Seismic Interpretation of The Exmouth Plateau, North Carnarvon Basin, Australia:
An Application of Data Conditioning, Seismic Attributes, and Self-Organizing Map on 2D Data
Downloaded 01/08/18 to 120.188.73.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Thang Ha*, Brad Wallet, and Kurt Marfurt - University of Oklahoma
Summary
In the last 15 years, 3D data have become commonplace in
seismic interpretation. However, 3D seismic survey are
often limited in size (at most ~100-km-long), thus insuffi-
cient for regional mapping of features that are hundreds to
thousands of km in length. Such 3D surveys are often ac-
companied by previously acquired 2D regional lines. Yet,
due to the 2D nature and older acquisition technique, these
2D lines are usually of lower quality and contain more
noise than the associated 3D data. In this project, we im-
prove seismic image quality, identify regional features and
local anomalies, and analyze seismic facies that are poten-
tially related to hydrocarbon production in the Exmouth
Plateau, North Carnarvon Basin, Australia, by simultane-
ously applying data conditioning, seismic attribute calcula-
tion, and Self-Organizing-Map (SOM) classification to
multiple vintage 2D lines.
Introduction
The North Carnarvon Basin is a major hydrocarbon reserve
in Australia (Chongzhi et al, 2013). It can be divided into
sub-basins (Figure 1). Among these sub-basins, the Ex-
mouth Plateau is the largest and cover most of the major
gas fields. Thanks to such a prolific amount of hydrocarbon
Figure 1. North Carnarvon Basin, Australia (Modified from Chongzhi et al). The study area consists of 55 2D seismic lines
forming a rectangular grid at the center of the Exmouth Plateau, where most of the major gas fields are located.
© 2017 SEG
SEG International Exposition and 87th Annual Meeting
reserve, numerous seismic surveys have been acquired over
the area. Most of the 3D surveys, however, are small
(<5000 km2) and concentrate around known reservoir loca-
tions. This leaves 2D data the only feasible choice to study
regional features and to explore new area that has not been
covered by 3D data.
Our project involves 55 lines forming a rectangular grid,
spanning an area of ~40000 km2 in the center of the Ex-
mouth Plateau (Figure 1). Unfortunately, those 2D lines
were acquired many years before the 3D surveys and thus
have lower image quality. Our goal is to improve seismic
interpretation of these 2D lines via data conditioning, seis-
mic attributes, and SOM classification in order to have a
better understanding of the regional geology as well as
local hydrocarbon-related facies.
Geologic Settings
In general, the North Carnarvon Basin has a geologic histo-
ry of a passive margin with multiple stages of extension,
subsidence, and late minor inversion that involve NE-SW
trending faults (Tellez Rodriguez, 2015). The main source
rocks of the Exmouth Plateau are the Locker Shale, depos-
ited in early Triassic. The Mungaroo fluvio-deltaic for-
mation (middle-late Triassic) and the Flag sandstone of the
Page 2117
 Seismic Facies Analysis: Exmouth Plateau, North Carnarvon Basin, Australia

a b
Time Time
Downloaded 01/08/18 to 120.188.73.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

(s) (s)

2.0 2.0

2.1 2.1
Amplitude Amplitude
Positive Positive
2.2 2.2
0 0
Negative 1km Negative 1km
Figure 2. Seismic image (a) before and (b) after data conditioning. Reflectors are clearer and better resolved (yellow ellipse),
faults are sharper (red arrows), and cross-cutting migration artifacts (green lines) are suppressed in the seismic image after spec-
tral balancing and structure-oriented filtering.
Time Sea Water
(s)
Quaternary Sediment
Fig 5
2.0

