The Windimurra Igneous Complex, Yilgarn Craton: An Archean Layered Intrusion Revealed by Seismic Data and 3D Modelling

Government of Western Australia
Department of Mines and Petroleum
RECORD 2015/12
THE WINDIMURRA IGNEOUS COMPLEX, YILGARN

CRATON: AN ARCHEAN LAYERED INTRUSION REVEALED
BY SEISMIC DATA AND 3D MODELLING
by
TJ Ivanic and J Brett
Geological Survey
of Western Australia
Record 2015/12
THE WINDIMURRA IGNEOUS COMPLEX, YILGARN

CRATON: AN ARCHEAN LAYERED INTRUSION
REVEALED BY SEISMIC DATA AND 3D MODELLING
by
Perth 2015
MINISTER FOR MINES AND PETROLEUM
Hon. Bill Marmion MLA
DIRECTOR GENERAL, DEPARTMENT OF MINES AND PETROLEUM

Richard Sellers
EXECUTIVE DIRECTOR, GEOLOGICAL SURVEY OF WESTERN AUSTRALIA

Rick Rogerson
REFERENCE
The recommended reference for this publication is:
Ivanic, TJ and Brett, J 2015, The Windimurra Igneous Complex, Yilgarn Craton: an Archean layered intrusion revealed by seismic data and
3D modelling: Geological Survey of Western Australia, Record 2015/12, 28p.
National Library of Australia Card Number and ISBN 978-1-74168-640-1
Grid references in this publication refer to the Geocentric Datum of Australia 1994 (GDA94). Locations mentioned in the text are
referenced using Map Grid Australia (MGA) coordinates, Zone 50. All locations are quoted to at least the nearest 100 m.
Disclaimer
This product was produced using information from various sources. The Department of Mines and Petroleum (DMP) and the State
cannot guarantee the accuracy, currency or completeness of the information. DMP and the State accept no responsibility and disclaim
all liability for any loss, damage or costs incurred as a result of any use of or reliance whether wholly or in part upon the information
provided in this publication or incorporated into it by reference.
Published 2015 by Geological Survey of Western Australia

This Record is published in digital format (PDF) and is available online at <www.dmp.wa.gov.au/GSWApublications>.
Further details of geological products and maps produced by the Geological Survey of Western Australia
are available from:
Information Centre
100 Plain Street
EAST PERTH WESTERN AUSTRALIA 6004
Telephone: +61 8 9222 3459 Facsimile: +61 8 9222 3444
www.dmp.wa.gov.au/GSWApublications
Contents
Abstract...................................................................................................................................................................1
Introduction.............................................................................................................................................................1
Previous work...................................................................................................................................................4
Igneous stratigraphy.........................................................................................................................................7
Discordance......................................................................................................................................................9
Contact relations and structural overprint........................................................................................................9
Input data for the 3D model..................................................................................................................................10
Geophysical data ...........................................................................................................................................10
Physical properties.........................................................................................................................................11
Seismic processing.........................................................................................................................................11
GeoModeller .................................................................................................................................................11
Construction of the 3D model........................................................................................................................13
Lithological packages and shear zones..........................................................................................................13
Results...................................................................................................................................................................14
Interpretation of the seismic lines across the WIC ........................................................................................14
Results from 3D modelling............................................................................................................................21
Discussion.............................................................................................................................................................23
Implications of the first order data and constraints........................................................................................23
Thickness estimates................................................................................................................................23
Discordance.............................................................................................................................................23
Ultramafic zone.......................................................................................................................................24
Implications of the inversion results..............................................................................................................24
Evidence for a genetic model.........................................................................................................................25
Conclusions....................................................................................................................................................25
Acknowledgements...............................................................................................................................................27
References.............................................................................................................................................................27
Figures
1. Simplified geological map of the northern Murchison Domain ...................................................................2
2. Interpreted geological map of the Windimurra Igneous Complex ...............................................................3
3. Photograph looking south in the Windimurra Vanadium pit ........................................................................4
4. Location of the Youanmi seismic lines in the vicinity of the Windimurra Igneous Complex ......................5
5. Previous interpretation of the igneous stratigraphy of the Windimurra Igneous Complex ..........................6
6. Igneous stratigraphic columns for layered igneous complexes of the Murchison Domain .........................8
7. Seismic data and interpretation along the Youanmi seismic lines ..............................................................16
8. Gravity and magnetic data used in 3D modelling ......................................................................................18
9. Revised igneous stratigraphic column for the Windimurra Igneous Complex ..........................................24
10. Schematic genetic model for the formation of the Windimurra Igneous Complex ....................................26
Tables
1. Summary of rock properties used in Ahmat (1986) ...................................................................................11
2. Summary of down-hold geophysics results from MNDD02 and MNDD04 ..............................................12
3. Summary of rock density determinations for GSWA samples ...................................................................12
4. Coordinate limits for the modelled volume ................................................................................................12
5. Details of the modelled stratigraphic column ............................................................................................13
6. Magnetic susceptibilities and densities for each unit .................................................................................15
7. Modelled shear zones in the ‘fault matrix’ .................................................................................................15
8. Summary of thickness constraints for the Windimurra Igneous Complex .................................................20
9. The volumes of the various modelled units in Model 2 .............................................................................21
10. Coordinate limits for the modelled volume ................................................................................................21
11. The volumes of the various modelled units in Model 3 .............................................................................22
12. Inversion parameters for runs 1–4 ..............................................................................................................22
13. The volumes of the various modelled units ................................................................................................22
Appendix*
1. 3D model metadata table to accompany ‘Windimurra, 2015: 3D geomodel series’
* The 3D geomodel to accompany this Record is provided separately on a USB. The Record may be accessed
directly on our website independent of the USB.
iii
iv
GSWA Record 2015/12 The Windimurra Igneous Complex, Yilgarn Craton: 3D modelling
The Windimurra Igneous Complex, Yilgarn Craton:

