Geoscience Canada Volume 17 Number 2
Uses (and Abuses) of
Ore Deposit Models
in Mineral Exploration
C.J. Hodgson
Dopertment of Geologicel Sclonces
‘Queen’ University
‘Kingston, Ontario K7L 3N6
Eaditors Note: The folowing attic orignally
appeared in Exploration '87 Proceedings:
The Role of Exploration in Resource De-
velopment. t's reprinted here withthe per-
mission of the Oector, Ontario Geological
‘Survey.
Abstract
[An ore deposit model's a conceptual andior
‘empirical standard, embodying both the de-
scriptve features ofthe deposit type, and an
explanation of these features in terms of
geological processes. The descriptive fear
tures of models serve as criteria for explora
Uonareaselection (ares selectioncritera’
How they are used in this function depends
onthe scale of ther spatial association with
0F@, on our confidence that they are reliable
Indicators of ore, and on the extent to which
they are preferentially associated with eco-
rnomically better deposits. The geological,
geochemical, and geophysical techniques
Usedin explotation. and exploration strategy
depend on area selection ertera. The rea-
tive importance ofarea selection eritriacan
be dotermined from their relative Requency
of aesociation with ore in a representative
‘samplecthe deposit population, resuiting in
an empiical model. A genetic model is de-
rived by considering the geneticrelationship
of area selection criteria to ore. The weak
Inks in model bullding are the lack of effort
which goes into systematically assembling
the data on the known population of depos-
its, and the weak scinific underpinnings of
the geneticirterpretation. Both ofthese fac-
tors influence exploration by leading toinap-
propriate assessments ofthe relative impor-
tance of area selection criteria In addition,
there are a number of human foibles which
ccommonty lead to shortcomings In the de-
velopment and use of models. The mast sig
nificant of these is our tendency to rely 0
much on too simple models. We do this to
{avoid the discomfort of uncertainty and con-
{fusion which inevitably comes when we are
called on to assess exploration sitions.
The history of exploration for messive
base-metal sulphide deposits and gold e-
posits inthe Canadian Shield provides @ good
iMystration oftpeinniuence of models on area.
selection criteria, and thereby, on explora
‘ton strategy and techniques.
Introduction
‘A madel in geology Is @ conceptual andior
‘empirical standard which embodies the es-
sential features of some population of natu
ralgeciogical phenomena. Athough a model
‘can be strictly descriptive, most contain in-
terpretive elements that explain the relation-
ships among the vorious descriptive fea-
tures in terms of geological processes. Mod-
‘ets of ore deposits are widely used in mineral
‘exploration as a basis for predicting the ex
ploraton potential of areas which may range
in scale rom largo regions down toindlvidual
‘ore zones. Everyone uses models, but often
Ite thoughtis given to how models are bit,
how they influonce exploration programs,
and how they ean be improved,
The objective ofthis paper Is to examine
the structure ofore deposit models and their
roleinthe mineral exploration process. nis,
amous book, The Structure of Scent Rov.
‘lutions, Kunn (1962) argued that macels are
‘used, consciously oF not, to predict in vi-
‘ualy allot our everycay interactions with our
‘environment, cluding in scientiicresearch,
this view is accepted, then models should
represent as close an approximation as pos-
sible to realty if they are lo properly guide
‘us. However, t should be emphasized that
‘models are a two-edged sword. On the one
hand, they are @ powerful means of organiz-
Ing datain a form that enhances understand-
Ing and prediction. Buton the other hand, by
‘operating to exclude perception of di
whieh does not fit the model, they have 9
‘oporitic effect, that may lead to unjustified
‘confidence in the application of the model
LARGE AREA
GEOLOGICAL CRITERIA FOR
EXPLORATION AREA SELECTION
EXPLORATION TECHNIQUES
‘SMALL AREA)
Figure 1 The exploration process: exploration
techniques (ge0'0g¢a. geophysical. and geo
‘chemical are used fo measure the detnbution of
‘rea zaiectioncmerissothalthe most prospect
Darts of laze exploration aoe can be selected
Tortutber exploration
cry
‘These concepts are ilustrated by describing
the infuence on exploration strategy and
techniques of historical changes in models
for base-metal, volcanogenic massive su
hide, and for gtd deposits.
