negative anomalies at high lat-
itudes. The ratio of the strength of remanent magnetization to induced
magnetization is called the Koenigsberger ratio. Koenigsberger ratios
greater than 1 indicate that remanence plays a greater role than induced
magnetization and must be considered in anomaly analysis. Standard
tables (Carmichael, 1989) list Koenigsberger ratios ranging from 0.02 to
10 for sedimentary rocks.
Relatively few case studies have been published that combine aero-
magnetic interpretation with the detailed rock magnetic investigation
required to establish unequivocally the source of anomalies in sedimen-
tary rocks. In one such study, Reynolds et al. (1991) found different mag-
netic sources in their investigation of aeromagnetic anomalies over
sedimentary regions of Oklahoma, Alaska, and Wyoming-Idaho-Utah. At
Cement oil field (Oklahoma), they found evidence for secondary produc-
tion of ferrimagnetic pyrrhotite (Fe7S8) as a result of hydrocarbon seepage.
Although the observed anomalies at Cement field (see Donovan et al.,
1979) are caused by cultural features such as buried well casing and
pipelines (Boardman, 1985), the amount of pyrrhotite there probably
would be sufficient to generate very subtle anomalies in the absence of
cultural interference (Reynolds et al., 1990). At Simpson oil field (Alaska),
Reynolds et al. (1991) found that ferrimagnetic greigite (Fe3S4), some of
which is related to hydrocarbon seepage, had sufficient natural remanent
magnetization to contribute to magnetic anomalies. In the Wyoming-
Idaho-Utah thrust belt, they found that detrital magnetite, commonly con-
centrated in heavy-mineral laminations in the Middle Jurassic Preuss
Sandstone, was the source of the magnetic anomalies. In this and other
studies, no direct evidence has been found of magnetic anomalies caused
by secondary magnetite over oil fields.
ANWR Stratigraphic/Tectonic Setting
The stratigraphy of the ANWR coastal plain can be divided broadly into
three parts: (1) Precambrian to Middle Devonian metasedimentary base-
ment rocks, (2) a Mississippian to Early Cretaceous passive margin
sequence derived from a landmass to the north (Ellesmerian sequence),
and (3) a Cretaceous- and Tertiary-age foredeep and passive margin
sequence derived from the orogenic highlands of the Brooks Range to the
south (Brookian sequence). Stratigraphic and structural features con-
tribute to the overall geologic complexity of the region (Bird and Bader,
1987). Elements that contribute to stratigraphic complexity include the
poorly known northern erosional limits of the Ellesmerian sequence and
the extent of lateral facies changes within the northeasterly prograding

131
Geologic Applications of Gravity and Magnetics: Case Histories
06 State of the Art 4/16/07 4:57 PM Page 132
Brookian sequence. The Brookian deposits consist largely of older deep-
water marine turbidites and younger mudstones, siltstones, and sands.
There is no evidence in shothole samples or nearby wells for any volcanic
units within the basin, although in some of the sand units, as much as
10% of the grains consist of volcanic lithic fragments (Bird et al., 1987,
Table 7.6).
The region has a long and complex structural history culminating in
the multistage production of a thin-skinned fold and thrust belt with sig-
nificant basement involvement. Figure 2 is a cartoon of a generalized
south-to-north profile in the central part of ANWR. Basement is thrust to
the surface in the south (as in the Shublik and Sadlerochit Mountains).
Farther north, the Ellesmerian and Brookian sequences are deformed on
a north-directed thrust. North of that, Brookian rocks are back thrust (in
a passive roof duplex) above the advancing thrust front. Locally complex
structures are formed in the upper levels of the relatively incompetent
Brookian sequence, especially in the back-thrust region. There is a fairly
abrupt transition between the deformed zone and the nearly flat-lying
units of the undeformed zone to the north.
Cartoon cross-section, ANWR coastal plane
S N
Sadlerochit
Shublik Mtns. Mtns. deformed
zone
undeformed
zone
?
Younger Brookian sedimentary rocks Ellesmerian sedimentary rocks
Older Brookian sedimentary rocks Pre-Mississippian basement rocks
Figure 2. Cartoon geologic cross-section (unlocated) showing basement-involved
thrusting forming the Shublik and Sadlerochit Mountains, complex folding and thrust-
ing within the deformed section of the basin, and the relatively undeformed section of
the basin to the north.
Previous Aeromagnetic Work—The 1981 USGS Survey
An experimental aeromagnetic gradiometer survey of the ANWR coastal
plane was flown in July 1981 by the U.S. Geological Survey using a
Beechcraft E50 twin-prop aircraft fitted with three proton-precession
magnetometers, one on each wing tip and one on a tail stinger. The sur-
vey was flown with a nominal flight-line spacing of 1 mile (1.6 km) and
State of the Art
at an altitude of 300 ft (90 m) above ground. Flight lines were mostly east-
west in the eastern part of the survey and north-south in the western part
of the survey.
The intent of the three-magnetometer configuration was to enable the
collection of horizontal gradient data both parallel and perpendicular to
the axis of the aircraft. These gradient data were to be used for direct
detection of short-wavelength magnetic features and to aid in the inter-
polation of magnetic field values between flight lines. In fact, problems
with magnetic compensation of the wing-tip sensors made determination
of gradients unreliable (Cunningham et al., 1987). Published maps from
the survey (possibly those in Donovan et al., 1984, and certainly those in
Cunningham et al., 1987) show calculated gradients based on the differ-
ence between a given total-field reading from the tail stinger and another
tail-stinger reading taken 2.2 seconds (a distance of about 520 ft, or 160
m) later. The raw data from this survey are available from the National
Geophysical Data Center (NGDC) in Boulder, Colorado (Survey #0260-
926).
The magnetic gradients obtained from this 1981 survey have been
interpreted as evidence for shallow magnetic anomaly sources of the
intensity expected for various models of chemical magnetization
associated with hydrocarbon microseepage (Donovan et al., 1984; Cun-
ningham et al., 1987). However, the aeromagnetic data simply show that
there are significant magnetic sources within the near-surface
sedimentary rocks; they give no specific indication of the cause for those
anomalies. Furthermore, the directional nature (along flight-line path)
of the gradient employed gives a biased picture of the distribution
of anomalies, thus making it difficult to assign significance to their
spatial distribution.
Interpretation of Detailed Aeromagnetic Data
Because the existing USGS aeromagnetic data were deemed inadequate,
we acquired proprietary aeromagnetic data for the ANWR 1002 Area
from BGM Airborne Surveys, Inc. (World Geoscience). These data were
collected at 300 m above terrain along flight lines spaced 800 m apart.
The data were corrected and leveled diurnally by the company (see
Wilson et al., 1997), and supplied as leveled flight-line data. Using USGS
software (Phillips et al., 1993; Phillips and Grauch, in press), we pro-
jected and gridded the leveled flight-line data to a 200-m grid, then used
wavelength filtering techniques to separate the gridded data into long-
and short-wavelength components. The long-wavelength component
results primarily from magnetic sources within the basement. The short-
wavelength component is caused primarily by sedimentary magnetic
sources near the ground surface. The filtering program removes a planar
132
surface from the data prior to processing, then adds this surface to the
longest-wavelength component after filtering. Thus, the short-wavelength
component is automatically leveled and surface corrected.
Figure 3 compares the filtered components to the original data along
a north-south profile (profile 1, Figure 1). The short-wavelength magnetic
anomalies have their greatest amplitudes over the east-northeast-trending
Brookian fold and thrust belt (deformed zone), and have much lower
amplitudes over the undeformed sedimentary section to the north-north-
west (undeformed zone) and the largely nonmagnetic sedimentary sec-
tion to the south-southeast (nonmagnetic zone). Over the fold-and-thrust
belt, the magnetic anomalies are lineated strongly along strike.
We extracted magnetic profiles from the short-wavelength grid along
the locations of cross-strike seismic reflection profiles. Two-dimensional
magnetic models then were constructed to attempt to honor the struc-
tural and stratigraphic interpretations of the seismic data and fit the
observed magnetic anomalies. Figure 4 shows the magnetic model for a
relatively undeformed section containing north-northwest-dipping units
and producing weak magnetic anomalies. The modeling assumed that the
total magnetization was aligned with the Earth’s magnetic field. To fit the
magnetic anomalies and the observed dips, the model required magne-
tized sedimentary layers with total magnetization directions, both paral-
lel (represented as positive intensities) and antiparallel (represented as
negative intensities) to the Earth’s field. Total magnetization intensities in
the model vary from –3.75 × 10–2 to 4.61 × 10–2 A/m, suggesting
S N
40
nonmagnetic deformed zone undeformed
20
zone zone
Magnetic field (nT)
0
-20
-40
all wavelengths
-60
short wavelengths
-80
-100 long wavelengths
-120
profile 1
0 10 20 30 40 50
Distance (km)
Figure 3. Magnetic profiles extracted from grids of unfiltered and filtered aeromag-
netic data along profile 1 (Figure 1). The filtering was designed to separate long-wave-
length anomalies produced by sources within the basement from short-wavelength
anomalies produced near the top of the sedimentary section. Note that the intensity of
the short-wavelength field goes to zero over both the nonmagnetic section to the
south and the undeformed section to the north. This provides confidence that the zero
level of the profile is correct for modeling purposes.
06 State of the Art 4/16/07 4:57 PM Page 133

pattern, the most likely explanation is that the magnetizations represent

profile A remanence acquired by the Tertiary marine sediments via depositional
Residual magnetic intensity (nT)
processes as sedimentation proceeded, perhaps discontinuously. In this
3
S N way, such detrital remanent magnetization (DRM) would have been
acquired over time intervals that spanned many reversals of the geomag-
0

netic field (see Butler, 1992; Harland et al., 1990). Another explanation
-3

dashed = observed; solid = calculated.

is that the magnetization is carried by minerals produced chemically in
flight surface
the sediments at some time after deposition. Such chemical remanent
magnetization (CRM) may account for the high magnetizations and
inferred high Koenigsberger ratios. Nevertheless, if CRM is responsible, it
0

most likely would have been acquired as the sediments accumulated,

normal reversed without major time lags, to produce a stratigraphically ordered polarity
magnetic magnetic
intensity (A/m) intensity (A/m) sequence. CRM acquisition long after deposition but before deformation,
>0.05 >0.05
via processes involving introduced fluids, seems unlikely to result in the
observed anomaly patterns that so closely track seismic stratigraphy and
Depth (km)

0.025 - 0.05 0.025 - 0.05 bedding attitude. Finally, lithologically controlled overprints of normal
<0.025 <0.025
polarity on thick sequences of originally reversed magnetization might be
considered as a mechanism to produce the observed pattern.
seismically interpreted interfaces Tests for all these possibilities are planned, using samples returned
from the area of the anomalies. Tests would include magnetic-property
measurements (magnetic susceptibility as well as direction and magni-
tude of remanent magnetization with demagnetization experiments) and
seismic
basement the chemical and textural characterization of magnetic minerals to deter-
5

0 5 10 15 20 mine their origins.

V.E. = 2 Distance (km)

Preliminary Conclusions
Figure 4. Magnetic model along seismic profile A. The short-wavelength observed and calculated magnetic profiles are shown at the top. The magnetic units in the
model are color-coded by polarity and total magnetization intensity. Heavy lines on the cross-section are structural and stratigraphic boundaries interpreted from Model studies suggest that short-wavelength aeromagnetic anomalies over
the seismic reflection data; subhorizontal boundaries are representative of many parallel reflectors. Seismic ties to wells west of the 1002 Area indicate that the the coastal plain at ANWR are produced by the erosional truncation and
magnetized sedimentary rocks are Paleocene in age (Bird and Magoon, 1987). faulting of dipping normally and reversely magnetized sedimentary layers.
In the models, these layers were made to conform to seismically mapped
Koenigsberger ratios greater than 1 and high remanent intensities, espe- across the dipping layers seen in the seismic reflection data. In this case, stratigraphic horizons. The presence of persistent normal and reversed
cially for the antiparallel layers. The magnetic anomalies of the model are it would be difficult to explain the close correlation between the lineated magnetizations within stratigraphic layers suggests a detrital origin for the
produced by the erosional truncation of the dipping layers near the magnetic anomalies and the regional strike of the sedimentary units. remanance. Remanent magnetization intensities and Koenigsberger ratios
ground surface. Thus, neither of these modifications is likely to produce a successful or values are apparently high, but still within the range of reported values for
We believe the layers with total magnetization directions antiparallel geologically reasonable model. sedimentary rocks (Carmichael, 1989). Planned laboratory work on shot-
to the Earth’s field provide evidence for the presence of reversed rema- Figure 5 shows a magnetic model for a more complex structural sec- hole samples and, ideally, on oriented samples will yield quantitative mag-
nent magnetizations. The modeling could be modified in two ways in an tion crossing the Marsh Creek anticline and producing more intense mag- netic property information. Other explanations are possible, however, such
attempt to fit the anomalies with strictly induced magnetization. First, the netic anomalies. Again, parallel and antiparallel magnetized layers as chemical magnetization acquired as the sediments accumulated, or
zero level of the observed magnetic field could be shifted downward until conforming to seismically mapped structures were required. Total mag- combinations of primary and lithologically controlled secondary magneti-
nearly all the anomalies were positive. In this case, it would be necessary netization intensities vary from –11.53 × 10–2 to 6.92 × 10–2 A/m. zations. The pattern of anomalies is not one expected for conditions and
for the model to fit nearly constant positive magnetic anomalies over the The inferred reversed and normal magnetizations that arise from the alteration related to hydrocarbon seepage (Reynolds et al., 1991).
undeformed zone to the north and the nonmagnetic zone to the south, a models can be interpreted in more than one way. On the basis of polarity
virtual impossibility. Second, the magnetic contacts could be made to cut reversals that correspond to seismically traced stratigraphy and outcrop

133
Geologic Applications of Gravity and Magnetics: Case Histories
06 State of the Art 4/16/07 4:57 PM Page 134

State of the Art

134
Donovan, T. J., O’Brian, D. O., Bryan, J. G., and Cunningham, K. I., 1986, Near-surface

Residual magnetic intensity (nT)

profile B magnetic indicators of buried hydrocarbons, aeromagnetic detection and separation
of spurious signals: Assn. of Petr. Geochem. Expl., 2, 1–20.
10
S N Elmore, R. D., Engel, M. H., Crawford L., Nick, K., Imbus, S., and Sofer, Z., 1987,
Evidence for a relationship between hydrocarbons and authigenic magnetite:
Nature, 325, 428–430.
-10 0

Goldhaber, M. B., and Reynolds, R. L., 1991, Relations among hydrocarbon reservoirs,
dashed = observed; solid = calculated. epigenetic sulfidization, and rock magnetization: Examples from the south Texas
flight surface coastal plain: Geophysics, 56, 748–757.