Fig 4 Tertiary Carbonate

2.5
Sequences

3.0 Muderong Shale

Amplitude
Positive
0 Triassic Deposit
3.5 Negative 5km
Figure 3. A representative portion of a 2D seismic profile in our project, with interpreted events. Major faults (red lines) termi-
nate against the top of the Muderong shale, consistent with post-rifting subsidence configuration of the area in middle Creta-
ceous. MTCs within the Muderong shale and Tertiary carbonate are chaotic, exhibiting low coherency (Figure 4). Within the
carbonate sequences, wavy pattern of turbidite channels can be observed, exhibiting strong negative curvature (Figure 5).
Barrow delta (early Cretaceous) are the main reservoir conditioning. Thin layers are more resolved, cross-cutting
rocks. During middle Cretaceous, post-rifting subsidence migration artifacts are suppressed, while faults are sharper,
enables a thick deposition of the transgressive Muderong enabling us to have an over-all better interpretation. To
shale throughout the entire basin, which act as a regional perform seismic facies analysis, we compute seismic attrib-
seal (Chongzhi et al, 2013). During the Tertiary, the Aus- utes, including coherence, envelope, and structural curva-
tralian Plate drifted northward to warmer tropical zones, ture, and then input those attributes into SOM automatic
facilitating development of carbonate sequences, including classification (described in detail by Zhao et al, 2016).
the Mandu Limestone formation (Smith, 2014).
Results
Methods A representative portion of 2D seismic data with interpret-
To improve seismic image quality, we follow Hutchinson ed events is shown in Figure 3. Most of the deeper faults
et al’s data conditioning workflow (2016) and apply spec- (below 2.5s time-depth) terminates within the Muderong
tral balancing to the data, followed by an edge-preserving shale, consistent with the post-rifting tectonic setting dur-
structure-oriented filtering (Zhang et al, 2016). Figure 2 ing the Cretaceous. Within the Muderong shale and upper
compares seismic vertical profiles before and after data sequences, mass transport complexes (MTCs) that exhibit

© 2017 SEG Page 2118

SEG International Exposition and 87th Annual Meeting
 Seismic Facies Analysis: Exmouth Plateau, North Carnarvon Basin, Australia

Time Amplitude
(s) Positive
Downloaded 01/08/18 to 120.188.73.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

2.6 0
Negative
Coherence
2.8 High

Low
3.0 1km
Figure 4. Co-rendered coherence and seismic amplitude profile of the region containing Mass Transport Complexes (MTCs).
Seismic amplitude is plotted in red-white-blue color scheme. These MTCs are chaotic, exhibiting low coherency, and are tens-of
-kilometers wide. We found MTCs both within the Muderong shale and the upper carbonate sequences.
Time Amplitude
(s) Positive

2.0 0
Negative
Curvature
2.2 Positive
0
Negative
2.4 1km
Figure 5. Co-rendered negative curvature and seismic amplitude profile of the region containing turbidite channels. Highly neg-
ative curvature values corresponds to channels’ axes, similar to Wallet’s observations (2016). These channels have thickness
ranging from 20 to 100ms (equivalent to ~ 25-125m for v=2500m/s) and half-width ranging from 0.5 to 1.5km.
Time Time
(s) Tertiary Carbonate (s)
Sequences
Tertiary Carbonate
2.0
Sequences
3.0

2.5
Muderong Shale
Amplitude
Positive
0 4.0
3.0 2km
Negative
Figure 6. 3D perspective view showing perpendicular seismic
amplitude profiles at the SW corner of the study area. Orange
arrows point to two amplitude anomalies within the Muderong
shale. These anomalies are apparently 2-km-long. We interpret
these anomalies as bright spots that could be potential gas-
charged reservoirs.
chaotic texture on amplitude profile can be easily seen on
coherence profile (Figure 4). On top of the Muderong
shale, wavy pattern of channels can be observed, which
corresponds to turbidite channels within a carbonate shelf
as described by Wallet (2016). These channels exhibit
strong negative curvature at their axes (Figure 5). Near the
3.5
Amplitude
Positive Muderong Shale
0
2km
Negative
Figure 7. 3D perspective view showing perpendicular seismic
amplitude profiles near the NE corner of the study area. Pink
arrows point to multiple dome-shaped features (~1km apparent
width) within the Muderong shale. These domes are possibly
carbonate mounds developed in the middle Cretaceous that
could be potential reservoirs if filled with organic material.
SW corner of the survey, we notice two bright spots within
the Muderong shale, which could be potential gas-charged
reservoir (Figure 6). In the NE corner of the survey, several
dome-shaped features emerge within the Muderong shale
(Figure 7). We interpret these domes as carbonate mounds
formed during the middle Cretaceous. If they are filled

© 2017 SEG Page 2119

SEG International Exposition and 87th Annual Meeting
 Seismic Facies Analysis: Exmouth Plateau, North Carnarvon Basin, Australia