an Archean layered intrusion revealed by
seismic data and 3D modelling
by
Abstract
The Windimurra Igneous Complex is situated in the western part of the Archean Yilgarn Craton, Western Australia within the Youanmi
Terrane. It is part of the plume-related and laterally extensive c. 2.81 Ga Meeline Suite, an anhydrous tholeiitic suite comprising five
layered mafic–ultramafic intrusions 25–80 km in diameter. The intrusions also host significant V–Ti mineralization in their fractionated,
Fe-rich upper zones. The Windimurra Igneous Complex is the only intrusion to contain two upper zones, separated by the Shephards
Discordant Zone. Recent mapping, aeromagnetic and gravity surveys, and the Youanmi Seismic lines have provided unparalleled 3D
constraints on the Windimurra Igneous Complex. These data have allowed direct assessment of the zone thicknesses and facilitated
construction of a 3D model. This model was refined by forward modelling of potential field data and tested against a geophysical
inversion. The results show that the complex is thicker than previously predicted at approximately 11 km, which make it the thickest
layered mafic–ultramafic intrusion identified globally. Relative to the consistency of layered igneous units, three significantly discordant
units are identified: (1) the Shephards Discordant Zone separating the older, basal eastern lobe from the western lobe; (2) the base of the
highly transgressive upper zone of the western lobe; (3) the roof zone and overlying Kantie Murdana Volcanics Member. The geometry
of these three features provides constraints on a multistage genetic model for magma emplacement. In addition, the indication of a
thick, subsurface ultramafic zone provides a potential target for Ni–Cr–PGE mineralization.
KEYWORDS: layered intrusions, Archaean, magma chamber, magmatic deposits,Youanmi Terrane
Introduction the results of a mantle plume affecting a wide region of

the Yilgarn Craton (Ivanic et al., 2010; Wyche et al., 2012;
Layered mafic–ultramafic intrusions are host to a variety of Wyman and Kerrich, 2012; Van Kranendonk et al., 2013).
ore deposits such as Ti–V and Ni–Cr–Cu–PGE (Platinum The central–northern part of the WIC (Fig. 2) is overlain
Group Elements). These deposits are associated with by the rhyolitic Kantie Murdana Volcanics Member of the
magma chamber processes and influenced, in particular, Yaloginda Formation of the Norie Group. These rhyolitic
by cumulate layering geometry. Therefore, it is essential to volcanics have also been directly dated at 2813 ± 3 Ma
understand the large-scale primary 3D geometry of these (Nelson, 2001). There are several economic vanadium
intrusions in order to assess the nature of magma chamber deposits hosted within the magnetitites of the upper zone
processes and prospectivity for ore deposit formation. (Fig. 3).
The Windimurra Igneous Complex (WIC) is a giant The WIC is dissected by larger shear zones (e.g. Challa
layered mafic–ultramafic intrusion ~2500 km2 (85 km Shear Zone and Yarloo Shear Zone) and also smaller
north–south by 37 km east–west) located in the central brittle faults. Hence parts of the WIC occur as sheared
Youanmi Terrane, Yilgarn Craton, Western Australia lenses to the south and east of the main body, especially
(Fig. 1). It is the largest, relatively intact and exposed, east of the Challa Shear Zone (unassigned gabbros,
mafic–ultramafic intrusion in Australia. Recent work Fig. 2). Based on aeromagnetic images, these features are
has directly dated the complex and assigned it to the mapped as up to 25 km north–south and 5 km east–west
c. 2813 Ma, anhydrous Meeline Suite (Ivanic et al., 2010; lenses (Figs 2 and 4).
Wingate et al., 2012), which includes other large layered
mafic–ultramafic intrusions such as the Youanmi Igneous There is an overall progression from more mafic rocks in
Complex (YIC). The Meeline Suite intrudes into the Norie the lower zone to large volumes of leucocratic rocks in
Group of the Murchison Supergroup (Van Kranendonk the middle and upper zones (Figs 5 and 6). Within these
et al., 2013). It is widely regarded that these intrusions are zones, megacyclicity on an approximately 200 m scale
1
Ivanic and Brett
118° 122° 118° 119°

117°
YE
R Gnanagooragoo Meekatharra
RR NE
10
00
26° NA RRA Igneous
60
TE
GA
Complex
BUR
-Y
T
EASTERN
VILL
Murchison
Domain GOLDFIELDS
U1
E
SUPERTERRANE
00
KUR
YOUANMI
80
KAL
30°
NAL
TERRANE 27° 27°
Southern GOO
PI
Cross
Domain
RLIE
00
INDIAN SOUTH WEST
0
OCEAN
TERRANE
10
34° 50 km
Terrane boundary
00
Cue
0
12
Fig. 4
00
0
Windimurra
14
Igneous
Fig. 2 Complex
10GA-YU2
0
28° 28°
700
10GA-YU3
800
Mt Magnet
600
0
0
600
0
0
800
00
93
400
100
112
Yalgoo
Youanmi
Igneous
29° Complex 29°
Narndee
Igneous
Paynes Find Complex
YO W
UA TE
S
NM R
IT 50 km
RA ERR
NE AN
E 117° 118° 119°
TJI53a 21.08.15
plutonic rocks supracrustal rocks
Granitic and Mougouderra Fm. Yalgoo Townsite
metagranitic rocks sedimentary rocks
Unconformity
Yalgowra Suite Glen Group
0
mafic–
800
Warriedar Suite Polelle Group Seismic line, with CDP points

ultramafic
intrusive Norie Group/
suites Boodanoo Suite felsic members, includes KMVM
Meeline Suite c. 2.95 Ga formations
Figure 1. Simplified geological map of the northern Murchison Domain, highlighting the location of mafic–ultramafic intrusive
suites relative to supracrustal greenstone groups and granitic rocks. Also shown are the locations of the three
seismic lines (10GA-YU1, 10GA-YU2, 10GA-YU3). Modified from (Ivanic et al., 2010). Inset shows broader setting of
the Murchison Domain within the Youanmi Terrane of the Yilgarn Craton. KMVM = Kantie Murdana Volcanics Member
of the Yaloginda Formation. Corner extents of Figures 2 and 4 indicated in green and red, respectively.
2
118°30'
Bald Rock Supersuite:
2640–2600 Ma post-tectonic granites
10
Metamonzogranites:
GA
2710–2640 Ma Big Bell and Tuckanarra suites
-Y
Mount Kenneth Suite:
U1
c. 2800 Ma
0
00
metagranitic rocks
15
ne
Kantie Murdana Volcanics
Zo
c. 2813 Ma
Member rhyolitic volcanics
ar
e
Sh Unassigned gabbros
loo
60 40
0
r
00
Ya
Corner Well Gabbro
16
75
70
28°00' Roof zone
00
Windimurra
c. 2813 Ma Meeline Suite

Mingyngura Hill 35 10GA-YU2
Igneous
6000
Complex Upper zone
9
49
c. 2813 Ma magnetitite-bearing rocks
16
EASTERN LOBE
Middle zone
50
00
Lower zone
2813
Ultramaﬁc zone
00
<2821 Border zone dolerites