MCA Harland, W. B., Armstrong, R. L., Cox, A. V., Craig, L. E., Smith, A. G., and Smith, D. G.,
1990, A geologic time scale: Cambridge Univ. Press.
0

normal reversed
magnetic magnetic
Henderson, R., Miyazaki, Y., and Wold, R., 1984, Direct indication of hydrocarbons
intensity (A/m) intensity (A/m)
from airborne magnetics: Expl. Geophys., 15, 213–219.
>0.05 >0.05
Kelly, J. S., and Foland, R. L., 1987, Structural style and framework geology of the
Depth (km)

coastal plain and adjacent Brooks Range, in Bird, K., and Magoon, L., Eds.,
0.025 - 0.05 0.025 - 0.05 Petroleum geology of the northern part of the Arctic National Wildlife Refuge,
northeastern Alaska: U.S. Geol. Surv. Bull. 1778, 255–270.
<0.025 <0.025
Magoon, L. B., Woodward, P. V., Banet, A. C. Jr., Griscom, S. B., and Daws, T. A., 1987,
seismically interpreted
interfaces Thermal maturity, richness, and type of organic matter of source-rock units, in Bird,
seismic
basement K., and Magoon, L., Eds., Petroleum geology of the northern part of the Arctic
National Wildlife Refuge, northeastern Alaska: U.S. Geol. Surv. Bull. 1778,
127–179.
V.E. = 2.25
5

0 5 10 15 20
Distance (km)
Figure 5. Magnetic model along seismic profile B. Color-coding of the magnetic units is identical to that in Figure 4. Seismic correlations indicate that the magnetized
sedimentary rocks in the core of the Marsh Creek anticline (MCA) are late Paleocene and Eocene in age (Bird and Magoon, 1987).
References Bruns, T. R., Fisher, M. A., Leinbach, W. J. Jr., and Miller, J. J., 1987, Regional structure
Bader, J. W., and Bird, K. J., 1986, Geologic map of the Demarcation Point, Mt. of rocks beneath the coastal plain, in Bird, K., and Magoon, L., Eds., Petroleum
Michelson, Flaxman Island, and Barter Island quadrangles, northeastern Alaska: geology of the northern part of the Arctic National Wildlife Refuge, northeastern
U.S. Geol. Surv. Miscellaneous Investigations Series Map I-1791, scale 1:250,000. Alaska: U.S. Geol. Surv. Bull. 1778, 249–254.
McCabe, C., and Sassen, R., 1986, Magnetic anomalies and crude oil biodegradation:
Geol. Soc. Am. Abstracts with Program, 18, 687.
McCabe, C., Sassen, R., and Saffer, B., 1987, Occurrence of secondary magnetite within
biodegraded oil: Geology, 15, 7–10.
Phillips, J. D., Duval, J. S., and Ambroziak, R. A., 1993, National geophysical data grids:
Gamma-ray, gravity, magnetic, and topographic data for the conterminous United
States: U.S. Geol. Surv. Digital Data Series CD-ROM DDS-9.
Phillips, J. D., and Grauch, V. J. S., Eds., in press, Geophysical map interpretation on the
PC: U.S. Geol. Surv. Digital Data Series CD-ROM.

Bird, K. J., and Bader, J. W., 1987, Regional geologic setting and history of petroleum
exploration: in Bird, K., and Magoon, L., Eds., Petroleum geology of the northern
part of the Arctic National Wildlife Refuge, northeastern Alaska: U.S. Geol. Surv.
Bull. 1778, 17–25.
Bird, K., and Magoon, L., Eds., 1987, Petroleum geology of the northern part of the
Arctic National Wildlife Refuge, northeastern Alaska: U.S. Geol. Surv. Bull. 1778.
Bird, K., and Molenaar, C. M., 1987, Stratigraphy, in Bird, K., and Magoon, L., Eds.,
Petroleum geology of the northern part of the Arctic National Wildlife Refuge,
northeastern Alaska: U.S. Geol. Surv. Bull. 1778, 37–59.
Bird, K., Griscom, S. B., Barsch-Winkler, S., and Giovannetti, D. M., 1987, Petroleum
reservoir rocks, in Bird, K., and Magoon, L., Eds., Petroleum geology of the north-
ern part of the Arctic National Wildlife Refuge, northeastern Alaska: U.S. Geol. Surv.
Bull. 1778, 79–99.
Boardman, J. W., 1985, Magnetic anomalies over oil fields: M.S. thesis, Colorado School
of Mines.
Butler, R. F., 1992, Paleomagnetism: Blackwell Scientific Publications.
Carmichael, R. S., 1989, Magnetic properties of rocks and minerals, in Carmichael, R. S.,
Ed., Handbook of physical properties of rocks: CRC Press, 229–287.
Cunningham, K. I., Roberts, A. A., and Donovan, T. J., 1987, Horizontal-gradient mag-
netic and helium surveys (chapter 14), in Bird, K. J., and Magoon, L. B., Eds.,
Petroleum geology of the northern part of the Arctic National Wildlife Refuge,
northeastern Alaska: U.S. Geol. Surv. Bull. 1778, 209–218.
Donovan, T. J., Forgey R. L., and Roberts, A. A., 1979, Aeromagnetic detection of dia-
genetic magnetite over oil fields: AAPG Bull., 63, 245–248.
Donovan, T. J., Hendricks, J. D., Roberts, A. A., and Eliason, P. T., 1984, Low-altitude
aeromagnetic reconnaissance for petroleum in the Arctic National Wildlife Refuge,
Alaska: Geophysics, 49, 1338–1353.
Donovan, T. J., Hendricks, J. D., Roberts, A. A., and Eliason, P. T., 1988, Low-level aero-
magnetic surveying for petroleum in arctic Alaska, in Gryc, G., Ed., Geology and
exploration of the National Petroleum Reserve in Alaska, 1974–1982: U.S. Geol.
Surv. Professional Paper 1399, 1, 623–632.
Reid, A. B., 1980, Aeromagnetic survey design: Geophysics, 45, 973–976.
Reynolds, R. L., Fishman, N. S., Hudson, M. R., 1991, Sources of aeromagnetic anom-
alies over Cement oil field (Oklahoma), Simpson oil field (Alaska), and the
Wyoming-Idaho-Utah thrust belt: Geophysics, 56, 606–617.
Reynolds, R. L., Goldhaber, M. B., and Tuttle, M. L., 1993, Sulfidization and magneti-
zation above hydrocarbon reservoirs: Applications of paleomagnetism to sedimen-
tary geology, Society for Sedimentary Geology Special Publication No. 49, 167–179.
Reynolds, R. L., Webring, M., Grauch, V. J. S., and Tuttle, M., 1990, Magnetic forward
models of Cement oil field, Oklahoma, based on rock magnetic, geochemical, and
petrologic constraints: Geophysics, 55, 344–353.
Wilson, C. R., Tsoflias, G., Bartelmann, M., and Phillips, J., 1997, A high precision aero-
magnetic survey near the Glen Hummel field in Texas: Identification of cultural and
sedimentary anomaly sources: The Leading Edge, 16, 37–42.
06 State of the Art 4/16/07 4:57 PM Page 135

The Rise and Fall of Early Oil-field

Technology: The Torsion Balance
Gradiometer
Robin E. Bell
Lamont-Doherty Earth Observatory
Palisades, New York, USA

R. O. Hansen
Pearson, deRidder and Johnson, Inc.
Lakewood, Colorado, USA

EDITOR’S NOTE: Remember Spindletop and Nash Dome? Gravity

gradiometry, a recent contribution to the acquisition “tool-
box,” is not new at all. While Pratson et al. (following this
paper) show the latest innovative techniques using high sen-
sitivity, measured gravity gradients, Bell and Hansen supply
the history behind the technology.

Today, elementary physics students take for granted such quantities as

“big G,” the universal gravitational constant. In fact, in the late 1700s the
value of this quantity was unknown, and the quest to determine it led to
some of the earliest geophysical instrumentation. Just after the
Revolutionary War in the United States, Henry Cavendish developed the
first system to measure the universal gravitational constant, the familiar
“big G.” Unfortunately, for geologists (at this time still mostly “gentlemen
scientists”), this apparatus produced data which were difficult to interpret
geologically, and it was far too large and cumbersome for field use. The
geologic limitation was that the system measured only the horizontal
derivative of a horizontal component of the gravity field, a quantity which
by itself is difficult to interpret. Thus no applications of this elegant yet
laboratory-bound instrument emerged.
Almost a full century later, the great Hungarian physicist Baron von
Eötvös designed an instrument which would revolutionize the petroleum
industry. As is often the case in revolutionizing technology, Eötvös used
“new” fiber technology to significantly reduce the instrument’s size and
thereby increase portability. Eötvös also added a significant new feature.
His master stroke was a design which suspended the weights on the tor-
sion balance at different elevations. This modification made it possible to
measure both the horizontal derivative of the horizontal field and the
Figure 1. An early torsion-balance instrument (6 ft tall) which was imported to Britain
for a series of demonstrations in London. The gravitational force acting on the small
mass on the upper left (point E) was different from the gravitational force acting on
the mass on the lower right (point N). The differing forces caused the rotation of the
system about the central axis of the instrument. This rotation, observed through the
telescope on the left (point C), was proportional to the local gravity gradients (from
“Eötvös Torsion Balance” by H. Shaw and E. Lancaster-Jones, Proceedings of the
Physical Society of London, 1923).
derivative of the vertical field (Figure 1). The vertical derivative was sig-
nificantly easier to interpret geologically.
Eötvös announced his invention in 1886. His work had been closely
followed by the head of the Hungarian Geologic Survey, Von Boeckh,
who had been frustrated by his inability to extend the mapping of clearly
exposed structures in the mountains beneath the flat-lying plains. Due to
prodding by Von Boeckh, Eötvös evaluated the instrument’s sensitivity to
geologic structures in the early years of this century and published the
results in 1908. The first experiments were on a frozen lake, where the
depths had been carefully mapped from a boat using a weighted line. The
lake gradients could be predicted with confidence. These early successes
Figure 2. Early map of gravity gradients over a salt dome in Germany. The dashed line
outlines the location of the low-density salt dome known from drilling. The arrows
indicate the direction and amplitude of the gravity gradient measurement (from “Die
bedeutung der drehwage von Eötvös fur die geologische forschnung nebst mitteilung
der ergbnisse einiger messungen” by W. Schweydar, Zeitschrift fur praktische
Geologie, 1918).
led to more ambitious efforts to map basement surfaces and, finally, dur-
ing World War I, to successfully mapping salt domes associated with oil
deposits in Germany, Hungary, and Czechoslovakia (Figure 2).
These efforts were followed closely by international oil companies.
Anglo-Persian acquired several instruments and began global operations
with the torsion balance. Efforts were made in 1914 to import this tech-
nology to the United States, but they were delayed by the war.
Interestingly, during World War I, E. W. Shaw published a landmark paper
in Science outlining the applications of gravity in oil exploration.
Meanwhile, new discoveries such as Spindletop were being made largely
on the basis of surface anticline structure, seeps, and geologic intuition.
Failures (the dry holes) were attributed to a variety of causes, including,