Mineral Exploration
Mineral exploration Involves the progressive
reduction in the size of the area being ex-
‘lored until @ mine is found (Figure 1). The
slarting polnt may be a region the size of 8
continent, or simply one level in @ rine.
‘Whatever its size, the objective of explora
tions to focus attention progressively on the
‘most favourable ports of the area so that as
exploration proceeds, the chances of an
‘economic mineral deposit being found con-
tinwously increase. Area reduction normally
takes place in steps which ore seperated by
‘2r0a selection “dacision points (Figure 2)
For example, in 8 regional reconnaissance
‘program, the frst decision point might follow
the completion of large-scale geochemical
‘and geophysical surveys, at which time a
‘number of claim groups might be staked or
optioned.
‘Area selection in mineral exploration is
based on the presence or absence of spec.
le geological features, or alternatively. geo
Physical and geochemical features which
Foflect geological features. These features
(Gee Figure 1). Area selection criteria are
concrete, measurable features, not con
‘cepts, For example, it might be considered
thal volcanic centres are an important re-
‘glonal-scale area selection criterion for vol
‘canogenic massive sulphide deposits. How
‘ever, “volcanic centre" is nota feature which
shows on the legend of most geological
‘mape. Rather, tis an interpretation based on
the distribution and configuration of specific
Ithologies and structures in on area. The
‘map patterns which indicate “voleanic cen-
{te are the concrete realy, and itis these
‘map pattorns that comprise the area selec-
tion enteria,
‘Mucho! the werk of exploration consists of
tiring the distribution of area selection
criteria in the exploration area. The term
“exploration strategy” can be used toreterto
the sequence of activites which resuits in
the progressive reduction inthe size of the
‘exploration area, The best exploration strat
egy optimizes he balance between cost and
effectiveness inthe area selection process
{see Figure 2). Two factors are citical 10
‘good exploration strategy:
(1) optimizing on the cost, in relation to the
ellectiveness, of methods used to determine
the presence or absence offeatures oa which
the area selection process 's based, and
{2)using the appropriate criteria for explora-
tion area selection, and correctly assessing
the relative importance of these. Most of the
papers given at Exploration '87 wore con-
‘corned withthe first factor. tis the secondaspect of exploration strategy that is the
‘main concern ofthis paper.
Models as Systems Which “Rate”
‘Area Selection Criteria
‘Area selection erteria must not only be fea-
tures which ore spatiaty associated with
ore, bulthey must aloo be gonot'cally related
10.re in some way. Without this gonetic ink
the spatial association would be fortuitous or
accidental, and therefore not arable guide
tore. In adéition, area selection criteria, to
bbe elu, must be relatively easily identified,
rormally by field techniques. For examole
the type of gold deposit which occurs in
greenstone Delis ‘s commonly associated
with certain types of felsic hypabyssal
intrusions. and therefore the presence of
these intrusions is a useful area selection
Criterion, readily applied in a field situation
(Hodgeon eta, 1982). Incontrast, the obser-
vation that Fuld incusions in Archean gold
deposits are CO;ich, while genetically sig
rifleant (Wood et al, 1986), Is nol, at our
present level of geological understanding. 2
iter thatcanbe practically applied nthe
area selection process in most instances. In
the case of epithermaltype gold deposits,
fon the other hand, the gas content of fui
BENEFIT ——o>
inclusions Is a eriteion for area selection
(Norman ofa, 1988)
‘Area selection critoria canbe orderedin a
three-dimonsional bierarchy, according to:
(1) the scale at which they are associates
With mineralization, and thus the scaie of
‘area selection for which they are used. (2)
the confidence one has that a feature Is an
‘essential (nt fortuitous) part ofthe ore en-
‘vironment, and (3) their relation (if any) tothe
‘economic quay ofa deposit
‘Area selection oiteria ave scale-specifc:
whats important atone scale may be rrele-
ant at another. For exemple, there is an
‘association of Archean gold mining camps
‘withthe contacts of mafic + ultramafic vol-
‘canic sequences with sedimentary rock se
‘quences, but this feature is ofite or no use
In selecting dil targets within mining camps.