00
90
10G
10
20 A-YU
3 Youanmi Upper zone
Ch
40
2813 Igneous
0
3
all
0
00
9
110
112
Complex
a
w c. 2819 Ma Middle zone

Sh
10 35 Greenstone supracrustal
Shear ndoo
Zone
ea
20 2820 – 2810 Ma lithologies (dominantly Norie

r Z
2
4 Group)
Wyema
on
30 2813 Geochronology sample site showing age

e
1 in Ma
3 w V mineralization
2813 WESTERN (W=Windimurra vanadium pit)
LOBE 60
2 Canegrass resource diamond drillholes
2677 (MNDD001–4)
PGE (Ni, Cu, Cr) mineralization
20
28°30' Igneous layering (right way up)
70
Chromitite seam
Fault
Shear zone
70 Trend of igneous layering
Muleryon Hill 40
Shephards Discordant Zone
Windimurra Igneous Complex
2819
Youanmi 20 Western Lobe
Igneous
0 0
Seismic line showing CDP points

90
Complex
70
75
TJI63 02.11.15
20 km
Figure 2. Interpreted geological map of the Windimurra Igneous Complex, highlighting available geochronological data and
mineralization styles. Note the location of the Shephards Discordant Zone separating the eastern and western lobes
and the location of three seismic lines (10GA-YU1, 10GA-YU2, 10GA-YU3). Area of map is a few kilometres wider than
that of the modelled volume (Fig. 9), which does not include modelling of the Youanmi Igneous Complex, shown
here to the southeast. The 3D geomodel layer ‘Images–Geological maps’ represents the original input data for the
3D geomodel and is superceded here.
3
Ivanic and Brett
New gravity and magnetic survey data acquired from