135
Geologic Applications of Gravity and Magnetics: Case Histories
06 State of the Art REV 1/29/09 12:37 PM Page 136
according to one historian, “the graves of cattle who had died of a hoof
and mouth disease.”
Following the close of World War I, word of Eötvös’ success rapidly
reached the United States and by 1922, Eötvös’ balances were imported
by Shell and Amerada. Late that year, an experimental survey across
Spindletop made it clear that the stru c t u re could be detected by the
instrument. The first discovery by the torsion balance, the Nash Dome,
was made in 1924 by Amerada. The map was a beautifully simple circle
of arrows pointing toward the center of the dome (Figure 3). The geo-
physical exploration industry was born in the United States.
A long string of successes followed, and contracting companies
appeared almost immediately. For a time, the torsion balance had no com-
petition whatsoever in the search for oil and gas. In the next 10 or so
years, the discovery of more than 1 billion barrels of oil and at least sev-
enty-nine producing structures was attributed to the application of this
instrumentation.
The instruments generally were about 2 m in height and were iso-
lated from the elements by a collapsible building resembling an outhouse
or a large tent (Figure 4). The most advanced design required 20 minutes
for an individual reading, a significant improvement over the several hours
Figure 3. The first hydrocarbon prospect, the Nash Dome, located with geophysics
in the United States. This prospect was identified by the distinctive suite of arrows
pointing toward the center of dense cap rock atop the salt dome (from “The Eötvös
torsion balance method of mapping geologic structure” by D. C. Barton, Geophysical
Prospecting, 1929).
State of the Art
that earlier designs needed. Unfortunately, the ground surrounding a tor-
sion-balance measurement had to be leveled in eight directions, often out
to 100 m. This obviously took a lot of time, and the resulting star-shaped
p a t t e rn left in vegetated areas never would have passed today’s environ-
mental regulations. This leveling to reduce near-zone terrain effects and
erecting the shelter meant that each station re q u i red, at the least, a few
hours to observe. The instrument was very sensitive to near effects, includ-
ing belt buckles and metal in the observ e r’s pockets, old cellar holes, and
o v e rhead wires. Despite these complications, measurements were excel-
lent quality even by today’s standards (resolution of about 1 Eötvös).
Interpretive tools for gradiometry data were slow to develop. The
early identification of salt domes and cap rocks was strikingly simple —
136
arrows resulting from the data pointed toward the salt dome. However,
these arrows became difficult to interpret over more complex structures.
Despite arguments in the literature about developing tools and visualiza-
tions which used all the gradient data recorded, the measurements were
increasingly integrated into gravity maps. This failure to develop inter-
pretive tools would contribute to the marginalization of gradiometry.
By the late 1920s, refraction seismology appeared on the scene.
Reflection seismology followed soon after. Curiously, the perceived roles of
gravity and seismology were very similar to those of today, with gravity re c-
ommended for reconnaissance work and seismology for detailed follow-
up. This assignment of responsibility was also as controversial as it is today.
The year 1930 brought both the Great Depression and the initial field
deployment of pendulum gravity meters. Although awkward to use and
not very accurate, the pendulum gravity meter was significantly faster
than the torsion balance and yielded quantities which most geologists
found easier to interpret. The trend of integrating the gravity gradient
readings to give results which could be compared directly with the grav-
ity meter continued. Finally, in the mid-1930s, the first modern astatized
spring gravity meters appeared. These were much smaller and more accu-
rate than the pendulum meter, and much faster and less sensitive to near-
Figure 4. Two field enclosures used with torsion-balance gradiometers: (a) domelike
structure used in Britain (from “The theory and practical use of the Eötvös Torsion
Balance” by Shaw and Lancaster-Jones, Mining Magazine, 1927) and (b) cloth-sided
shelter used widely in the U.S. Gulf Coast, with a Model T for scale (from Exploration
Geophysics by J. J. Jakosky, 1940).
06 State of the Art 4/16/07 4:57 PM Page 137
zone terrain than the torsion balance. This last feature made the new
meters useful in terrain where the torsion balance did not give usable
results. The new meters rapidly supplanted the torsion balance in U.S. oil
exploration (Figure 5). Despite the protests of some who claimed that the
new instrument had “lower resolving power,” the torsion balance was a
museum piece in the United States by 1940, although its use lingered in
eastern Europe for much longer.
However, the use of gradient measurements never disappeared com-
pletely. Through the 1950s and 1960s, enterprising geophysicists
attempted numerous experiments, mostly with vertical gradiometers syn-
thesized from paired gravimeter measurements. Most of these experi-
ments were aimed at mineral or geotechnical applications. Finally, in the
late 1960s came the realization that gradiometers had great potential in
moving-platform applications because of the absence of vertical heave and
Eötvös errors, and the modern era of gradiometers was born. Today, the
development of interpretive tools for gradiometry data is the challenge of
a new era of gradiometry.
Figure 5. The transition from torsion-balance gradiometry to gravity measurements
during the 1930s in the Gulf Coast region. The plot shows the number of geophysical
field parties active during each quarter from 1935 to 1939 (from “A brief history of
the gravity method of prospecting for oil” by E. A. Eckhart, Geophysics, 1940).
Results from a High-resolution, 3-D Marine
Gravity Gradiometry Survey over a Buried
Salt Structure, Mississippi Canyon Area,
Gulf of Mexico
Lincoln F. Pratson*, Robin E. Bell, Roger N. Anderson
Lamont-Doherty Earth Observatory of Columbia University
Palisades, New York, USA
Dan Dosch, John White, Clive Affleck,
Andrew Grierson
Bell Aerospace Co., Buffalo, New York, USA
Bryant E. Korn, Ronald L. Phair
Texaco Exploration & Production Inc.
New Orleans, Louisiana, USA
E. K. Biegert
Shell Exploration and Technology Company
Houston, Texas, USA
Peter E. Gale
Shell Offshore Inc., New Orleans, Louisiana, USA
*Now at Institute of Arctic and Alpine Research, University of Colorado,
Boulder, Colorado, USA
EDITOR’S NOTE: This is an important paper about a new technol-
ogy and a new gravity-acquisition system. The gradiometer
was developed at great expense and effort by the U.S. gov-
ernment. Developed and installed on nuclear submarines for
stealth navigation, this technology allows the sub to “feel” a
3-D gravity effect in a totally isolated and quiet mode.
Recently declassified, this extremely sensitive method of
measurement of gravity can now be applied to exploration
problems. This paper is a first step in understanding how to
interpret an exciting new data type.
Geologic Applications of Gravity and Magnetics: Case Histories
Abstract
The first test of the Bell Aerospace gravity Gradiometry Survey System
(GSS) for geologic applications was conducted in April 1994 in collabora-
tion with the U.S. Navy. The GSS is a recently declassified gravity sensing
system that contains the world’s only moving-base gravity gradiometer.
The system measures both gravitational acceleration and gravity gradi-
ents, yielding six measurements that define the local gravity field and its
gradients in three dimensions—a technologic advance in measuring grav-
ity analogous to the advance from 2-D to 3-D seismic profiling through
the towing of multiple rather than single hydrophone arrays. The gravity
gradiometry test survey was conducted over a buried salt structure south-
southeast of New Orleans in water depths of ~1500 m. The quality of the
survey data is excellent. In declassified grids of the data at 2-km wave-
lengths, gravity gradients are resolved to 0.5 Eötvös and gravity to 0.07
mGal. Simple models are used to illustrate the power of this data in sub-
surface structure definition. The potential utility of gravity gradiometry in
oil and gas exploration then is demonstrated through application of the
survey data in improving a geologic model of a part of the survey area
derived from 3-D seismic data.
Introduction
Even with the best 3-D seismic aquistion and processing techniques, the
geometry of salt structures and their encasing strata can be difficult to
resolve. The acoustic properties of salt and the often complex geometries
of salt bodies tend to attenuate and distort seismic reflections from the
edges of salt and from subsalt interfaces.
A natural data type for validating and improving seismic interpreta-
tions of salt structures is high-resolution gravity measurements. Both seis-
mic and gravity signals are dependent on sediment density. A geologic
model of an area derived from seismic data can be used to forward model
gravity anomalies against which measured gravity can be compared.
Similarly, inverse modeling using gravity measurements can constrain the
depth and morphology of density contrast surfaces poorly resolved in seis-
mic reflection data.
An instrument for aquiring high-resolution gravity measurements new
to oil and gas exploration is the gravity Gradiometry Survey System (GSS).
The GSS is a recently declassified military gravity sensor developed by
Bell Aerospace, a division of Lockheed-Martin. It contains two gravime-
ters and the world’s only moving-base gravity gradiometer (Jekeli, 1988).
This paper presents results from the first test of the GSS for use in
hydrocarbon exploration. The test was conducted in April 1994 in col-
laboration with the U.S. Navy. The GSS was used to survey a buried salt
137
06 State of the Art 4/16/07 4:57 PM Page 138

State of the Art

138
structure located south-southeast of New Orleans in water depths of The gradient of gravitational acceleration (or curvature/second deriv-
~1500 m (Figure 1). We discuss the accuracy of the data collected dur- ative of the potential field) is a nine-component tensor, expressed mathe- large
De

g
ing this survey and its utility in delineating buried structures. We then matically as tec shallow

=
tab masses

G
illustrate some of its potential applications in oil and gas exploration by 0.5

m
g' α le

r
⎡Txx Txy Txz ⎤

-2
using the gravity gradiometry data to improve a geologic model of a part by Detectable by
⎢ ⎥ Gm GS
of the survey area constructed from 3-D seismic data. We start, how- ∇T( x , y , z ) = ⎢Tyx Tyy Tyz ⎥ r -3
gra
SGSS gradiometer and
ever, by reviewing gravity gradiometry and its relation to gravimetry. 1 state-of-the-art
⎢⎣Tzx Tzy Tzz ⎥⎦ dio

What is Gravity Gradiometry?
Gravitational acceleration is the gradient (i.e., first derivative) of the grav-
ity potential field. In cartesian coordinates, gravitational acceleration,
denoted here as T, is a vector with components in the x (east-west), y
(north-south), and z (up-down) directions. Gravimeters measure the
amplitude of gravitational acceleration in the z direction (i.e., Tz), per-
pendicular to the theoretical, equipotential sea-level surface enveloping
the Earth referred to as the geoid. The measurements generally are
reported in units of milligals, where 1 mGal = 0.001 cm s–2. They are
used in the construction of free-air and Bouger anomaly maps.
Figure 1. Survey tracks overlaid on local bathymetry highlighted by false illumination.
White square box is area encompassed by geologic model shown in Figure 7. Bathy-
metric contours are in meters below sea level. Boundaries between black-and-white
bars framing image represent distance increments of 0.1° longitude in the horizontal
direction and 0.1° latitude in the vertical direction.
Because this tensor is symmetric (e.g., Txy = Tyx) and because the grav-
ity potential field honors Laplace’s equation (i.e., Txx + Tyy + Tzz = 0),
five of these gradients are unique (underlined) and can be used to deter-
mine the remaining four.
Gravity gradients can be derived from gravimeter measurements.
Derived gradients, however, are subject to three important limitations: (1)
The differentiation needed to compute the gradients elevates noise inher-
ent in gravimeter measurements and geologic information is lost, a con-
sequence that becomes significantly detrimental at finer scales of geologic
interest; (2) derived gradients estimate only the longer, smoother-wave-
length components of the gravity field, adding to the information loss at
finer scales; and (3) derived gradients are not independent measure-
ments, and therefore provide no independent information about varia-
tions in the gravity field.
Gravity gradiometry is the direct measurement of gravity gradients.
Gradient measurements are reported in units of Eötvös, where
1E = 10–9 s–2 or 0.1 mGal km–1. The GSS contains the only gradiometer
to have measured gravity gradients successfully from a moving platform
outside a laboratory environment.
GSS Gradiometry Detection Capabilities
Resolution as a function of depth/mass
Gravity gradiometers and gravimeters have complementary detection
capabilities. The detection crossover point—the depth at which a
causative body is just resolvable by both a gradiometer and a gravi-
meter—is independent of the mass of the body and dependent only on
the relative resolution (noise floor) of the two instruments (Figure 2).
Because of security concerns, the U.S. Department of Defense would
allow only the release of gridded GSS data at the time of this study. The
accuracy of the declassified grids, however, is very high. Mis-tie and spec-
tral analysis indicate that at wavelengths of 2 km, the gridded GSS gra-
diometry data resolve gravity gradients to 0.5 E. This means that the
gridded gradiometry data can detect smaller and/or deeper masses than
state-of-the-art gravimetry having a resolution of 0.2 mGal at 2-km wave-
lengths up to a depth of ~6 km (Figure 2).
Depth (km)
me gravimetry
ter
Not detectable by
either system
5
detection crossover point
Small deep
masses
10
1000 10 000 100 000 1 000 000
Minimum detectable anomalous mass (1000 metric tons)
Figure 2. Theoretical plot of the minimum mass detectable with the gridded GSS
gradiometry data as function of depth. Plot is based on a resolution of 0.5 E at 2-km
wavelengths determined from the test survey results. Also shown for comparison is a
theoretical plot of the minimum mass detectable with state-of-the-art gravimetry data
having a resolution of 0.2 mGal at 2-km wavelengths. Minimum mass detectable with
GSS gradiometry data increases with depth cubed. Mimimum mass detectable with
state-of-the-art gravimetry data increases with depth squared.
Edge detection
A principal advantage of using gravity gradients for subsurface exploration
is that they are excellent edge detectors and corner locators. As an illus-
tration, in the absence of noise, the gravity anomaly caused by a buried
cube is a broad bull’s-eye (Figure 3). The gradient anomalies, on the other
hand, constrain the size and shape of the cube, with individual gradients
highlighting different cube boundaries (Figure 3). For example, Txx and
Tyy delineate the north-south and east-west edges of the cube, and Txy
delineates the cube corners (Figure 3).
Gravity gradients are good at detecting not only sharp edges but also
rounded edges. Figure 4 shows the gravity and gradient anomalies asso-
ciated with a model of a simplistic ellipsoidal salt sheet and cylindrical
feeder stock. In this illustration, the gradient anomalies are simulated to
have the same resolution as the test survey data (i.e., 0.5 E @ 2-km wave-
lengths), and the resolution of the gravity anomaly is simulated to be
equivalent to that of state-of-the-art gravimetry (0.2 mGal @ 2-km wave-
06 State of the Art 4/16/07 4:57 PM Page 139

Txx -2.5 Txy 0.0 Txz 0.0

0.0 0.5 Txx Txy -0.5 0.0 0.5 Txz
0.0 0.0 -2.5 2.5 0.0
1.0 1.0
-7.5 5.0 -5.0
-5.0 5.0
2.5 2.5 10.0 -10.0 0.5
-15.0 -12.5
15.0 1.5 1.5
0.0 0.0 0.0
0.0
-15.0 -0.5
-0.5 0.0
5.0 -5.0
0.0
-5.0 0.0
2.5 -2.5
0.5
0.0 Eötvös Eötvös Eötvös
Eötvös A A'
0
Tyy 2.5
Tyz 0

1
Sediment Layer 1, ρ = 2.075
seafloor
0.0 0.5
Tyy 0.5 Tyz
7.5 1 1.0
-5.0 Salt, ρ = 2.15 A

km below sea level

0.0 -10.0 1.0 1.5
2
-15.0 2 -0.5
Sediment Layer 2,

km
3 0.0 -1.0
ρ = 2.225

km
5.0 0.0 3
-2.5 -7.5 -15.0 -12.5 -5.0 0.0 1.0 0.0
4 -0.5 0.5
0.5 4 -0.5
15.0 0.0
10.0 0.5
0.0 2.5 5.0 5 1.0
A' 5
2.5 6
0.0 Eötvös 6
0.0
0.0 Eötvös
0 1 2 3 4 5 6
mGal km
0.0 2.5 5.0 7.5 0 1 2 3 4 5 6
Tz km Tzz -0.2 Tz km
-0.5

-1.0
Tzz
-0.3 -1.5
-0.4

0.0
-0.7
-0.6 -0.3
3.0 20.0 -0.5
5.0
2.0 -0.2
1.0 0.0 1.0
-0.3 -0.1 0.1
-0.0 0.5
-0.2 -0.1 0.0

-0.1 mGal 0.0 Eötvös

Figure 3. Gravity and gradient anomalies associated with buried cube 3049 m (10 000 ft) on a side and having density contrast Figure 4. Gravity and gradient anomalies associated with simplistic model of ellipsoidal salt sheet and cylindrical salt stock.
with surrounding medium of 1.0 g cm–3. The five independent gradient anomalies plus Tzz are shown, arranged as they appear in Cross-section of model shown in left center of figure. Model densities (ρ) are in g cm-3. Contours for gradient anomalies are in
gravity gradient tensor. Contours in gradient anomalies are in Eötvös. Contours in gravity anomaly are in milligals. Gravity anom- Eötvös. Contours for gravity anomaly are in milligals. Black circle and ellipse in anomalies are salt stock and sheet perimeter,
aly (Tz) shown in place of Tzx. Black square in anomalies is cube outline. respectively.

lengths). Note how the different gradient anomalies highlight different
boundaries of the salt structure, constraining its form more accurately
than the gravity anomaly (Figure 4). Combinations of the gradients also
provide unique information. For example, when viewed together, the hor-
izontal gradients (Txx, Txy, and Tyy) highlight the perimeter of the salt
sheet (Figure 4).
Surface estimation
The GSS gradiometry data is optimized when integrated with the
gravimetry data. Figure 2 indicates that when used together, shallow
and/or short-wavelength mass variations are sensed by the gradiometer,
while deep and/or long-wavelength mass variations are sensed by the sys-
tem’s two gravimeters.
Integration of the gradiometry and gravimetry data is particularly
beneficial when using an inverse modeling approach to estimate the mor-
phology of a subsurface structure. An example of a need for this approach
is in resolving base-of-salt horizons, which often are imaged poorly in seis-
mic data. Figure 5a shows a hypothetical base-of-salt horizon beginning at
1372 m (4500 ft) below sea level, possessing many of the attributes that
make seismic imaging of the bottoms of salt structures difficult: thick sec-
tions (>1 km), steep walls ≤ 25°), and an irregular surface.
Figures 5b and 5c show attempts to reconstruct this horizon from the
gravity and gradient anomalies it would cause (see Appendix for details).
Figure 5b is the result when only the gravity anomaly due to this surface,
simulated to the resolution of state-of-the-art gravimetry, is used in the in-
version. This surface reflects the general assymetry of the original base-of-salt
horizon, but fails to capture its relief (< 500 m), slopes (≤ 8°), and notable
surface irregularities (e.g., bulb at southeastern terminus of horizon).
Figure 5c is the result when the gravity and gradient anomalies, sim-
ulated to the resolution of the GSS-gridded gravimetry and gradiometry
data, are used in the inversion. Mis-tie and spectral analysis indicate the
resolution of the gridded GSS gravimetry data is very high, 0.07 mGal at

139
Geologic Applications of Gravity and Magnetics: Case Histories
06 State of the Art REV 1/29/09 12:37 PM Page 140

State of the Art

140

a b c

Figure 5. (a) Hypothetical base of salt horizon having a density contrast with surrounding sediments of 0.075 g cm-3. (b) Result from an inversion for (a) using its gravity anomaly simulated to have a resolution of 0.2 mGal at 2-km wavelengths; see text for discussion.
(c) Result from an inversion for (a) using its gravity and gradient anomalies simulated to have gridded GSS data resolutions of 0.07 mGal and 0.5 E at 2-km wavelengths, respectively; see text for discussion.