(Figure 3) Large-scale area selection cnt
Fiaare moreimportantthansmal-scale crit
fia, since even the most technically song
program cannot sueceed fits carried out in
AREA SELECTION
CRITERIA
MAFIC » ULTRAMAFIC
Vocac oe
[ARV ROCK CONTACT
FELSIC INTAUBION
MAJOR CARBONATE
[ALTERATION ZONE
MAJOR FAULT
Loca srmucTuRES
‘20LD on INDICATOR.
ELEMENT/MINERAL
‘SHOWING
the wrong general area. However, the scale
at which an area solection criterion is appl
‘cable, and the confidence that one has that
the feature Is related to ore, tend 0 be In
vverselyrelated— itgenerallyis more oifficult
to characterize, and determine the genetic,
relations among large-scale phenomena
than it's among small-scale phenomena.
‘The larger scale ares selection criteria ore
‘commarly considered te define the goologi-
‘cal environment whichis favourable for min
tralzation, whereas the smaller scale fea-
tures define the depost (Figure 4).
‘One ofthe major prablams in expioration's
assessing the roliabilty of criteria for area
selection the second dimension ofthe ating
Hierarchy above. Thore are basicaly two ap-
proaches to the problem. in the first ap-
proach, the distribution of features in the
known natural population of deposit is re-
corded. Features are rated or theirreliabibty
‘2ccording to thelr relative requency of asso-
lation with ore, and their absence in areas,
TARGET
sanana cua one
came ‘@RoUP Bopy
‘Figure 2 Diagrmmetic representation ofthe exploration sequence intrms of
Conlbenett ata, Theoptinal exploration step shat succession of expo-
fation ectivtes wtih in eggregeta has the lowest fte cost. relative fo the
fecanomic retin Exch ie of exploration actly ges diminishing retumon
investment oss pureve, ands replaced by th nex most cast-afkce! pe
of ectty athe docson pons war te sizeof he aoa boing explored is
duced
epost
ENVIRONMENT
Figure 3 Are selection rte fr god capostsinthe Suprir Province ofthe
‘Ganectan Shield, showing how We relive importance of diferent eerie
‘changes 8 largot size decreases during the exploration process (From
Heagsen ot al, 1982).
LTHOGEOCHEMICAL
‘ANOMALY
Figure 4 Diagrammatic stration shoring how the dtnction Between gec-
legal caractonie oh “aepout”: and to geological characteristics of he
“depos envronman” depends on the sca of ie geological feu assoc
tod wan minaratzation
PorsTOn
(OF EPOETS eee co OIL FEATURES 8 zm
Figure § ideatzed method for bung @gonelic modlfor an ore depos ye,
‘and wang Hf ae ares selection extraGeoscience Cenade volume 17. Number 2
without ore, Thus, afesture which occurs in
rine out of ten deposits, but is otherwise
rare, would be considered a much more sig-
nificant area selection criterion than @ fea
ture which occurs in only half the deposits
‘and also occurs in areas without mineraliza-
tion, A model generated in this manner
{termed an “emprical moder”. To construct an
‘empirical model, the presence or absence of
‘each feature must be recorded for a statis-
{ically significant and unbiased sampie of he
‘entre population. Hereinlies the mein week-
ress of the purely empirical approach: deta
{or only a small part ofthe population cannot
'be used to formulate the model. For exam-
le, uid incision orisotope data ona single
‘deposit cannot be used as empirical area
‘selection criteria, because without applying
Interpretive, genetic arguments, isnot pos-
‘sible to say thatthe data wil be cherecteris~
tic of other deposits ofthe total population.
However, these data may be ciilical to un-