2004 onwards greatly improved the resolution of coverage
over the WIC area. The magnetic surveys have a line
spacing of 400 m, gridded at a resolution of 80 m. The
gravity surveys have a data spacing of 2.5 km, gridded at
a resolution of 400 m.
Recent and extensive mapping at 1:100 000 scale
(simplified in Fig. 2) incorporated the existing 1: 50 000
data (Ahmat, 1986) and provided new zone nomenclature
for the WIC and adjacent gabbroic and granitic intrusions
(Ivanic, 2011; 2012a,b; 2014; Zibra et al., 2013). These
map sheets provide five cross-sections constrained from
surface observations (Fig. 4). From this mapping and the
new gravity data, with preliminary 2D profile modelling,
it was concluded that the WIC was approximately 6 km
TJI64 10.06.15 thick, broadly concentrically layered and bowl shaped
Figure 3. Photograph looking south in the Windimurra with more steeply dipping igneous layering at the margins
vanadium pit showing moderately west-dipping (Ivanic et al., 2010, 2014; Ivanic, 2014). This is deeper
layers of magnetitite (dark layers) hosted in than the estimates of Ahmat (1986) and allow for a larger
magnetite-bearing leucogabbro (pale layers) of proportion of ultramafic rocks at the base of the complex.
the upper zone of the eastern lobe. Pit wall is 30 m
high. Note the multiscale cyclicity of mafic and felsic Additional support for this depth came from the results of
layers creating a heterogeneous rock volume on a
three seismic reflection lines. In 2010, seismic lines were
centimetre-to-decametre scale.
planned in order to traverse particular parts of the WIC
utilizing existing roads and tracks to provide sufficient
data for a 3D seismic interpretation of the complex
(Fig. 4). This goal was greatly enhanced by the inclusion
of a third line 10GA-YU3 to provide an east–west cross-
can be related to at least 13 documented reversals in the section of the WIC, supplementing 10GA-YU1 and
chemostratigraphy (Ahmat, 1986). The exposed complex 10GA-YU2 which both have terminations within the
has an extremely felsic overall composition compared to complex. These transects in the vicinity of the WIC had
other layered gabbros worldwide and has been described four particular foci: (1) to image across what was thought
as a having an anorthositic affinity (Ahmat and De Laeter, to be some of the deepest parts of the complex as seen
1982) and a high Ca–Fe tholeiitic composition (Ahmat, in gravity data; (2) to image across the SDZ and provide
1986). However, it is possible that its composition is more evidence for the mechanism of emplacement of rocks to
typically tholeiitic (c.f. Nebel et al., 2013a) if it can be the west of this zone; (3) to image and trace the extent of
shown that a large volume of ultramafic zone material economically significant magnetitite horizons of the upper
exists at depth. zone of the complex; (4) to image across the Challa Shear
Zone and the associated metamonzogranite plutons.
The magnetotelluric (MT) conductivity model along
Previous work the Youanmi seismic lines shows moderate to low
Previous work documenting the geometry of the complex conductivity for the ultramafic and lower zones of the
involved 1:50 000 surface mapping of igneous and complex. This feature appears to be a single, long-
structural data (Ahmat, 1986), which concluded that the wavelength region. Higher conductivity is evident in the
form of the WIC was tabular and broadly comprised surrounding foliated granitic rocks, possibly indicating
inwardly dipping igneous layering. This work included interconnected conductive minerals in these rocks;
forward modelling two gravity transects set in a regional, for example, within the biotite foliation planes. Short-
low-resolution, gravity survey area covering the WIC. wavelength anomalies are present in the upper parts of
The gravity modelling utilized simple slabs of estimated the WIC, corresponding to the magnetite-rich upper zone
densities and concluded that the WIC extended down to and roof zone. A 10 km low conductivity feature to the
an average of 3 km depth and locally down to 5 km in the east of the complex on this line is thought to represent late
vicinity of these transects. This was clearly at odds with granitic plutons.
the cumulative stratigraphic thickness logged at 13 km Until acquisition of the Youanmi seismic and MT survey,
from surface observations (Fig. 5). This discrepancy very little attention had been paid to the 3D form of
was explained by the presence of significant discordant the WIC as a whole. Initial interpretations for each of
features, which were mapped throughout the lower, middle the three seismic lines in the vicinity of the complex
and upper zones including the Shephards Discordant have been presented (Ivanic et al., 2014; Wyche et al.,
Zone (SDZ) (see further discussions below). The Kantie 2014). Here we present updated interpretations of the
Murdana Volcanics Member (including ‘mafic intrusives’ seismic data and integrate this in 3D with new gravity,
presumably of what we now call the roof zone of the and magnetic forward models, and drillcore data along
WIC), was modelled to extend down to a maximum of with surface mapping. This modelling is used to tie
1.5 km depth.
4
0 00
14
10
GA
-Y
U1
0 00
15
CHALLA2
00
10GA-YU2
0
16
28°00'
00
60
70
00
60
99
00
Mount Magnet
70
00
4
16
00
80
50
10GA
00
00
-YU3
90
000
10
AY
HIGHW
3
0
9
110
112
40
00
HERN
CHALLA1
NORT
T
GREA
28°30'
EW1
Youanmi
COOLA1
29°00'
Paynes Find
118°00' 118°30'
TJI42a 10.06.15
25 km
Figure 4. Location of the Youanmi seismic lines in the vicinity of the Windimurra Igneous Complex (cf. Fig. 1) in
relation to combined gravity data (colour) and first vertical derivative magnetic data (shading). In black
are the cross-section lines used in the 3D modelling where ‘EW1’ is the additional line used to further
constrain the southern part of the Windimurra Igneous Complex. Challa1, Challa2 and Coola1 are taken
from the cross-sections of the 1:100 000 geological map of the same name.
5
Ivanic and Brett
TJI65 10.06.15
6
Figure 5. (left) Previous interpretation by Ahmat (1986) of the In summary, the complex is divided into eight parts (see
igneous stratigraphy of the Windimurra Igneous Fig. 2):
Complex showing from left to right: perpendicular
stratigraphic height, zone nomenclature, 1. A doleritic border zone (equivalent to the ‘Border
traverse locations, lithologies, textures and the
Group’ of Ahmat, 1986). This is typically 100 m
chemostratigraphy of cumulus (solid lines) and
intercumulus (dashed lines) phases. Note the
thick and occurs along the boundary between sheared
diagonal dashed lines representing the Shephards supracrustal rocks and the lower zone gabbros. This
Discordant Zone; however, a clear interpretation for zone is thought to represent the chilled primary
this was lacking. Details of lithological and other magma, although it is typically recrystallized at
acronyms are available in Ahmat (1986). amphibolite facies conditions. Examples of granulite
facies contact metamorphism of pelitic xenoliths and
wall rocks (likely supracrustal rocks of the Norie
Group) are preserved in a few locations at the margins
of the complex.
surface observations with seismic constraints and to test
the validity of current interpretations of the form of the 2. The ultramafic zone is only exposed in the small
complex; in particular, the form and relationships of the outcrops close to Muleryon Hill adjacent to the Challa
more controversial units: the Kantie Murdana Volcanics Shear Zone as a lens detached from the main body of
Member, the SDZ and the ultramafic zone. A variety of the WIC. It is characterized by abundant peridotite
geometries of the various components of the complex with common accessory-disseminated chromite.
were assessed using forward and inverse 3D modelling of
gravity and magnetic data. 3. The lower zone hosts olivine-rich gabbros and
gabbronorites which grade upwards into more
Utilizing these geometric constraints, inferences are made leucocratic gabbroic rocks, typically without oxides.
concerning the large-scale magmatic process history for These rocks are modally layered on a centimetre-
the emplacement of the WIC. In terms of broad layering to -metre scale and the rock types are repeated on a
geometry, we show that the complex can be divided 200 m vertical scale. Several kilometres of lower zone
into two discrete bodies: a lower eastern lobe and an are apparently repeated in the western lobe, west of the
upper western lobe with a combined thickness of 11 km. SDZ.
Most significantly, we show that there is a voluminous
ultramafic zone that likely underlies the entire complex, 4. The middle zone is composed of troctolitic rocks with
which is a potential target for mineral exploration. intercumulus magnetite, which are layered in a way
similar to the lower zone. This zone appears to be
repeated in the western lobe, west of the SDZ.
Igneous stratigraphy 5. The upper zone is marked by the incoming of
Understanding the stratigraphic relations of the various vast thicknesses of cumulus magnetite and the
zones of the complex is essential for reconstructing disappearance of Mg-olivine (Fig. 5). Within a few
its geometry. This stratigraphy represents the primary hundred metres of the base of the upper zone, Fe-
magmatic control on the igneous sequence onto which are olivine appears as a cumulus phase with magnetite.
superimposed crosscutting and deformative features (see The majority of rocks in this zone are composed of
next section). Furthermore, an appreciation of the variation magnetite-bearing leuconorite and anorthosite, with
in the lithologies within these zones allows for a realistic magnetitite locally abundant. The upper zone of
representation of the complicated density and magnetic the eastern lobe is truncated by the SDZ, whereas
structure. the upper zone of the western lobe appears to have
intruded as a single pulse that scoured down into
The stratigraphy of the WIC (Fig. 6) is comparable in the cumulates of the middle zone of the western
composition to many other large layered mafic–ultramafic lobe (Nebel et al., 2013b). This zone is thicker and
intrusions such as the Bushveld Complex (Cawthorn and apparently more developed in the western lobe;
Walraven, 1998) and others in the Youanmi Terrane. For however, it is likely that much of the upper zone of
example the Youanmi, Atley, Barrambie, Lady Alma, and the eastern lobe is not exposed. There are several
Igneous Complexes are noted to range from 3 to 5 km economic vanadium deposits hosted within the
in thickness (Fig. 6) and possess distinctive magnetitite magnetitites of the upper zone (Fig. 3).
units in their upper zones. Notably, the Narndee and
Gnanagooragoo Igneous Complexes lack these and are not 6. The Corner Well Gabbro is a late phase of peridotitic-
assigned to the same suite as the WIC. gabbroic pipes, 0.3–2 km in diameter, which have
been observed to intrude into the middle and lower
The cumulative thickness of the exposed layers of zones of the western lobe of the complex (and possibly
gabbroic and ultramafic rocks in the WIC is 13 km; the lower zone of the eastern lobe). This phase may
however, they exhibit significant lateral aggradation be consanguineous with the magmas responsible for
or offlap geometry (Fig. 5; Ahmat, 1986). Thus, this the main lobes of the WIC; however, geochemical
geometry yields a true thickness estimate of about 6 km evidence is not conclusive.
(Ivanic et al., 2010) as derived from surface constraints
(Fig. 6). This is approaching that of the 8.1 km thick 7. The roof zone of the WIC comprises kilometre-scale
Bushveld Complex. tabular plutons of unlayered dolerite and porphyritic
7
Bushveld BOODANOO YALGOWRA
Igneous SUITE SUITE
Complex1 (c. 2.8 Ga) MEELINE SUITE (c. 2.81 Ga) (c. 2.72 Ga)
(2060 Ma)
Narndee Windimurra Youanmi Lady Alma Atley Barrambie Barloweerie Gnanagooragoo
2
Igneous Igneous
Ivanic and Brett
Igneous Igneous Igneous Igneous gabbro Igneous