2-km wavelengths. However, much of the surface detail recovered in this
inversion results from the inclusion of the gradients, which also are mea-
sured to a high, but independent, accuracy. The result of the inversion is
a surface that captures the proper relief, steep slopes, and most of the
irregularities of the original horizon.
GSS Test Survey and Results
Survey target
The GSS gravity gradiometry test survey was conducted over a larg e ,
buried salt overhang south-southwest of New Orleans, in water depths of
~1500 m (Figure 1). The salt structure surveyed (Figure 6) is similar in
dimensions and shape to the simple salt structures modeled above.
However, it is significantly more complex. The central portion of the
structure has been depressed and thinned by sediment loading (Figure 6),
during the Pliocene. Variations in salt thickness (Figure 7) show that this
loading was focused along the west side of the structure, which, in
squeezing salt from beneath it, led to inflation and thrusting of salt to the
east. The Lentic I horizon (Figure 7), a regional stratigraphic marker
located above the salt structure, parallels bathymetry in this area, indicat-
ing that movement of the salt pretty much ceased in the late Pliocene.
A 3-D geologic model of the southern half of the salt stru c t u re (Figures
1 and 7), including the encasing strata, was provided by Te x a c o
Exploration & Production Inc. and Shell Off s h o re Inc. prior to the GSS sur-
vey. The model consisted of stratigraphic horizons interpreted from 3-D
seismic data collected by We s t e rnGeophysical. The oil companies also pro-
vided information needed to constrain the regional gravity field and the
density of the stratigraphic units in the geologic model. The density infor-
mation included a density log (Figure 8) from the nearest well, appro x i-
mately 15 km northwest of the survey area. The log showed that sediment
densities were, to first ord e r, constant within the model units (Figure 8). A
single, average density was assigned to each model unit. The geologic
model and additional information then were used to compute the gravity
and gradient anomalies predicted for the southern third of the survey are a .
Survey execution
The GSS survey over the salt structure was conducted by the U.S. Navy.
The survey was carried out within a rectangular area 27 km long by 7 km
wide (Figure 1). The long axis of the area was aligned northwest-south-
06 State of the Art 4/16/07 4:57 PM Page 141

Figure 6. Seismic reflection profile of salt struc-

0 1 ture surveyed with GSS (adapted from Hodgkins Shale density (g cm –3)
km and O’Brien, 1994). Profile lies along long axis of
1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6
survey area shown in Figure 1.

1800

east along the trend of the salt structure.

2000
The survey was run at a speed of 10 knots.
Track lines, oriented northwest-southeast
and northeast-southwest, were spaced
2200
500 m apart. The track-line spacing was
significantly denser than needed to resolve
the targeted structure, but was designed to 2400
allow for systematic evaluation of optimal
track-line spacings for future GSS surveys.
2600

Time (ms)
Results
SE NW
Over most of the model area surveyed
2800
(Figure 1), the trends and magnitudes of
SE the measured gravity and gradient anomalies (Figure 9) compared
NE extremely well to those computed from the geologic model of this region
3000
SW (Figure 10). The excellent comparisons verified the high accuracy to
0 which the GSS measures gravity gradients.
sea floor
In one quadrant in the model area, however, a major discrepancy 3200
-1
occurred between measured and calculated anomalies. This discrepancy
Lentic I was highlighted when the calculated gradients and gravity anomalies
-2
(Figure 10) were differenced from the measured anomalies (Figure 9) to 3400
-3 produce residual anomalies (Figure 11). The residual anomalies indicated
km below sea level

salt top the signature of the error. Closed, circular contours in maps of the Tz, Tzz,
-4 and Tyy residual anomalies (Figure 11) indicated that excess mass existed 3600
in the southwestern part of the model area.
-5
Model error identification
-6 Figure 8. Shale density and velocity logs from well 15 km northwest of geologic model
The gradient and gravity anomalies for each horizon in the geologic area (box, Figure 1). Note that up to 3200 milliseconds, density increases in steps.
-7 model were evaluated separately to isolate the source of the model error
producing the large, bull’s-eye residual anomaly. A major error in the map-
9 -8
8 ping of the base-of-salt horizon or one of the deeper horizons in this area
7
6
was considered doubtful after modeling showed that the former would
5 -9 need to be elevated by more than a kilometer to compensate for the
4
km 3 1 0
2 3 2 Figure 7. Model of southern third of salt structure (box, Figure 1) consisting of strati- residual anomaly. Such an adjustment would have required an unreason-
1 5 4
0 7 6 graphic horizons interpreted from 3-D seismic data. Horizons: red = sea floor; orange able mistake in the processing or interpretation of the 3-D seismic data
8 km = Lentic I; purple = top of salt; light green = base of salt; light blue and dark blue = used to construct the model. A comparably large, localized increase in the
subsalt 1 and 2; purple = mid-Cretaceous unconformity.

141
Geologic Applications of Gravity and Magnetics: Case Histories
06 State of the Art 4/16/07 4:57 PM Page 142

State of the Art

142

Txx -13 Txy Txz 11

Txx Txy 5
6 Txz 7
-13
7 5
8 -14
0
7 -6
4 -13 -7 0 -13
-10 6 -8 -14
-8 5
0 -27
-15 4 -9
4 -30 4
-12 3 -1
-14 -7 0
14 5 14 -2
-3 13 -3
10 -4
Eötvös -4 -17
Eötvös
0 900 0
600
300
900
1200
9
8
Tyy 2
9 Tyz 600 3000
2000
300
1000
1200
4000
9
8
Tyy 0 -1 9 Tyz -4
8 -9 -6 8 -6 0
7 -4
-1
7
0
6 7 0 6 7

km
km
5 6 5 6

km
km
4 -8 5 4 -11 5
-11 -15
3 4 3 -14 -12 4
-16 -14 15 -13
2 2 18
0 3 0 0 3
1 12 1 2
Salt isopach (meters) -3 2 Salt isopach (meters) 1
0 -2 -1 0 2 21
1 32
1
0 1 2 3 4 5 6 7 8 9 0 0 1 2 3 4 5 6 7 8 9 23 25
0 0
km km
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9
Tz km Tzz Tz km Tzz 14
17

9
8 9
7 24 8 30
6 7 21
31 6
0 5
4 0
-13 0 -15
0
-7
-7 -5 -6
-6 -10 -14
mGal mGal

Figure 9. Gridded GSS gravity and gradient anomalies measured in model area. Contours for gradient anomalies are in Eötvös. Figure 10. Gridded gravity and gradient anomalies computed for survey area from model shown in Figure 7. Contours for gradient
Contours for gravity anomaly are in milligals. Salt isopach map shown in left center of figure for reference. anomalies are in Eötvös. Contours for gravity anomaly are in milligals. Salt isopach map shown in left center of figure for reference.

density of the subsalt units also was considered unlikely, based on the This left an unmodeled subregion of relatively dense material as the surface was approximated by a simple, inverted-bowl-shaped horizon
regional stratigraphy. This left three more feasible possibilities as to the most likely cause for the majority of the bull’s-eye residual anomaly. This (Figure 13a). The two seismic profiles which partially imaged the wedge
source of the bull’s-eye residual anomaly: (1) excess salt in the south- interpretation was confirmed when it was learned that the geologic (Figure 12) were used to predict an appropriate relief and diameter for
western quarter of the geologic model, (2) a subregion of relatively dense model lacked inclusion of a horizon for a localized wedge of older, denser this bowl. The seismic profiles, however, did not constrain the location of
material above the salt, or (3) a combination of excess salt overlain by a sediments overlying the salt in the western part of the model area. Two the wedge apex. Consequently, the gradient and gravity residual anom-
region of relatively dense material. seismic reflection profiles provided by Western Geophysical reveal the dis- alies (Figure 11) were analyzed for information on where to locate the
The first possibility was evaluated through simple 2-D modeling of the position of this wedge on top of the salt (Figure 12). inverted-bowl horizon within the model area (Figure 1).
gravity and gradient anomalies caused by a tabular salt body. The results Closer examination of the gradient residual anomalies revealed that
of this modeling showed that salt thickness in the the southwestern quad- Forward model of missing structure the maximum Tzz and Tyy anomalies both are located approximately 1
rant of the model would have had to be in excess of more than 2 km to Two attempts were made to use the gradient and gravity anomalies to km southwest of the maximum gravity anomaly (Figure 11). Simple 2-D
compensate for the residual anomaly. Because the salt in this region of the reconstruct the wedge and improve the geologic model. In the first modeling of the gravity and vertical gradient anomalies associated with a
model is interpreted to be only 150 m thick (Figures 9–11) on average, attempt, a forward modeling approach was used. The wedge was relatively dense, skewed, triangular body explained this observed offset
the first and third possibilities were discounted. assigned a density equivalent to the first subsalt unit (Figure 7), and its (Figure 14). It was found that although the maximum gravity anomaly is
06 State of the Art 4/16/07 4:57 PM Page 143

Txx 2 Txy 2
Txz 5
0
0
2
-2

-1 1
0
0 10
0 -1
-1 -1
1 0
0 -5
-2 -1 -1
0 1 0
-1 0
0
Eötvös
0
600 900
3000
2000
300
1000 4000
1200
9
8
Tyy 3
9 Tyz -4

7 5
8
6 7 -3
5 km 6 -2
1

km
4 5
1
3 0 4
-1
2 3
0 0
1 -4
Salt isopach (meters) 2
0 2
0
-4 1
0 1 2 3 4 5 6 7 8 9 -2 9 6
0
km
0 1 2 3 4 5 6 7 8 9
Tz -1.0 km Tzz -5
-3
-0.4
0.0
-2
0.4 -1

0.8
0.6 6 2
3 4
0.4
0.0 mGal

Figure 11. Residual gravity anomalies from differencing computed anomalies (Figure 10) from measured anomalies (Figure 9).

located over the triangle’s center of mass, the vertical gradient anomaly
peaks over its apex (Figure 14). This showed that the best match to the
residual anomalies would be achieved by placing the inverted-bowl hori-
zon in the geologic model so that its apex coincided with the maximum
Tzz and Tyy residual anomalies (Figure 11). Thus the gradients provided
crucial information in constraining the geometry of the missing horizon.
The construction of the inverted-bowl horizon was based on the
assumption that the gradient and gravity residual anomalies were caused
solely by the absence of the sediment wedge in the geologic model. If this
assumption was correct and the shape and density contrast of the inverted-
bowl horizon accurately reflected the mass of the missing sediment wedge,
then the gradient and gravity anomalies associated with the inverted-bowl
horizon should have matched the residual anomalies exactly.
Figure 12. Two seismic profiles over sediment wedge that was not included in model shown in Figure 7.

143
Geologic Applications of Gravity and Magnetics: Case Histories
06 State of the Art 4/16/07 4:57 PM Page 144

State of the Art

144
sediment wedge had a constant density equivalent to the first subsalt unit,
5 and (3) the sediment wedge did not extend beyond the eastern boundary
a 9
4 of the salt structure.
8 3 Figure 13b shows the horizon generated using the type of inversion
2 scheme outlined in the Appendix. The horizon has the same general relief
7 1 and apex location as the inverted-bowl horizon solved for in the forward
Txz (E)

mGal, km, Eötvös

0
Tzz (E)
modeling approach (compare Figures 13a and 13b). However, this new
6
-1 horizon (Figure 13b) extends more to the east and north and its apex is
5 -2 more skewed, a characteristic predicted by the simple triangular body
used to understand the offset between the T and Tzz-Tyy anomaly max-
km

-2125 -3 Sediment wedge (km)

4 -4 ima (Figure 14).

-1825
-5 Which horizon, the forward model or inverse model, more accurately
3
-1600
-6 Tz (mGal) reflects the true structure of the missing sediment wedge will be deter-
2 -7 mined with the addition of future seismic interpretations. However, the
-2650 0 5 10 15 20
results from the forward and inverse modeling of the residual gravity gradi-
km
1 ent anomalies already have improved the geologic model used in this study.
0 Conclusions
Figure 14. A 2-D model of gravity and Tzz and Tzx gradient anomalies over relatively
0 1 2 3 4 5 6 7 8 9 dense, skewed triangular sediment wedge. Note that gravity anomaly peaks over Interpretation of gravity gradiometry for exploration purposes still is
wedge center of mass (dotted line), and Txz and Tzz anomalies peak over wedge apex
b 9 -2650
(dashed line). developing. However, this case study of gravity gradiometry integrated
8
with a seismically derived geologic model over a complex salt structure
already clearly demonstrates that:
7
1) Gradient and gravity anomalies measured using the Bell Aerospace
-2125 Comparison of Figures 15 and 11 shows that the match of the Tzz gravity Gradiometry Survey System (GSS) provide an independent,
6
anomalies was the closest, both in shape and magnitude. The Tyy and Txx multifold constraint on geologic models derived from seismic inter-
5 -1825 anomalies were similar in shape but not in magnitude. The other gradient pretations.
km

anomalies and the gravity anomaly showed more marked differences.