Time SOM 2D color table

(s)
Downloaded 01/08/18 to 120.188.73.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

1.5

2.0 Amplitude
Positive
0
Negative
Time-Depth(s)
2.0
2.5
2.5
3.0

5km

Figure 8. 3D perspective view showing two profiles of co-rendered SOM results and seismic amplitude, along with the top
Muderong shale horizon. Bright yellow color corresponds to concave channels. Gold color corresponds to chaotic MTCs. Green,
blue, dark red, and purple colors corresponds to bright, coherent reflectors.
with organic material, these mounds could be potential Acknowledgements
reservoirs. We thank Schlumberger for a license to the Petrel interpre-
Lastly, we present a 3D perspective view composed of two tation software, and all of the sponsors of our AASPI con-
profiles of SOM output and a time-structure surface of the sortium. Thanks to Geoscience Australia for providing the
top Muderong shale in Figure 8. The top Muderong shale data used in this project.
can be easily traced across all lines, further confirming the
widespread deposition of transgressive shale during the
middle Cretaceous. Each color of the SOM represents a
different combination of coherency, amplitude envelope,
and structural curvature. Bright yellow color corresponds
to concave channels. Gold color corresponds to chaotic
MTCs. Green, blue, dark red, and purple colors corre-
sponds to bright, coherent reflectors.

Conclusions
By applying data conditioning to the 2D data, we are able
to improve seismic image quality. Seismic attributes, in-
cluding coherence, envelope, and structural curvature, help
us interpret regional geological features (such as the wide-
spread Muderong shale and the turbidite channels within
the Mandu limestone formation), as well as local anomalies
(such as bright spots and carbonate mounds). Analysis of
seismic facies, including MTCs and turbidite channels, can
be accelerated with the aid of SOM classification.

© 2017 SEG Page 2120

SEG International Exposition and 87th Annual Meeting
 EDITED REFERENCES
Note: This reference list is a copyedited version of the reference list submitted by the author. Reference lists for the 2017
SEG Technical Program Expanded Abstracts have been copyedited so that references provided with the online
metadata for each paper will achieve a high degree of linking to cited sources that appear on the Web.

REFERENCES
Downloaded 01/08/18 to 120.188.73.23. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Chongzhi, T., B. Guoping, L. Junlan, D. Chao, L. Xiaoxin, L. Houwu, and W. Dapeng, 2013, Mesozoic
lithofacies palaeogeography and petroleum prospectivity in North Carnarvon basin, Australia:
Journal of Palaeogeography, 2, 81–92, http://doi.org/10.3724/SP.J.1261.2013.00019.
Hutchinson, B., J. Qi, F. Li, O. Olorunsola, and K. J. Marfurt, 2016, Pitfalls in poststack data
conditioning: Distinguishing enhanced geology from enhanced artifacts: 86th Annual
International Meeting, SEG, Expanded Abstracts, 1783–1787, http://doi.org/10.1190/segam2016-
13970185.1.
Smith, M., 2014, Cenozoic stratigraphy of the North Carnarvon Basin: Insights for the growth history of
carbonate margins of the North-West Shelf: Master’s thesis, University of Western Australia.
Tellez Rodriguez, J. J. J., 2015, Seismic-sequence stratigraphy framework and architectural elements for
Cenozoic strata at the Rankin Platform sub basin, north Carnarvon basin, Australia: Master’s
thesis, University of Oklahoma.
Wallet, B. C., 2016, Attribute expression of channel forms in a hybrid carbonate turbidite formation:
Interpretation, 4, SE75–SE86, http://doi.org/10.1190/INT-2015-0108.1.
Zhang, B., T. Lin, S. Guo, O. E. Davogustto, and K. J. Marfurt, 2016, Noise suppression of time-migrated
gathers using prestack structure-oriented filtering: Interpretation, 4, SG19–SG29,
https://doi.org/10.1190/segam2015-5921220.1.
Zhao T., J. Zhang, F. Li, and K. J. Marfurt, 2016, Characterizing a turbidite system in Canterbury basin,
New Zealand, using seismic attributes and distance-preserving self-organizing maps:
Interpretation, 4, SB79–SB89, https://doi.org/10.1190/INT-2015-0094.1.

© 2017 SEG Page 2121

SEG International Exposition and 87th Annual Meeting