8000 Complex3 Complex
Complex2 Complex4 Complex5 Complex (2711 Ma) Complex6
(2800 Ma) (2813 Ma) (2919 Ma) (<2821 Ma)
W. Lobe E. Lobe
7000
Upper Zone
zone
upper
zone
zone
upper zone
upper
upper
zone
zone
upper
upper
?
6000
S
zone
undesignated
middle
middle zone
5000
middle zone
main zone
lower zone
H E P H A R D S
Main Zone
D
middle zone
lower zone
4000 ? ?
8
middle zone
Vertical thickness (m)

I S C O R D A N T
3000 predominantly leucocratic
? norite
Z
rocks
lower zone
zone
lower
hornblende-bearing mafic
–ultramafic rocks pyroxenite
O N E
dolerite (predominantly
lower zone
<3 mm grain size) troctcolite
2000
Critical Zone
diorite (apatite-bearing) wehrlite
magnetitite interlayered with peridotite-gabbronorite-
leucogabbro anorthosite cycle
? ? peridotite including dunite-
1000 anorthosite harzburgite-lherzolite
gabbro magnetitite layers
ultramafic zone
zone
Lower Zone
ultramafic
gabbronorite chromitite layers
0
TJI45b 21.08.15
Figure 6. Comparative igneous stratigraphic columns (adapted from Ivanic et al. 2010) for layered igneous complexes of the Youanmi Terrane, compared with the Bushveld Igneous
Complex. Simplified stratigraphy is aligned with occurrence of the lowermost upper zone, typically associated with thick magnetitite horizons in the Meeline Suite. Note
the lobe and zone interpretation from Ivanic et al. (2014) which shows the repetition within the east and west lobes of the Windimurra Igneous Complex. Also note the
chromitite horizons in comparison to the critical zone of the Bushveld Igneous Complex. Notes: 1adapted from Cawthorn and Walraven (1998); 2zone nomenclature adapted
from Scowen (1991); 3age from Ivanic et al. (2010); 4age from Gill (2011), zone interpretation from Ivanic (2014); 5Wang (1998); 6Ivanic (2009), zones adapted from (Parks, 1998)
dolerite and gabbro. The basal contact with layered deformation associated owing to the inferred liquid-
gabbros of the lower zone is not exposed, but its present accommodation of any movements.
upper contact locally has spinifex-textured gabbros
in direct contact with rhyolites of the c. 2813 Ma The SDZ, which is at least 20 km long, represents a
Kantie Murdana Volcanics Member. According to significant break in the igneous stratigraphy (several
aeromagnetic data, the roof zone also appears to kilometres, see Fig. 6). However, the precise origin of
truncate layered rocks on either side of the SDZ. This this feature is still not fully understood since arguments
zone is interpreted to represent a final phase of shallow- for both structural and magmatic processes have been
level mafic plutonism whose upper contact chilled postulated (Ahmat, 1986; Bunting, 2004). It is clear that
against and intruded into overlying rocks, separating the SDZ separates the complex into two main parts: an
them from the layered sequence below. The magma eastern lobe and an apparently transgressive western lobe,
may not have been consanguineous with other parts both of which contain a lower, middle and upper zone of a
of the WIC. Immediately overlying the roof zone are similar nature. However, what becomes of this discordant
tonalitic rocks of the c. 2813 Ma Mount Kenneth Suite relationship underneath the Kantie Murdana Volcanics
and 5–50 m scale apophyses of roof zone dolerite Member and to the west is not clear.
intrudes into these and the Kantie Murdana Volcanics
Member. Another significant discordant feature is located at the
base of the upper zone magnetitites in the vicinity of the
8. Unassigned units of the complex in several locations Canegrass vanadium/iron resource in the western lobe of
along the Challa Shear Zone are detached from the complex (diamond drillholes 1–4, Fig. 2). Here, the
any known stratigraphy and are not of distinctive upper zone has an undulating basal contact which appears
composition, most units being metagabbros with to have scoured into underlying middle zone rocks,
variable preservation of igneous textures. Knowledge creating a ‘base breccia’ identified in drillcore (Nebel
of these units is also poor because the outcrop is minor et al., 2010). The relationship of this feature to the SDZ is
(Ivanic, 2011) due to the proximity of the Challa not clear from surface observations.
salt lake system. They include: Boodanoo Hill area,
Watsons Well area, West Challa area, Moolyawarda
Hill area, and the North Bore area (east of the Challa Contact relations and structural
Shear Zone) as mentioned in Bunting (2004).
overprint
In addition to the eight parts described above, there are
several other components which may be attributed to the There are many unknowns remaining concerning the
complex: contact relationships of the WIC and younger structures in
the area. These are mainly at depth, but also near surface
a) A significant ultramafic zone directly underlying the where many key contacts and faults are under cover. 3D
lower zone and not exposed at the surface. This was modelling presented here aims to clarify many of these
postulated to account for the excess in leucocratic rocks issues and what follows is a description of the state of our
in the exposed complex as a whole (Ivanic et al., 2010). knowledge prior to the modelling in this study.
This unit might include a preserved feeder system.
Little is known about the basal contact of the WIC as
b) Amphibolitic xenoliths up to 300 m in size within it is unexposed. It is inferred from seismic data to be
metagranitic rocks in a c. 50 km radius around the predominantly intruded by later granitic rocks (see below;
complex. Ivanic et al., 2014). However, at the time of intrusion of
the complex, the substrate was likely c. 3 Ga basement
c) It is possible that either the Youanmi or the Atley gneisses of the Youanmi Terrane. A chilled border zone
Igneous Complex (Fig. 1) or both could have been (<500 m thick) and an ultramafic cumulate belonging to
part of the WIC as separate ‘lobes’, now separated by an ultramafic zone of the WIC was likely in contact with
a 12–32 km width of granitic plutons. this host rock.
d) Significant thicknesses of portions of the complex may The WIC is in sheared contact with host rocks of the
have been lost through erosion. Norie Group in several places around its northwestern,
e) An area of high-gravity response to the south of the northern and eastern sides. In these areas, some xenoliths
Narndee Igneous Complex (Fig. 4), approximately of supracrustal country rocks are preserved, which
30 x 30 km, may be a sinistrally displaced part of suggest that the complex intruded into Norie Group host
the western lobe of the WIC (i.e. with c. 80 km rocks. The border zone is postulated to have formed by
displacement). chilling of magma along these contacts (Ahmat, 1986)
immediately prior to the accumulation of the layered series
of zones.
Discordance The roof zone, as described above, is thought to represent
A discordant zone within a layered magmatic body is a a complex zone of late-stage intrusive gabbroic rocks
plane of ‘unconformity’ between two layered packages which were chilled locally at their upper contact
whereby one package either onlaps onto or truncates the with the Kantie Murdana Volcanics Member. Contact
layers of the underlying package. This is differentiated metamorphism is thought to be responsible for
from a fault or shear zone because there is little or no recrystallization of rhyolitic rocks of this member. The
9
Ivanic and Brett
Mount Kenneth Suite of granodioritic plutons is also Abundant, north–south trending, steeply dipping brittle
thought to be associated with the rhyolitic volcanics faults in the central part of the complex offset contacts
(Ivanic, 2012a) and a few of these are mapped at the between the various zones described (Fig. 2). Although the
surface adjacent to the Kantie Murdana Volcanics Member. displacement cannot be measured directly, the mechanism
is consistent with normal faulting (including antithetic
Metamonzogranites (c. 2700 Ma Big Bell and the faults) and 50–500 m of offset based on map patterns.
c. 2680 Ma Tuckanarra Suites) are observed to be in A set of northeasterly trending brittle faults in the south
sheared contact with the lower, middle, upper and border of the main body of the complex dextrally offset the
zones of the complex, where it is evident that they truncate Wyemandoo Shear Zone with strike-slip displacement
many of the macroscopic igneous layers on aeromagnetic of up to 1000 m based on map patterns. Another set of
maps. These layers are locally deformed into parallelism brittle faults in the northeastern part of the complex trends
with shear zones and metamonzogranites. Bald Rock northwest and dextrally offsets igneous layering, the
Supersuite granites intrude across the southern part of the border zone and lateral shear zones around the complex
complex, where they are interpreted to have cut across by between 50 and 3000 m.
sheared layering of the complex (Ivanic, 2011; 2012b;
2014). In addition, Wogala Suite Li-rich pegmatites
(belonging to the Bald Rock Supersuite) crosscut layers
of the lower zone of the complex. According to seismic Input data for the 3D model
observations (Ivanic et al., 2014), it is likely that these
four suites of granitic rocks intrude into the basal contact Good geophysical data coverage in the area around
of the complex at considerable depth, possibly obscuring the WIC has enabled this study to advance the limited
much of the original basal contact with basement gneisses. understanding of the complex derived from surface
geological observations and to augment conceptions
There are several shear zones observed around the about the vertical extent and original architecture of the
western, northern and eastern parts of the complex where complex. The construction of the 3D model was intended
metagranitic rocks are typically in contact with sheared specifically to be simplified (i.e. incorporating only
supracrustal rocks (Fig. 2). The Yarloo Shear Zone deforms first order structures and lithological zones), utilizing
metagranitic rocks at the northwest of the complex and only orientation and contact data from well understood
appears from aeromagnetic data to be crosscut in the south regions of outcrop and in the seismic data, validated
by the Challa Shear Zone (Ivanic, 2012a). The Challa against gravity data. Moreover, a low-resolution model is
Shear Zone is the most significant shear zone to affect the considered suitable given that there is still a high degree of
WIC since it is responsible for transportation of lenses of uncertainty about the location of several key components
the complex up to (and possibly further than) 30 km away of the complex. The effective resolution of the 3D map is
from the core of the complex (apparently on a large scale, coarse; it is essentially a 1:1 000 000 scale product with a
with a sinestral sense). However, there are locally abundant 1 km resolution.
dextral shear sense indicators observed along this shear
zone and hence it is likely that it has subsequently been Although providing limitations in terms of detail, this
reactivated in a dextral sense at c. 2680 Ma, the age of syn- methodology has enabled a geological approach to