4
The differences between the gravity and gradient residual anomalies
3
(Figure 11) and the modeled anomalies (Figure 15) indicate that the
-1600
inverted-bowl horizon was a good first-order estimate of the shape of the
2 missing structure. A more accurate forward model of the sediment wedge
could have been achieved by refining the shape of the inverted-bowl hori-
1 -1825 zon until all the residual gradient and gravity anomalies were matched.
0
The excercise clearly demonstrated that measured gradient anomalies
represent powerful independent constraints in the forward modeling of
0 1 2 3 4 5 6 7 8 9
subsurface structures.
km
Inverse model of the missing structure
Figure 13. (a) Structure map of inverted-bowl-shaped horizon used to estimate sedi-
ment wedge in Figure 12 through forward modeling of residual anomalies shown in In the second attempt to reconstruct the missing sediment wedge from
Figure 11. (b) Structure map of result from gravity and gradient inversion for sedi- the gravity and gradient anomalies, an inverse modeling approach was
ment wedge in Figure 12 using residual anomalies in Figure 11. Contours are in used. Three important assumptions were inherent in this approach: (1) All
meters below sea level.
of the residual anomaly signal was caused by the sediment wedge, (2) the
2) Gradient and gravity anomaly measurements can be used to identify
deficiencies in seismic-based models and constrain their causes.
3) Gradient and gravity anomaly measurements also can be used to
improve seismic-based models through forward and inverse modeling
techniques.
Acknowledgments
We thank Pat Millegan and Dick Gibson for the opportunity to contribute
to this special volume, as well as several anonymous reviewers for their
efforts in improving the manuscript. We also thank Bell Geospace Inc. for
permission to analyze the gravity gradiometry data highlighted in this
study, and Texaco Exploration & Production Inc. for the interpreted seis-
mic surfaces used to model the subsurface structure in the area surveyed.
Finally, we thank Western Geophysical for permission to use the two seis-
mic lines shown in Figure 12.
06 State of the Art 4/16/07 4:57 PM Page 145

Gravity Gradiometry
Txx Txy Txz E. K. Biegert
2 6
Shell Exploration and Technology Company
3
1 Houston, Texas, USA
5 2
0 -4 1
-1
-2 0 Gravity gradiometry was one of the first
-3
3 -1 geophysical measurements used in petro-
2
1
4 leum exploration early in this century,
but with the development of accurate
Eötvös
relative gravity meters, gradiometer mea-
Tyy 9 Tyz surements have not been used for geo-
8 logic studies until recently. Conceptually,
7 the fundamental measurement is the dif-
1
6 ference between a matched pair of
-1

km
0
5 -3 -2 accelerometers. Since they are linked
-1 -4
4
-2
3
together physically, common mode accel-
0
-3
2 1
erations are eliminated, resulting in a
1 measurement that is not sensitive to
2
mGal
0 local platform motion. In a moving-plat-
0 1 2 3 4 5 6 7 8 9 form environment, short-wavelength
Tz -2.0 km Tzz noises are attenuated significantly in the
-4
-1.0 -3 gradiometer output. Once converted
0.0 -2 1
-1
2 back to conventional gravity by integra-
3
0.2 0
4
4
tion, the gradient signal can be combined
3
5 2
optimally with the conventional gravity
0.4 6 1 signal to extend the bandwidth to the
0.2
0.0
shorter wavelengths. This means that a
gradiometer system is suited ideally for
acquiring data in rough seas or from air-
Figure 15. Anomalies associated with forward model of sediment wedge shown in Figure 12a. Compare with residual anomalies craft. The gradients it measures provide
shown in Figure 11. valuable cross-line information about the
causative body, much in the way a swath
of seismic data provides cross-line infor-
References accuracy of the GSS gravimetry data after gridding, which mis-tie and mation that a single 2-D seismic line
Hodgkins, M. A., and O’Brien, M. J., 1994, Salt sill deformation and its implications for spectral analysis indicated was 0.07 mGal at 2-km wavelengths. The grids lacks. This greatly enhances the informa-
subsalt exploration: The Leading Edge, 13, 849–857. of the gradient anomalies were simulated to the test survey accuracies as tion content of a gradiometer survey over
Jekeli, C., 1988, The Gravity Gradiometry Survey System (GGSS): EOS, 69, 116–117. well (i.e., 0.5 E at 2-km wavelengths). a conventional gravity survey, especially
The computed anomalies then were used in an inverse approach to in reconnaissance surveys with widely
Appendix reconstruct the hypothetical base-of-salt horizon. In the inversion, the spaced traverses.
The inverse modeling results in Figures 5b and 5c were arrived at as same density contrast (0.075 g cm–3) across the horizon was used, and
follows: Grids of the gravity and gradient anomalies were computed at a the edges of the horizon were constrained to lie at 1372 m (4500 ft)
uniform grid-cell spacing of 500 m, assuming a density contrast across the below sea level. An iterative solution was used to converge on a surface
horizon of 0.075 g cm–3. Two grids of the gravity anomaly were gener- that produced gravity and/or gradient anomalies that best matched the
ated. In one, the gravity anomaly was simulated to an accuracy equiva- computed anomalies within the limits of their simulated resolutions.
lent to state-of-the-art gravimetry data having a resolution of 0.2 mGal at
2-km wavelengths. In the other, the gravity anomaly was simulated to the

145
Geologic Applications of Gravity and Magnetics: Case Histories
06 State of the Art 4/16/07 4:57 PM Page 146
The Role of Vertical Magnetic Gradiometry
in Hydrocarbon Exploration
H. O. Seigel and T. J. McConnell
Scintrex Limited, Concord, Ontario, Canada
EDITOR’S NOTE: Vertical magnetic gradiometry has the potential to
provide better definition of geologic features and of cultural
noise. The interpretation of subtle intrasedimentary effects is
dependent on the proper sampling of the anomaly and on
proper identification and removal of cultural anomalies.
Introduction
The application of vertical magnetic gradiometer methods to hydrocarbon
exploration was initiated some thirty years ago, by Pure Oil Co. (now
Union Oil) and Aero Service Corp., using both helicopters and fixed-wing
aircraft (Slack et al., 1967).
Both types of installation used two total-field sensors, with a vertical
separation of 100 ft (30 m). For the helicopter version, both sensors were
suspended on tow cables. For the fixed wing, one sensor was in a tail
stinger and the second was towed below the aircraft. The magnetic inter-
ference of the aircraft on the stinger sensor was compensated for by pas-
sive means.
The objectives behind the use of aeromagnetic gradiometry at that
time were:
1) elimination of the effect of diurnal activity, particularly in northern
latitudes, and
2) derivation of additional geophysical information regarding the source
of magnetic anomalies
The efficacy of these early gradiometer attempts was limited by sev-
eral factors, namely
1) noise introduced by the uncontrolled geometry of the towed bird sen-
sor (or sensors). This could amount to as much as 25 pT/m (verti-
cally) and 5 pT/m horizontally (1pT = 10–3 nT; 1nT = 1 gamma).
2) imperfect compensation of aircraft motion noise on the stinger-
mounted sensor, typically several nT at that time, and
State of the Art
146
3) limiting sensor and instrumental noise levels of the dual-cell rubidium The Merits of Vertical Gradient Information
magnetometers used at that time. These were typically of the order of
50 pT per sensor, or 75–100 pT on the gradiometer.
Vertical gradiometry can enhance the interpretability of aeromagnetic sur-
veys in several ways. For one thing, it provides a better definition than do
the total-field data of such geologic features as fault zones and contacts.
For various reasons, including the above-mentioned noise levels and In addition, it provides much better resolution of nearby magnetic
the high cost of helicopter-borne surveys, vertical aeromagnetic gradiom- sources. Gradient measurements also automatically remove the diurnal
etry was not accepted widely for hydrocarbon exploration. The justifica- noise and the regional gradients of the Earth’s magnetic field.
tion for such measurements remained valid, however. Figure 2 (Hood, 1979) shows that the vertical gradient data offer far
Over the past thirty years, many improvements have been made in better resolution of nearby magnetic sources that does the total field. This
aeromagnetic instrumentation and practice. Current optically pumped enhanced resolution is vitally important in determining the depth to these
total-field sensors (typically, cesium-vapor based) have resolutions of the sources; otherwise, such depth estimates may be excessive.
order of 1 pT (e.g., Scintrex CS-2). Using multiparameter, computer-soft-
ware-based compensation systems, figures of merit for all aircraft maneu-
vers of much less than 1 nT can be achieved. The term figure of merit is
a measure of the envelope of magnetic noise created by the aircraft when
undergoing a complete range of manovers.
Because of such high-resolution magnetometers and excellent aircraft
compensation systems, good-quality gradiometer data now may be
obtained, using dual-stinger installations, with vertical separations of only
2 m between sensors, as in Figure 1. In addition, the advent of precise
position determination by GPS has automated the computer data pro-
cessing and presentation completely, thereby greatly reducing survey
costs. As a result, the incremental cost of adding gradiometer measure-
ments to total-field aeromagnetic surveys is very small indeed.
With these advances in instrumentation, it is timely to revisit the
application of vertical magnetic gradiometry to hydrocarbon exploration.
Figure 1. Geophysical survey aircraft with 2-m tail-stinger-mounted vertical magnetic
gradiometer.
Figure 2. Comparative resolving powers of total-field and vertical gradient magnetic
measurements (after Hood, 1979).
06 State of the Art 4/16/07 4:57 PM Page 147
Figure 3. Vertical gradient and total-field profiles over a wide dipping dike; inclination
of the Earth’s field = 75° (after Hood, 1979).
Figure 3 (Hood, 1979) demonstrates how the vertical gradient defines
the contacts of a wide magnetic body, much better than is possible from
the total-field response of the body. In fact, it can be shown that the zero
gradient value will delineate the contact of major rock formations, pro-
viding that some measurable magnetization contrast exists across that
contact. It may be noted that it is possible to calculate the vertical gradi-
ent of the total magnetic field from the total-field data (McGrath et al.,
1976). The process for doing so, however, entails the gridding of the
observed total-field data, which suppresses much of the shorter wave-
length information. This short-wavelength information is, however, pre-
sent in the directly measured vertical gradient data.
Figure 4 (Hogg, 1989) presents a comparison among the contour
maps of the observed total magnetic field map of an area (A), the mea-
sured vertical magnetic gradient (B), and the calculated vertical magnetic
gradient (C). Hogg concluded that there is much fine detail in the mea-
sured gradient map that is not present in the calculated gradient. He also
noted that if a subsequent interpretation process should be based on the
vertical gradient parameter, the gradient should be measured rather than
calculated.
A further benefit of gradiometer measurements lies in their use for the
removal of cultural noise. Interference which is encountered increasingly
in conducting high-resolution aeromagnetic surveys in hydrocar-
bon exploration is the response of cultural sources. These may
include large metal buildings, pipelines, and oil-well casings, for
example. When the survey relies on detecting and interpreting
low-order (e.g., 1-nT) signatures due to suprabasement or base-
ment structures, then these cultural features may distort such sig-
natures seriously and create an erroneous interpretation.
In such circumstances, the use of measured gradiometer data
can serve to:
1) clearly resolve the near-surface cultural effects from the effects
of deeper geologic sources, and
2) provide the means for mathematically modeling the cultural
effects, thereby stripping them from the total-field data prior to
interpretation of the latter
Figure 5 shows the calculated total magnetic field anomalies at
a terrain clearance of 1500 ft (450 m) of one cased well, and of
three wells 2500 ft (750 m) apart. It can be seen that even with
this wide separation of sources, we are near the limit of resolution
of the total-field data at the survey elevation. The juxtaposition of
nearby cultural sources readily can simulate a long-wavelength
anomaly, which would be interpreted as a deeply buried geologic
source.
Figure 6 shows the vertical gradient of the magnetic field for
the same three wells. The spatial resolution of these wells is
enhanced greatly, to the point that each readily can be modeled
quantitatively and thereafter stripped from the original total-field
data. Depending on the nature of the cultural source, a simple
model usually will suffice for this purpose, such as a single pole for
a long vertical oil-well casing, or a dipole for a metal building.
Thus we can see that the addition of vertical gradiometer mea-
surements to total-field aeromagnetic surveys can increase the
quality and interpretability of these surveys significantly, at very lit-
tle additional cost.
References
Hogg, R. L. S., 1989, Recent advances in high sensitivity and high resolution
aeromagnetics, in Proceedings of Exploration ’87: Ontario Geological
Survey, Special Volume 3.
Hood, P. J., 1979, Magnetic methods applied to base metal exploration, in
Geophysics and geochemistry in the search for metallic ores, Geological
Survey of Canada Economic Geology Report 31, 7–104.
Figure 4. Comparison among contour plan maps of total field (A), measured vertical gradient (B),
and calculated vertical gradient (C) (after Hogg, 1989).
147
Geologic Applications of Gravity and Magnetics: Case Histories
06 State of the Art 4/16/07 4:57 PM Page 148

State of the Art

General Philosophy 7
edited by David A. Chapin
LaCoste & Romberg
Austin, Texas, USA
Where do we go from here?
Future interpretation technology could
take us in the following directions:
1) improving our understanding of the
physical processes behind the mag-
netic responses we measure
2) further integration with ancillary data
sets (e.g., radiometrics, electromag-
netics, satellite imagery, digital eleva-
Figure 5. Calculated total magnetic field anomalies of one and three wells, 2500 ft tion models)
(750 m) apart, at a terrain clearance of 1500 ft (450 m).
3) more sophisticated, robust, efficient,
user- and geology-friendly modeling
software
4) analysis, sensivity, and reporting of
noise sources and their effect on
interpretation
5) improving our understanding of the
meaning and limitations of the auto-
mated “interpretation” techniques
(e.g., Euler, anomaly correlation, ter-
racing)
148
In 1990, I put together my wish list of
future trends in gravity and magnetics for
an internal company report. It is a valu-
able lesson to revisit this list. Not only
are we heading in these directions, but it
shows us that we still have a long way to
go. Here is my list, in no particular
order:
1) more stress on basic bread-and-butter
interpretation work
2 integrated software tools in a work-
station environment
3) larger databases and associated conti-
nentwide evaluations
4) better inversion of gravity data
5) a better understanding of high-
frequency signals
6) more and better rock properties
databases
7) improved magnetic depth picking
8) borehole gravity as a more widely
used porosity/density tool
David Chapin

Stephen W. Reford
Paterson, Grant & Watson Ltd.
Toronto, Ontario, Canada

Figure 6. Calculated vertical magnetic gradient anomalies of the same cultural

sources.

McGrath P. H., Kornik, L. J., and Dods, S. D., 1976, A method for the compilation of
high quality calculated first derivative aeromagnetic maps, in Report of Activities,
Part C, Geological Surveys Canada, Paper 76-1C, 9–17.
Slack, H. A., Lynch, V. M., and Langan, Lee, 1967, The geomagnetic gradiometer:
Geophysics, 32, 877–892.’
07 Glossary 4/17/07 8:50 AM Page 149

Editors:
Mike Alexander
Corine Prieto
Contributors:
Richard Hansen
Patrick Millegan
Dave Oxley
Jeff Rowe
Stephen Reford
Integrated Geophysics Corporation
Integrated Geophysics Corporation

Pearson, deRidder and Johnson, Inc.

Marathon Oil Company
Mobil E&P Tech Center
Geoterrex
Paterson, Grant & Watson Ltd.

7
Glossary
07 Glossary 4/17/07 8:50 AM Page 150
The following terms are intended for readers not versed in potential-fields
terminology or for those readers interested in a review.
ACCUDRAPE™
A trademark of Geoterrex/CGG Canada, Ltd., it is a software and hard-
ware package that employs real-time navigation, allowing a gentle drape
with consistent traverse and tie-line terrain clearances at intersections.
ALTREX™
A trademark of Geoterrex/CGG Canada, Ltd., it is an airborne multi-
method tool for hydrocarbon ALTeration EXploration. The technique
maps anomalous surface conductivity and magnetic responses which may
be the result of hydrocarbons at depth.
Analytic signal method
The analytic signal method, known also as the total gradient method, as
defined here produces a particular type of calculated gravity or magnetic
anomaly enhancement map used for defining, in a map sense, the edges
(boundaries) of geologically anomalous density or magnetization distri-
butions. In exploration potential-field applications, the term analytic sig-
nal loosely refers to the calculated modulus of the gravity or magnetic
anomaly field’s three mutually orthogonal spatial (x, y, z) derivative terms.
Mapped maxima (ridges and peaks) in the calculated analytic signal of a
gravity or magnetic anomaly map locate the anomalous source body
edges and corners (e.g., basement fault block boundaries, basement lithol-
ogy contacts, fault/shear zones, igneous and salt diapirs, etc.).
Analytic signal maxima have the useful property that they occur
directly over faults and contacts, regardless of structural dip which may
be present, and independent of the direction of the induced and/or rema-
nent body magnetizations. Various extensions to the analytic signal
method (as defined here) exist. For example, some extensions to the
method include as an additional solved parameter the anomalous source
body depth(s).
Automated depth estimation
A variety of techniques, which include Werner deconvolution, the Euler
method, Naudy’s method, Phillips’ method, CompuDepth™, and the
analytic signal method, which analyze digital magnetic profiles or maps
to obtain estimates of source body depth without specific user identifica-
tion of key portions of anomalies. This contrasts with profile techniques
such as Peters’ method (half-slope) or Vacquier’s method (straight slope),
Glossary
which may be implemented as computer programs but require interactive
identification of special points on anomalies.
Bouguer gravity field
The gravity field obtained after latitude, elevation, Bouguer, and terrain
corrections have been applied to the measured (observed or raw) gravity
data. The Bouguer (named after Pierre Bouguer, a French geodesist) grav-
ity field often is noted as simple Bouguer for the gravity field before apply-
ing terrain corrections or complete Bouguer for the gravity field after
applying terrain (and sometimes curvature) corrections.
The gravity anomalies observed in the Bouguer field are caused by lat-
eral density contrasts within the sedimentary section, crust, and subcrust
of the earth.
A Bouguer gravity field measured above sea level and accurately cor-
rected to sea-level datum is not equivalent to gravity measured at sea
level. Anomalies caused by mass inhomogeneities between station eleva-
tion and datum and which were measured at the original station eleva-
tions remain in the data unless special corrections are made.
CompuDepth™
An automated depth-estimation method based on the exponential behav-
ior of magnetic dike or contact anomalies in the spatial frequency domain.
Correlation techniques for estimating the exponent are implemented by
back transforms into the spatial domain.
Compu-Drape™
A registered trademark of Paterson, Grant & Watson Limited (PGW), it is
PGW’s implementation of Cordell’s chessboard technique for continua-
tion of profile or gridded magnetic data between arbitrary surfaces (e.g.,
barometric-to-drape transformation, draping on the magnetic basement ).
Density
Mass per unit volume, expressed in grams per cubic centimeter. Rock or
formation densities usually are measured as either saturated bulk densi-
ties or grain densities. For gravity interpretation, the contrasts between
rock-bulk densities are of primary interest because they are responsible for
the anomalous gravity field.
Rock-bulk densities have been shown to vary as a function of geologic
age, lithology, and depth of burial. Rock densities typically range from
1.9 g/cm3 to 3.0 g/cm3.
150
Density contrast
The density of one rock unit relative to another. Density contrasts can be
either positive or negative. For example, if Rock A = 2.30 g/cm3 and
Rock B = 2.40 g/cm3, then the density contrast of Rock A relative to Rock
B is –0.10 g/cm3. Conversely, the relative density contrast of Rock B rel-
ative to Rock A is +0.10 g/cm3.
Gravity anomalies caused by density contrasts within the earth’s sed-
imentary section, crust, and subcrust can be analyzed and interpreted as
lithologic and/or structural anomalies.
Density-depth function
The relationship between the change in density with a change in depth.
In many areas of the world with thick clastic sections, the increase in den-
sity with an increase in depth has been shown to be primarily a function
of compaction. However, age, lithology, and porosity also may influence
the relationship. The relationship is important in gravity modeling
because a gravity anomaly may be caused by a gradational change in den-
sity rather than a relatively abrupt density contrast, such as that which
may occur at a fault, contact, or unconformity.
Density model
A model of the geology in which layers or bodies of given lithologies are
replaced by equidensity layers or bodies which may or may not corre-
spond to specific geologic formations.
Depth slicing
Generically, the use of linear filters to isolate (based on wavelength crite-
ria) anomaly contributions to a map derived from source bodies in a cer-
tain depth range. Numerous techniques are used to carry out the
isolation.
DIGHEM©
Copyrighted by Dighem/CGG Canada, Ltd., it is a multifrequency heli-
copter-borne electromagnetic system.
Dike model
See Prism. Dike-model descriptions include wide, narrow, thin, vertical,
and inclined.
07 Glossary 4/17/07 8:50 AM Page 151
Elevation correction
The sum of the free-air and Bouguer corrections to observed or “raw ”
gravity. The Bouguer correction requires an estimation of bulk density to
calculate and eliminate the gravitational effect of the subsurface mass
between point of gravity measurement and a datum.
Euler method
A profile-based or map-based depth-estimation method based on the con-
cept that the magnetic fields of localized structures are homogeneous
functions of the source coordinates and therefore satisfy Euler’s equation.
This equation therefore can be solved parametrically for the source loca-
tions. In recent years, use of this method has become more widespread
because it has been automated to work with either grid or profile data.
Forward modeling
Computation of the potential field anomaly expected from a geological
model.
Frequency domain
A domain is where a mathematical function (the independent and depen-
dent variables x, y, maybe z, and perhaps more) exists. In the frequency
domain, the independent variable has been transformed from a distance
such as miles (seconds in the case of seismic) to frequency such as
cycles/mile (a spatial frequency versus a temporal frequency such as
cycles/second). The dependent variables then become the strength and
phase of that frequency. See also Space domain.
Free-air gravity field
The gravity field after the free-air correction. This correction is applied to
observed or “raw ” gravity readings to correct for the change in gravity
caused by the difference in elevation of the gravity station relative to
datum elevation (usually sea level). The change in gravity with elevation
is related inversely to the change in distance between the meter’s center
of mass (meter elevation) and the Earth’s center of mass.
Gamma (g)
Simply, the units in which magnetic survey maps often are contoured; 1
gamma = 1 nanotesla. A unit for stating the magnitude of the magnetic
field vector B represented by the number of lines of induction passing
through a unit area perpendicular to the vector direction.
1 gamma = 10–5 gauss
= 10–5 lines/cm2
= 10–1 line/m2
= 10–9 weber/m2
= 10–9 tesla = 1 nanotesla
= 10–2 millioerstred
Gardner’s equation
An empirically derived equation which describes the relationship
between bulk densities and acoustic velocities of rocks: ρ = 0.23v0.25.
The equation was derived by G. H. F. Gardner et al. from laboratory
and field observations of brine-saturated (nonevaporite) rock types.
Experience has shown that the equation (or some modification of it) is
valid for many sedimentary basins of the world.
©
GEOTEM
Copyrighted by Dighem/CGG Canada, Ltd., it is a fixed-wing airborne
digital time-domain electromagnetic system.
Gravity unit
A unit of acceleration used with gravity measurements. Abbreviated as
g.u. Measurements in gravity units formerly were used widely, but mea-
surements in milligals are now more common; 1.0 milligal = 10 gravity
units.
Gradiometer
A device or set of devices which measures the value of a field in at least
two different points in space at the same time. The gradient is the differ-
ence in field values per unit of distance between the sensors. By measur-
ing a field’s gradient (that is, its first derivative or rate of change with
distance), the total field itself may be computed with varying degrees of
accuracy. For potential fields, the direction of the measurement relative to
the Earth is critical. Is the gradient being measured horizontally or verti-
cally, and in the case of magnetics, what is the orientation relative to the
Earth’s magnetic field? Even with these possible difficulties, measuring
just the gradient has the advantage of removing nongeologic field signals
such as, when measuring gravity, those introduced by the normal accel-
erations of the survey aircraft.
GRIDEPTH™
A registered trademark of Robertson Research International Limited (for-
merly Simon Petroleum Technology Limited, formerly . . .). It is the com-
Geologic Applications of Gravity and Magnetics: Case Histories
mercial implementation of 3-D Euler deconvolution, originally developed
by Reid et al. and implemented by PGW (UNIX) and Geosoft (PC).
High-density basement
The deepest significantly thick high-density unit(s) within the geologic
section of an area, which provide a major positive density contrast. The
rocks above the major density contrast are usually younger sediments
and/or volcanics, typically having densities ranging from approximately
1.9 g/cm3 to 2.6 g/cm3. Those below the major density contrast are usu-
ally older sedimentary, volcanic, and/or crystalline rocks, typically having
densities ranging from 2.6 g/cm3 to 3.0 g/cm3. High-density basement
may or may not be equivalent to crystalline and/or magnetic basement.
High-resolution aeromagnetics
This might be termed more correctly “high-precision aeromagnetics.” The
term has gained wide acceptance in the industry to denote surveys flown
at low terrain clearance (80-150 m), with close line spacings (100-500
m), recorded at high sample rates (0.1–0.25 s), and acquired with high-
sensitivity magnetometers (0.001–0.005 nT).
International Geomagnetic Reference Field (IGRF)
The most widely used mathematical models for fitting the main magnetic
field of the Earth at a given time (e.g., 1965). The models consist of spher-
ical harmonic coefficients derived from observatory and satellite data.
They are used to objectively remove long-wavelength components from
survey data to obtain the anomalous magnetic field (TI), which contains
the shorter wavelength components of exploration interest.
Inverse modeling
A technique whereby a 2-D or 3-D density, susceptibility, or geometric
(geologic) model is computed to satisfy (invert) a given observed gravity
or magnetic field.
Magnetic basement
Magnetic basement usually is equated to crystalline (felsic and mafic) or,
sometimes, metamorphic basement. It is the unconformity on which an
essentially nonmagnetic sedimentary section has been deposited. Large
exposures of basement (e.g., the Canadian Shield) show it to be litholog-
ically and magnetically heterogeneous. Very thick sequences of highly
magnetic volcanics sometimes may be considered equivalent to a mag-
netic basement.
151
07 Glossary 4/17/07 8:50 AM Page 152