tectonic Tuckanarra Suite plutons in the area. dominate the study. Emphasis is given initially to high-
quality geological map data with subsequent constraints
The Wyemandoo Shear Zone in the southeast of the imposed by geophysical data. The coarse resolution
complex displaces igneous layering and the SDZ. has allowed a geological focus on fundamental model
This shear zone juxtaposes lenses of metamorphosed geometries, which can be tested through forward
supracrustal rocks and amphibolitized gabbro lenses, geophysical modelling. One advantage of this approach is
with less sheared gabbroic rocks of the complex on either more flexibility, compared with an ‘over-constrained’ or
side. There is also another minor shear zone with similar highly complex model. The result of this methodology is
characteristics in the northwestern part of the complex (the to provide first order information on the likely form of the
Mingyngura Hill area Fig. 2; Ivanic, 2012a), which brings various components of the complex, which has enabled
supracrustal rocks into contact with the internal parts of some speculation as to the magmatic processes necessary
the complex. to explain this geometry.
Tight folding on a 500 m scale is evident in the Kantie
Murdana Volcanics Member and also in supracrustal rocks
immediately to the north of the complex. These locations Geophysical data
are both in the region of tightest curvature in the large-
scale layering of the WIC. Axial planes are typically steep Grids of magnetics and gravity were sourced from the
and northerly trending with an associated, moderately Western Australian state geophysical compilations gridded
north-plunging, lineation. This is consistent with an at ~400 m for gravity and 80 m for magnetic data. The
overall east–west shortening along a north–south axis and magnetic data is in units of nanotesla (nT) RTP (reduced
moderate downwarping along a north-plunging axis. If the to pole). The gravity data was converted from gravity
complex is taken as an originally circular form which has units (µms-2) to milligals (mgals). A first order trend was
become an ovoid strain ellipse, this would indicate ~2:1 removed from both gravity and aeromagnetic surveys to
east–west shortening. A synformal axis of hinge formation remove regional effects that could not be accounted for
is indicated on the Windimurra and Challa 1:100 000 map in the 3D model. In order to counter the effect of bodies
sheets (Ivanic, 2012a, b). around WIC, the forward modelling and inversion were
10
conducted on a smaller volume than for the whole 3D stacking velocity analysis. High bedrock and interval
model. The magnetic data was upward continued by velocities mostly exceeding 6250 ms -1 (and up to
1000 m to attenuate very short wavelengths that could not approximately 6500 ms -1 locally,) correspond with
be accommodated with the voxel element size used for the the location of the WIC (Jones et al., 2012). Seismic
inversion. Grids were then imported into the GeoModeller attenuation is low, resulting in high-frequency reflections
project. MT data was not utilized in this study. with good resolution down to the base of the complex,
thus facilitating detailed interpretation. Full details of
The region of outcrop of the WIC has a substantially
acquisition and processing are described in Costelloe and
higher gravity response than surrounding metagranitic
Jones, 2014.
rocks (Fig. 2) as would be expected for more
ferromagnesian plutonic rocks. Superimposed on this, Several workshops were held in 2013 in order to
distinct high-gravity features occur within the WIC, which make initial geological interpretations along all parts
are focused in its central parts but with a bias towards of the Youanmi seismic lines (Wyche et al., 2014; see
the southwestern sheared margin of the western lobe. Acknowledgements). Following this, we present a
Magnetic data provides excellent trend lines of igneous summary of the interpretation in the vicinity of the WIC
layering features (Fig. 2). In the lower zone olivine-rich which includes additional details. Once finalized, the
cumulates create thin linear trends of high magnetic interpreted cross-sections along the seismic lines were
response, interpreted to represent metamorphism to adapted and imported for use with 3D modelling software
magnetite + serpentine. In the middle and upper zones, the (Fig. 7 showing interpreted sections).
incoming primary magnetite translates to thick and high
magnitude magnetic highs. Faults may also be interpreted
based on the offsets of these features. Regions showing
distinctly low magnetic response in the south of the WIC GeoModeller
and in the northwest are thought to be caused by magnetic The software used to build the 3D model (Model 3)
remanence effects. was GeoModeller by Intrepid Geophysics (Melbourne).
GeoModeller is a 3D geological modelling tool used for
building complex, steady-state, 3D geology models and
Physical properties for performing forward and inverse geophysical modelling
directly from solid 3D geology. Later these shapes were
Rock property data (density and magnetic susceptibility)
translated to GOCAD for presentation purposes.
were compiled from several sources (Tables 1–3) for the
purpose of estimating average rock volume values for the The first consideration for the 3D model was the extent
various modelled packages: of the bounding box volume. We selected an area which
encompassed the YIC and the Narndee Igneous Complex
1. Values used for previous gravity modelling (Ahmat,
in case they were to be modelled at a later stage (Table 4).
1986; Table 1)
The depth was defined at 50 km (i.e. down to below the
2. Maximus Resources downhole geophysics data from Moho), as the Moho was well defined on the Youanmi
two of four drillholes from the Canegrass prospect seismic lines and the height above the surface, sufficient
(Fig. 2, Table 2) to model a substantial thickness of displaced stratigraphy.
3. GSWA samples analysed at the Australian National Second, we considered the resolution of observations to
University (Table 3) incorporate for the purpose of fulfilling the objectives
of this study. Hence, the resolution of one kilometre-
4. Values used in the gravity forward modelling of the scale was chosen (i.e. a discretization value of 1000 m),
seismic line (Gessner et al., 2013). so that we could model features on the order of the
scale of the gravity observations (2.5 km separation
The magnetic susceptibilities and densities of each between observations) and we would not over-define
package are shown in Table 1 with expected standard the model. The complexity of the brittle faulting (all
deviations shown for bimodal and unimodal distributions. estimated at <1 km displacement) could be avoided
Bimodal distributions were defined for units that are without significant disruption of gravity modelling.
expected to be distinctly heterogeneously layered in terms
of their density or magnetic susceptibility. The basal layer,
the mantle (unit 1, Table 1), is deemed not to have any Table 1. Summary of rock properties used in Ahmat (1986)
magnetic effect on aeromagnetic data.
Average density
Lithology
(g/cm3)
Seismic processing Generic ‘Windimurra Gabbroid’ 2.89
Details of seismic acquisition and processing for the Shephards Discordant Zone 3.10
WIC including specific observations on its seismic ‘Middle Zone – Upper Zone’ 3.15
characteristics, were presented in Jones et al., 2012. Felsic schists 2.70
Seismic processing obtains seismic velocity in two
Mafic intrusives, Kantie Murdana Volcanics 3.05
ways: bedrock velocity as a byproduct of refraction Member
statics analysis, and interval velocity as a derivative of
‘Granitoid’ 2.65
11
Ivanic and Brett
Table 2. Summary of downhole geophysics results from MNDD002 and MNDD004 form the Canegrass
prospect. Note that high average densities and standard deviations here are expected given the high
abundance of magnetite and fayalitic olivine in many of the intersected lithologies. Their interlayering
with anorthositic rocks is interpreted to have yielded high standard deviations.
Average magnetic
Standard Average Standard
Region susceptibility
deviation density (g/cm3) deviation
(S.I units)
Drillhole MNDD004 0.30 0.27 3.44 0.36
Drillhole MNDD002 0.50 0.28 4.13 0.54
Upper zone sheet (from MNDD002 <150m) 0.57 0.26 4.19 0.65
Middle zone average (from lower half 0.27 0.16 3.52 0.19
MNDD004)
Table 3. Summary of rock density determinations for GSWA samples
GSWA Density
Lithology Rock unit
sample no. (g/cm3)
178101 granitic rock Tuckanarra Suite 2.65
185922 gabbro Yalgowra Suite 3.06
178197 monzogranite Jungar Suite 2.64
198220 monzogranite Walganna Suite 2.65
198209 psammitic schist Norie Group 2.66
198210 foliated metagranite Jungar Suite 2.66
198132 sheared gabbro Unassigned amphibolite 2.92
In addition, the known complexity of the repetitive modal was considered too thin to include in this model;
igneous layering of the WIC could not be accurately (8) Proterozoic sills and dykes. These are too thin at
estimated far beneath the surface. Hence it would be <100 m and also likely have similar geophysical properties
averaged at this scale in terms of geometry and physical to the gabbroic rocks of the WIC; (9) the regolith
properties. (maximum thickness approximately 100 m); (10) the
topography (variation <100 m).
With the region and resolution defined in this way, the
appropriate means to divide the WIC into amenable rock A compilation of depth-converted seismic sections and
volume packages would be to subdivide the complex existing cross-sections from 1: 100 000-scale mapping was
into several components based broadly upon the zone imported and georeferenced in GeoModeller. The surface
nomenclature (above). External units to the WIC would topography was added as a flat horizon at 450 m elevation
also be readily modelled at this scale, i.e. the large-scale (‘Windi_SRTM_Elevation’). A simplified interpreted
crustal features such as granites, basement rocks and bedrock map was added to this topographic horizon with
associated major shear zones adjacent to the WIC. relevant structural data (‘rec2015_12_fig2_prelim’ in 3D
geomodel).
Features that were too small to model here are as
follows: (1) brittle faulting and smaller shear zone
geometry; (2) the border zone; (3) the Corner Well Table 4. Maximum and minimum values in the coordinate
Gabbro pipes; (4) separation of the roof zone gabbros system MGA zone 50 for the modelled volume in
from the felsic volcanics and banded iron-formation metres
of the Kantie Murdana Volcanics Member and Mount
Kenneth Suite granitic rocks; (5) individual megacyclic Minimum Maximum Length
units as even the largest of these are <500 m thick; Easting 555 000 705 000 150 km
(6) regions of the upper zone rich in magnetitite as
Northing 6 750 000 6 950 000 200 km
these are interlayered with a variety of lithologies
on a metre scale; (7) the Wyemandoo Shear Zone Depth -50 000 10 000 60 km
12
Construction of the 3D model 7. ‘EW1’ — A supplementary cross-section for extra

control on the model geometry at the southern tip of
Map data was digitized onto the surface section in the complex
GeoModeller according to the boundaries of the rock
packages defined. Structural data for igneous layering 8. Surface section (‘Windi_SRTM_Elevation’ in 3D
was added to the model where known, and was as widely geomodel)
distributed as possible. Faults were drawn and their
structural orientation data were added for these where
known. Following this, boundaries and orientation data Lithological packages and shear
were entered from the seismic lines and then from cross-
sections taken from the 1:100 000 maps. zones
This geological data were entered iteratively from simple, The simplified stratigraphy used is shown in Table 5 using
large-scale features initially, with more intricate detail ‘base of’ unit definition (i.e. not ‘top of’). Within the WIC
added later and as deemed necessary for geophysical the main components identified from map and seismic
modelling. Initially, all shear zones and faults were data were the hidden ultramafic zone, the lower zone of
ignored in order to assess the potential size of the complex the eastern lobe, the middle and upper zones of the eastern
prior to deformation (Model 1). Subsequently, the Challa lobe (grouped because the upper zone is locally <1 km
Shear Zone was incorporated along with shear zones thick), the lower and middle zones of the western lobe, the
around the northern part of the complex (Model 2). upper zone of the western lobe, and the roof zone merged
The smaller-scale Wyemandoo Shear Zone and north- with the Kantie Murdana Volcanics Member.
south-trending brittle faults were not modelled given the In the seismic lines, a significant distinction is apparent
relatively coarse resolution of this model. between the middle crust and the upper crust at the
The interpreted geometries from the three seismic lines 4–5s interval. Therefore the ‘Litho’ package (Table 5) is
10GA-YU1, 10GA-YU2 and 10GA-YU3 (see results equivalent to the ‘Yarraquin Seismic Province’ identified
section) were input into the model. A list of cross-sections in (Korsch et al., 2014) consisting of high-amplitude,
utilized in the 3D model is given below (see also Fig. 4): parallel, east-dipping reflectors. This package is deemed
to have a higher density than the upper crustal ‘basement
1. ‘YU1-curve’ — Southern 10 km of seismic line granites’ owing to higher pressure and higher Fe-Mg
10GA- YU1 content of less evolved rocks.
2. ‘YU2-curve’ — Western 30 km of 10GA-YU2 Table 6 shows the magnetic and density values used for
each unit. These are adapted from values in Tables 1–3.
3. ‘YU3-curve’ — Eastern 40 km of 10GA-YU3 The fault matrix for the three major shear zones in the
model is shown in Table 7 where the Challa Shear Zone
4. ‘COOLA1’ — 1:100 000 map sheet cross-section: is the latest feature, crosscutting the older shear zones
Coolamaninu E-W (the Yalgar and ‘eastern’ shear zones). The faults were
constrained to cut all stratigraphy except for ‘post-tectonic
5. ‘CHALLA1’ — 1:100 000 map sheet cross-sections:
granites’, ‘lithosphere’ and ‘lithospheric mantle’.
Challa (S) linked to Windimurra E-W
6. ‘CHALLA2’ — 1:100 000 map sheet cross-section:
Challa (N)
Table 5. Modelled stratigraphic column describing the features merged
Unit order and codename Unit name Comments