Magnetic sedimentary section
A surface or zone within the geologic column where magnetic suscepti-
bility contrasts are significant enough to generate magnetic anomalies
which could delineate sedimentary geology. Susceptibility variations
within the sedimentary column generally are considered to be near zero
except where relatively magnetic sediments (e.g., pyroclastics, arkoses,
some shales) are present.
Milligal (mGal)
The unit of acceleration used with gravity measurements.
2 be posted on the following maps: Bouguer, free-air, regional, or residual
1 Gal = 1 cm/sec
1 Gal = 1000 mGal
1 mGal = 10 gravity units
N A trademark of Geoterrex/CGG Canada, Ltd., it is a suite of Petroleum
The degree of homogeneity in Euler’s equation, interpreted physically as
the falloff rate with distance and geophysically as a structural index (SI).
Values vary from 1 to 3, according to magnetic or gravity source body
geometry.
Nanotesla (nT)
Simply, the units in which magnetic survey maps often are contoured; 1
nanotesla = 1 gamma. A unit for stating the magnitude of the magnetic
field vector B represented by the number of lines of induction passing
through a unit area perpendicular to the vector direction.
1 nanotesla = 10–9 tesla
= 10–9 weber/m2
= 10–1 lines/m2
= 10–5 lines/cm2
= 10–5 gauss = 1 gamma
= 10–2 millioersted
Naudy’s method
An automated profile-based depth-estimation method wherein anomaly
type and location are identified by crosscorrelation of the observed mag-
netic profile with theoretical anomalies. The depth to a dikelike or plate-
like source then is estimated from parameters relating source body
half-width, depth, and data sampling interval.
Neural network
A member of a class of software that is “trained” by presenting it exam-
ples of input and the corresponding desired output. For example, the
Glossary
input might be a magnetic anomaly and the required output the depth to
the source of that anomaly. Training might be conducted using synthetic
data, iterating on the examples until satisfactory depth estimates are
obtained. Neural networks are general-purpose programs which have
applications outside potential fields, including almost any problem that
can be regarded as pattern recognition in some form.
Observed gravity field
The term observed gravity also is used often in lieu of raw gravity or mea-
sured gravity. Incorrectly, but often, the term observed gravity map may
gravity field.
PAM™
AeroMagnetic enhancement products that assists the interpretation of
magnetics.
Phillips’ method
An automatic depth-estimation method in which the source parameters
are estimated from the autocorrelation function of the magnetic anomaly.
Like Werner deconvolution, the method uses a dike or contact model.
Plate
A term used to describe a sheetlike magnetic source body with limited
vertical dimension. That is, its thickness may range from 0.1 to 1.0 times
its depth to top. Its anomaly character is similar to that of a set of dipoles.
Potential field
A field which obeys a differential equation known as Laplace’s equation.
Gravity and magnetic fields are both vector potential fields. Most explo-
ration gravity work uses the vertical component of the gravity field, and
most exploration magnetic work uses the scalar total intensity of the mag-
netic field.
Pseudodepth slice
A pseudodepth slice can be described as a frequency filter without hard
edges, progressively reducing signal amplitude away from a desired fre-
quency band. The direct mathematical relationship between frequency
and depth in magnetics analysis makes this a technique for sharpening a
152
desired frequency-depth range while retaining a subdued expression of
frequencies outside the range.
Pseudogravity
An approximation of a gravity field derived from a magnetic field mea-
sured at, or transformed to, the magnetic pole. The process requires con-
version of susceptibility values to density values and a vertical integration
of the magnetic data.
Prism
A term used to describe a magnetic source body which can be considered,
for practical purposes, parallelepiped, which is semiinfinite in vertical
dimension. That is, its depth to bottom is at least four times its depth to
top. Its anomaly character is similar to that of a monopole or line of poles.
A 2-D prism (semiinfinite normal to the plane of section) sometimes is
referred to as a dike model.
Raw gravity
Also called measured gravity or observed gravity. The gravity field mea-
sured at a gravity station before latitude, free-air, Bouguer, or terrain cor-
rections are applied.
Reduction-to-equator (RTE)
A mathematical transformation of the total-magnetic-intensity (TI) field at
its observed inclination (I) and declination (D) to that of the magnetic
equator (i.e., I = 0°).
Reduction-to-pole (RTP)
A mathematical transformation of the total-magnetic-intensity (TI) field at
its observed inclination (I) and declination (D) to that of the north mag-
netic pole (i.e., I = 90°, D = 0°).
Regional gravity field
The long-wavelength component of the Bouguer gravity field usually
attributed to density variations considered to be deeper than general
exploration interest (e.g., the gravity component due to crustal density
variations or undulations of the crust/mantle interface). A subjective
regional often can be designed to enhance residual anomalies of primary
interest.
07 Glossary 4/17/07 8:50 AM Page 153

Residual gravity field Strike filter (pass or reject) Werner deconvolution

The shorter wavelength component of the Bouguer gravity field attributed
to density contrasts within high-density basement and/or the lower den-
sity overburden. Anomalies in the residual field are usually of exploration
interest.
A first residual is a difference field obtained by subtracting the
regional gravity field from the Bouguer gravity field.
A band-pass filter designed to pass or attenuate components of a poten-
tial-field data set along a predetermined angle (strike).
Structural model (2-D)
A gravity or magnetic structural model is a 2-D density and/or suscepti-
bility model of given or assumed geology. The geology of an area can be
An automated profile-based depth-estimation method derived from
S. Werner’s analysis of magnetic anomalies from sheetlike bodies.
Polynomials representing a total-field anomaly or its derivative (horizon-
tal gradient) can be solved simultaneously to estimate the depth, dip, hor-
izontal location, and susceptibility of the source body (thin sheet or
interface).

SAMS™ modeled by representing lithologic layers as equidensity and/or equisus-

A trademark of World Geoscience Corp., it signifies Sedimentary
AeroMagnetic Surveys which are characterized by close line spacing and
low-altitude (high-resolution) airborne acquisition.
Second-vertical derivative
A second-vertical-derivative map of a potential field may be calculated by
application of a frequency-domain or space-domain filter to a potential-
field grid file. The result is an anomaly enhancement or residual map
related to the “curvature” of the input field. Inflection points on the
anomalies of the input field will be zero values on the derivative map and
may have special interpretation significance.
Space domain
A domain is where a mathematical function (the independent and depen-
dent variables x, y, maybe z, and perhaps more) exists. In the space
domain, distance (1 if by profile, 2 if by map measured in perhaps feet,
kilometers, degrees, seconds, etc.) is the independent variable and some
quantity (milligals, gammas, density, seismic amplitude, etc.) is the depen-
dent variable. See also Frequency domain.
SPI™ (Source Parameter Imaging)
A trademark of Geoterrex/CGG Canada, Ltd., it is a method for deter-
mining depth, susceptibility, and dip of magnetic sources. The method
assumes contact or thin-sheet models and uses the complex analytic signal.
STARMAG™
A trademark of Texaco, it is a method for estimating residual basement
depth using a neural network with smoothed estimated basement depths
and magnetic field anomalies as input.
ceptibility layers and/or blocks. The layers are formed by contrast bound-
aries which may or may not correspond to specific geologic formation
boundaries. Where high density or susceptibility contrasts exist in nature,
the model may correspond closely to those geologic formations.
For 2-D modeling, the density and susceptibility models of the geol-
ogy and the observed gravity and magnetic anomalies for the model are
assumed to be semiinfinite. For 2-D modeling, the third dimension y (in
and out of the plane of the profile) is approximated by one or more given
distances, thus providing a quasi-3-D model.
Susceptibility
A measure of the degree to which a substance may be magnetized. It is a
ratio k of the intensity of magnetization I to the causative magnetic field
H. It typically is expressed in micro cgs units for oil and gas exploration
work. Susceptibility has been shown to be proportional to the volume per-
centage of magnetite contained in a rock. Susceptibility contrast is the
susceptibility difference between two rocks or geologic bodies. See
Density contrast.
Three-dimensional (3-D) model
A network or grid of values which models a geologic surface represented
as a surface of density contrast (gravity) or susceptibility contrast (mag-
netics). The output of a forward model is based on the calculated gravity
or magnetic effect of a specified input surface. The output of an inverse
model is the geometry of an appropriate (but nonunique) surface calcu-
lated by inverting the input gravity or magnetic field.
Total magnetic intensity anomaly
The total-magnetic-intensity anomaly field (TI) is the resultant field after
correcting TF, the total magnetic (observed) field for a regional gradient
field, such as an IGRF.

153
Geologic Applications of Gravity and Magnetics: Case Histories
07 Glossary 4/17/07 8:50 AM Page 154

This page has been intentionally left blank

08 Bibliography 4/17/07 8:56 AM Page 155

8 Annotated
Bibliography
08 Bibliography 4/17/07 8:56 AM Page 156
This list was compiled to provide references to both “classic” historical
papers and books on gravity and magnetics interpretation and similar
resources of instructional value. It was compiled by Dick Gibson, but the
core of the list (including many of the annotations, some of which are
quoted) comes from extensive bibliographies prepared by David Chapin
and Alan Reid. Many other people contributed references, and some of
their comments are quoted as well. Although the focus is on interpreta-
tion, a few references to other aspects of gravity and magnetics are
included. The list contains references that range from theoretical treat-
ments to case histories, and we hope that it will be a useful pointer for
explorationists. It is not an exhaustive list, but it is intended as a starting
point and guide.
Abdoh, A., Cowan, D., and Pilkington, M., 1990, 3-D gravity inversion
of the Cheshire Basin: Geophys Prosp., 38, 999–1011.
A clear example of inversion applied to real geology.
Atchuta Rao, D., and Ram Babau, H. V., 1984, On the half–slope and
straight-slope methods of basement depth determination: Geophysics,
49, 1365–1368.
Critical and useful examination of some classic methods.
Barnett, C. T., 1976, Theoretical modeling of the magnetic and gravita-
tional fields of an arbitrarily shaped three-dimensional body:
Geophysics, 41, 1353–1364.
A breakthrough in flexibility and computing speed.
Beyer, L. A. , 1991, Borehole gravity surveys, Short Course notes, Soc.
Expl. Geophys.
An excellent source for general principles of borehole gravity. Has very
good figures and references.
Bhattacharyya, B. K., 1965, Two-dimensional harmonic analysis as a tool
for magnetic interpretation: Geophysics, 30, 829–857.
A classic; provides the mathematical basis for many of the tools of the
trade.
Bhattacharyya, B. K., 1978, Computer modeling in gravity and magnetic
interpretation: Geophysics, 43, 912–929.
A review of the theory of various modeling schemes.
Annotated Bibliography
Blakely, R. J., 1995, Potential theory in gravity and magnetic applications:
Cambridge Univ. Press.
Simply the best thing there is on the mathematics you need to get seri-
ous about potential fields as applied to exploration geophysics. —Alan
Reid
Blakely, R. J., and Connard, G. G., 1989, Crustal studies using magnetic
data, in Pakiser, L. C., and Mooney, W. D., Eds., Geophysical frame-
work of the continental United States: Geol. Soc. Am. Memoir 172,
45–60.
Valuable information for tectonic studies.
Bostrom, R. C., 1989, Subsurface exploration via satellite: Structure visi-
ble in Seasat images of North Sea, Atlantic continental margin and
Australia: AAPG Bull., 73, 1053–1064.
Well-written early announcement of the expressions of features of explo-
ration interest that are detectable from satellite.
Bowin, C., Wollenhaupt, W. R., and Simon, B., 1975, Mascons: A two-
body solution. J. Geophys. Res., 80, 4947–4955.
An explanation of gravity anomalies on the moon.
Breiner, S., 1973, Applications manual for portable magnetometers:
Geometrics.
A highly readable, although somewhat dated, guide for the acquisition
and interpretation of magnetic data. Includes good illustrations of basic
concepts.
Chenot, D., and Debeglia, N., 1990, Three-dimensional gravity or mag-
netic constrained depth inversion with lateral and vertical variation of
contrast: Geophysics, 55, 327–335.
A complete treatment of depth inversion.
Cordell, L. E., and Grauch, V. J. S., 1985, Mapping basement magnetiza-
tion zones from aeromagnetic data in the San Juan Basin, New
Mexico, in Hinze, W. J., Ed., The utility of regional gravity and mag-
netic anomaly maps: Soc. Expl. Geophys., 181–197.
Simple but effective. —Alan Reid
Eggers, D. E., and Thompson, D. T., 1984, An evaluation of the marine
magnetic gradiometer: Geophysics, 49, 771–779.
A method of collecting diurnal-free marine magnetic data.
156
Ervin, C. P., 1977, Short note: Theory of the Bouguer anomaly:
Geophysics, 42, 1468.
A short and important paper reiterating basic concepts.
Evjen, H. M., 1936, The place of the vertical gradient in gravitational
interpretations: Geophysics, 1, 127–136.
An early paper on gradiometry, which has become one of the “hot top-
ics” in gravity exploration in the 1990s.
Gay, S. P. Jr., and Hawley, B. W., 1991, Syngenetic magnetic anomaly
sources: Three examples: Geophysics, 56, 902–913.
Case studies demonstrating the existence and possible genesis for shal-
low, intrasedimentary magnetic sources.
Gay, S. P., 1961, Standard curves for interpretation of magnetic anomalies
over long tabular bodies: Geophysics, 28, 161–200.
A classsic aid to interpretation.
Grant, F. S., 1985, Aeromagnetics, geology and ore environments, I.
Magnetite in igneous, sedimentary and metamorphic rocks: An
overview: Geoexploration, 23, 303–333.
This is one of too few papers that help us try to understand what min-
eralogy lies behind the anomalies we interpret. —Stephen Reford
Gunn, P. J., 1975, Linear transformations of gravity and magnetic fields:
Geophys. Prosp., 23, 300–312.
A clear and complete exposition on Fourier domain operations, which
dominate many aspects of processing and interpreting gravity and mag-
netic data.
Gunn, P. J., Ed., 1997, Airborne magnetic and radiometric surveys: AGSO
J., 17, no. 2.
A valuable, wide-ranging set of papers covering all aspects of magnetic
work. Australian focus with examples from other parts of the world.
Hammer, S., 1983, Airborne gravity is here!: Geophysics, 48, 213–223.
Stirred up a considerable controversy, but announced a significant
advance.
08 Bibliography 4/17/07 8:56 AM Page 157
Hanna, William F., Ed., 1987, Geologic applications of modern aeromag-
netic surveys: USGS Bull. 1924.
Proceedings of a workshop with useful and instructional examples, espe-
cially for regional tectonics.
Hartman, R. R., Teskey, D. J., and Friedberg, J. L., 1971, A system for
rapid digital aeromagnetic interpretation: Geophysics, 36, 891–918.
The best accessible paper on Werner deconvolution.
Heiskanen, W. A., and Meinesz, F. A. V., 1958, The Earth and its gravity
field: McGraw Hill Book Co.
A classic. Contains theory and techniques for dealing with large-scale
gravity data sets.
Hinze, W. J., Ed., 1985, The utility of regional gravity and magnetic
anomaly maps: Soc. Expl. Geophys.
Thirty-three papers describing continent- and basin-scale data sets and
their analysis. Worldwide application. Something of an inspiration for
the present case-histories volume. —Dick Gibson
Hospers, J., Finnstrom, E. G., and Rathore, J. S., 1985, A regional gravity
study of the northern North Sea (56-62°N): Geophys. Prosp., 33,
543–566.
The effects of isostasy on gravity as applied to petroleum exploration. A
valuable exploration-oriented paper. Vital. —Alan Reid
Klitgord, K. D., and Popenoe, P., 1984, Florida: A Jurassic transform plate
boundary: J. Geophys. Res., 89, 7753–7772.
A modern, integrated regional study which shows the usefulness of grav-
ity and magnetics to interpret plate boundaries and relative motions.
LaFehr, T. R., 1983, Rock density from borehole gravity surveys:
Geophysics, 48, 341–356.
One of the first descriptions of borehole gravity technology and its appli-
cation to determination of parameters useful in exploration.
Leaman, D. E., 1994, Criteria for evaluation of potential field interpreta-
tions: First Break, 12, 181–191.
This paper sets out procedures for extracting the maximum possible
quantitative geologic information from gravity and magnetic data, using
rigorous methods analogous to those used in interpretation of seismic
profiles. Recommended!
Lines, L. R., Tan, H., and Treitel, S., 1991, Velocity and density imaging
between boreholes: CSEG Recorder, 16, 9–14.
A unique case study which integrates borehole gravity with the process-
ing and interpretation of cross-well tomography.
Millegan, P. S., 1990, Aspects of the interpretation of Mesozoic rift basins
in northern Sudan using potential fields data: Soc. Expl. Geophys.
Expanded Abstracts, 1, 605–607.
Useful exploration case history in a relatively frontier area.
Nabighian, M. N., 1972, The analytic signal of two-dimensional magnetic
bodies with polygonal cross-section: Its properties and use for auto-
mated anomaly interpretations: Geophysics, 37, 507–517.
This report describes one of the first (then) new approaches to magnetic
depth estimation in some years.
Naudy, H., 1971, Automatic determination of depth on aeromagnetic pro-
files: Geophysics, 36, 717–722.
Digital technique for magnetic depth estimates.
Nettleton, L. L., 1976, Gravity and magnetics in oil prospecting:
McGraw-Hill Book Co.
The classic textbook, out of print and somewhat dated but still highly
readable and informative.
Nettleton, L. L., 1971, Elementary gravity and magnetics for geologists
and seismologists: Soc. Expl. Geophys. Monograph Series, 1.
A simple little handbook that provides a great introduction to classic
(and standard) approaches to gravity and magnetic surveying and
interpretation.
Okubo, Y., Graf, R. J., Hansen, R. O., Qgawa, K., and Tsu, H., 1985,
Curie point depths of the island of Kyushu and surrounding areas,
Japan: Geophysics, 53, 481–494.
A case history which shows how to make Curie-point depth estimates
and how to interpret the results.
Patterson, N. R., and Reeves, C. V., 1985, Applications of gravity and
magnetic surveys: The state-of-the-art in 1985: Geophysics, 50,
2558–2594.
An overview, with some regional case studies, of gravity and magnetic
mapping and interpretation.
Geologic Applications of Gravity and Magnetics: Case Histories
Pawlowski, R. S., 1992, Tutorial: Gravity anomalies for nonspecialists:
The Leading Edge, September, 41–43.
A nice review of what gravity anomalies really are, based on models.
Pearson, W. C., Wiener, J. M., and Moll, R. F., 1990, Aeromagnetic struc-
tural interpretation using neural networks: A case study from the
northern Denver-Julesberg Basin: 60th Ann. Internat. Mtg., Soc. Expl.
Geophys., Expanded Abstracts with Biography, 587–590.
Artificial intellegence techniques applied to the interpretation of mag-
netic data.
Peters, L. J, 1949, The direct approach to magnetic interpretation and its
practical application: Geophysics, 14, 290-–20.
A respected method, widely used in the past and still appropriate in
some cases of low-resolution data or simple, quick, ballpark magnetic
depth estimation.
Prieto, C., Perkins, C., and Berkman, E., 1985, Columbia River basalt
plateau—an integrated approach to interpretation of basalt-covered
areas: Geophysics, 50, 2709–2719.
Integration of seismic reflection, well data, magnetotellurics, and gravity
and magnetics data to analyze subvolcanic plays.
Reford, M. S., 1964, Magnetic anomalies over thin sheets: Geophysics,
29, 532–536.
A catalog of models and their computed results.
Reid, A. B., 1980, Aeromagnetic survey design: Geophysics, 45,
973–976.
Planning surveys can be as important as acquiring the data. This is
almost the only treatment of this topic that has been published.
Reid, A. B., Allsop, J. M., Granser, H., Millett, A. J., and Somerton,
I. W., 1990, Magnetic interpretation in three dimensions using Euler
deconvolution: Geophysics, 55, 80–91.
3-D digital depth to magnetic source using the Euler method. This
method is rapidly gaining favor for its ability to quickly analyze a
large amount of data. —David Chapin
Shuey, R. T., and Pasquale, A. S., 1973, End corrections in magnetic pro-
file interpretation: Geophysics, 38, 507–512.
Development of the 2.5-D geometry for modeling magnetics. This is the
method used in most current modeling programs.
157
08 Bibliography 4/17/07 8:56 AM Page 158

Simpson, R. W., and Jachens, R. C., Gravity methods in regional studies,
in Pakiser, L. C., and Mooney, W. D., Eds., Geophysical framework of
the continental United States: Geol. Soc. Am. Memoir 172, 35–44.
Important for basin analysis and tectonics.
Simpson, R. W., Jachens, R. C., Blakely, R. J., and Saltus, R. W., 1986,
A new isostatic residual gravity map of the conterminous United
States with a discussion on the significance of isostatic residual anom-
alies: J. Geophys. Res., 91, 8348–8372.
Discusses the importance of isostatic gravity maps in the context of tec-
tonic analysis.
Skeels, D. C., 1967, What is residual gravity?: Geophysics, 32, 872–876.
An instructional discussion of regional/residual separation.
Soc. Expl. Geophys., 1948, 1956, Geophysical Case Histories
(2 volumes).
Another inspiration for the present volume. These case histories contain
classic exploration examples of salt domes, reefs, and basement tecton-
ics (before the term existed). U.S. focus, but some examples from basins
around the world.
Annotated Bibliography
158
Thompson, D. T., 1982, EULDPH: A new technique for making com-
puter-assisted depth estimates from magnetic data: Geophysics, 47,
31–37.
The original Euler deconvolution paper, describing an automated depth
estimator.
Vacquier, V., Steenland, N. C., Henderson, R. G., and Zeitz, I., 1951,
Interpretation of aeromagnetic maps: Geol. Soc. Amer. Memoir 47.
A classic, widely quoted and, for many years, widely used, and still used
by lots of people.
Wynn, J. C., Ed., 1986, Special issue on geophysics in archeology: Soc.
Expl. Geophys., 51.
Several papers describe magnetic surveys to define features such as
tepee rings and hearths.

Spector, A., and Grant, F. S., 1970, Statistical models for interpreting aero-
magnetic data: Geophysics, 45, 293–302.
A classic that has had enormous influence on the subject.

Steenland, N. C., 1965, Oil fields and aeromagnetic anomalies:

Geophysics, 30, 706–739.
A classic of direct application to hydrocarbon exploration.

Steenland, N. C., and Brod, R. J., 1960, Basement mapping with aero-
magnetic data—Blind River Basin: Geophysics, 25, 586–601.
A high-quality interpretation of aeromagnetic data using good old-
fashioned common sense.

Talwani, M., 1965, Computation with the help of a digital computer of

magnetic anomalies caused by bodies of arbitrary shape: Geophysics,
30, 797–817.
A classic that was the basis for many modeling techniques.
09 Index 4/17/07 9:00 AM Page 159

9
Index
09 Index 4/17/07 9:00 AM Page 160

Index
160
A Chaco-Parana Basin, 74 Eucutta Field, 69–70, 71
chemical remanent magnetization, 133 Euler deconvolution, 85, 98, 148
acid mine drainage, 112–113
Chocolate Mountains, 105–110 exploration, diamond, 104
algorithms, Talwani, 20
Cicuco Field, 29 exploration process, 6
alteration plume, 125, 126, 127
Cleveland Oil Field, 3 exploration risk, 6
Altiplano, 35
Conroe Dome, 3 exploration strategy, 7–8
ambiguity, 102
continental-scale data, 32–34
amplitude variation with offset (AVO), 9, 74
Continental Offshore Stratigraphic Test (COST) #1 well, 47
analytic signal, 92
continuation, 41, 92
F
Andes, 34, 35 Fairway Field, 127
Copper River Basin, 37
anhydrite, 22 faults, 20, 72, 81
crustal anomalies, 72
anomalies, 93 filtering, 10–12, 96
cultural editing, 95
anticline, 22 folds, 21
cultural effects, 92, 130, 146
Appalachian Valley, 120 Fourier analysis, 79
cultural features, 96
Arbuckle Formation, 75 Fourier transform, 11
cultural noise, 93, 147
archaeology, 104 frequency analysis, 10–12
Arctic National Wildlife Refuge (ANWR), 18–19, 130–133 frequency-depth relationship, 63
Arkoma Basin, 75 D frequency filter, 88, 92
Arvaca High, 29 Decade of North American Geology (DNAG), 32–33, 42, 54 future technology, 148
authigenic minerals, 18 decorrugation, 37
Aves Ridge, 53 Defense Mapping Agency (now National Imagery and
Mapping Agency [NIMA]), 33
G
gamma ray-density logs, 116
B density logs, 116
gamma ray-neutron logs, 116–119
Denver Basin, 124, 127
Bacon Flat Field, 126 geochemistry, 124
Denver Julesberg Basin, 127
band-pass filter, 97 gradiometer, 135
diagenetic minerals , 90
basalt, 24 gradiometry, 136–137, 145–146
diamond exploration, 104
basement, 26, 40, 72, 88 Grant Canyon Field, 126
Diamond Field, 69–70
basement lithology, 41 gravimeter, development of, 3
dike sill, 25
bathymetry, 33 Grenada Basin, 53
dipping beds, 21
bathymetry effects, 33 gridding, 32, 37, 96
Bayes’s theorem, 13 Gulf Coast, 9–10, 14, 17, 82, 137
borehole (core-hole) gravimeter, 113, 117–119 E Gulf of Mexico, 14, 46, 72, 74, 137
borehole gravity, 46, 148 Eagle Springs Field, 125 Gunsan Basin, 63
Brookian sequence, 131 East China Sea, 59
Brooks Range, 131 edge detection, 138
Bureau Gravimetrique Internationale (BGI), 33 edge maps, 96
H
El Dificil Field, 29 Hartzog Draw Field, 30
hematite, 90
C electromagnetics (EM), 105, 113, 120, 122, 124
Heuksan Basin, 63
Ellesmerian sequence, 131
Canadian Shield, 41, 42–43 high-resolution aeromagnetics (HRAM), 18-19, 93–94, 130
environmental, 104
cap rock, 9, 84 historical perspective, 3
Eötvös, 135
carbonates, 94 Holitna Basin, 37
error budget, 13
Cement Field, 17–18, 89, 131
09 Index 4/17/07 9:00 AM Page 161
horizontal gradient, 96, 126
House Creek Field, 30
hypothesis testing, 93
I Mesquite gold mine, 105
ideal body analysis, 93, 102
igneous, 24, 74
illite, 89
induced polarization (IP), 105
intrusives, 24
Irish Sea, 85, 86
isostatic correction, 13
isostatic residual, 35, 92
isotopes, 121
K Monte Carlo simulation, 102
Kate Spring Field, 125
kimberlite pipes, 104
Koenigsberger ratio, 131
Kuskokwim Basin, 39
L National Geophysical Data Center, 34, 37
large-scale data, 32
Lesser Antilles, 53
leveling, 37, 79
limestone, 24
line spacing, 97
lithologs, 116, 118
Little Nancy Field, 127
Llano Basin, 29
Lower Magdalena Basin, 29
M Okinawa Trough, 59, 61
Macarenas High, 29
mafic, 24
maghemite, 90
magnetic depth estimation, 93, 98
magnetic gradiometer, 120
magnetic gradiometry, 146
magnetic minerals, 131
magnetite, 90 Permian Basin, 113–114
magnetometers, 3,17 pipelines, 122
magnetotactic bacteria, 18 Pine Prairie Dome, 82
Marsh Creek anticline, 133, 134 pitfalls, 32, 33, 53–54
Mercia Mudstone, 87 planning, 7–8
Pollen Field, 127
micromagnetics, 17, 125 potassium, 121
Mid-Continent Rift, 80 Powder River Basin, 29
Middle Tanana Basin, 37 Project Magnet, 63
Minchumina Basin, 37 profile residual, 10–11
mineralization, 105 prospect definition, 74
mineralogy, 131 pseudodepth slice, 88, 92
mining, 104 pseudogravity, 92
Mississippi Salt Basin, 68, 69, 70 pseudomagnetics, 92
mis-tie, 79 pyrite, 90
modeling, 84, 85 pyrrhotite, 90, 99
Morecambe Fields, 85, 86–87
mylonites, 107
Q
quality control, 95
Quill Field, 127
N
Nash Dome, 3, 84, 136
R
radiometrics, 124
National Imagery and Mapping Agency (NIMA), 33
nonuniqueness, 101, 102 radiometric spectrometry, 120
radionuclides, 121
Norphlet, 71
North African Craton, 30 Railroad Valley, 124–125
reconnaissance, 6
North Sea, 74
nuclear waste, 120 Reddell Oil Field, 82
reduction-to-the-equator, 53
reduction-to-the-pole, 38, 51–52, 92, 96
O reef, 24, 74, 95
Oak Ridge, 120 regional, 36
Ochoa series, 114 remanent magnetization, 131
reservoir management, 7–8
ore bodies, 106 residual, 36, 84, 92, 108, 142
Orocopia Schist, 106 residualization, 9, 13
Ouachita overthrust belt, 75 resistivity, 112–113, 121
resolution, 7–8, 79, 102, 138, 146
resource evaluation, 7–8
P Richton Dome, 10
Paradox Basin, 124, 126
ring residual, 11
Patchuta Creek Field, 69 risk, 74
161
Geologic Applications of Gravity and Magnetics: Case Histories
09 Index 4/17/07 9:00 AM Page 162

Index
162
risk reduction, 6 T
rock density, 85, 116
Taiwan Basin, 59
Rubiales High, 29
Telemzane High, 30
Ryukyu Arc, 59–60
tensor, 138
terraced map, 38
S three-dimensional modeling, 74
Sadlerochit Mountains, 13 Tokai Basin, 59
salt, 9–10, 14, 68, 71, 82, 84, 137, 142 torsion balance, 3, 135–137
salt dome, 11, 15 Trap Spring Field, 125
salt pillow, 22 thrust faults, 75, 107
salt reef, 22 Tunghai Shelf, 59, 61
Scripps Observatory, 34
Sea of Japan Basin, 54 U
Senkaku Basin, 59
Urengoy Field, 81
seepage, 131
shale diapir, 26
Sherwood Sandstone, 87 V
Shublik Mountains, 132 velocity modeling, 74
Sierra Reef, 120 vertical derivative, 96
Spindletop, 135 vertical gradient, 92, 125–126, 146
St. George Basin, 47 vertical gradient magnetics, 120, 122
strategy, 7–8 Viola Formation, 76
structural interpretation, 92 volcanics, 22, 74, 130
Smackover, 71
South Hartshorne Field, 75
South Yellow Sea, 63 W
stratigraphy, 131 wavelength filtering, 10–12, 108
sulfur, 132 Werner deconvolution, 98
sulfur generation, 115 West Siberian Basin, 79
suprabasement effects, 26 Wiggins Arch, 72
survey design, 95
survey planning, 7–8 Y
survey spacing, 104
Yoyo Reef, 93
survey, regional, 104
Yukon-Koyukuk-Kobuk Basins, 39
susceptibility, 14, 40–41, 44, 60, 81, 88, 131
Susitna Basin, 37
syncline, 22