V01_BaldRockSupersuite Post-tectonic granites ≡ Bald Rock Supersuite
V02_KantieMV_RoofZone Roof zone and Kantie Murdana Each are <1 km thick on their own so have to be merged
Volcanics Member
V03_UpperMiddleZonesWest Upper and middle zones of the western Upper zone treated as a single sheet (single pulse; Nebel
lobe et al., 2013b) plus similar lithologies of the middle zone
V04_LowerZoneWest Lower and middle zones of the western Units immediately above the Shephards Discordant Zone
lobe
V05_UpperMiddleZonesEast Middle and upper zones of the eastern Grouped because the upper zone is locally <1 km thick
lobe
V06_LowerZoneEast Lower zone of the eastern lobe Large thick zone
V07_UltramaficZone Ultramafic zone Large thick zone
V08_Metagranites_Basement Pre- and syn-tectonic granitoids, upper Mainly Tuckanarra, Big Bell, and Mount Kenneth suites; may
crust basement include >2900 Ma basement
V09_YarraquinSeismicProvince Mid-lower crustal gneisses ≡ ‘Yarraquin Seismic Province’ identified in Korsch et al., 2014
LithoMantle Lithospheric mantle Tabular layer below the Moho
13
Ivanic and Brett
Results complex as a whole. The upper part of the complex, in

contrast, is strongly anisotropic in terms of the mineralogy
(and hence, density) of its lithologies and shows some
layering complexity visible especially in its central part
Interpretation of the seismic lines on 10GA-YU3 (Fig. 7c). This may be partly due to out-of-
across the WIC plane reflections caused by 3D effects, but also some local
complexities associated with brittle faulting.
The interpretations of the seismic data in the vicinity of
the WIC are presented in Figure 7 and described below. The maximum true thickness of the lower, strongly
The interpretations were checked for 3D consistency and layered region is 1 s TWT (about 3 km) and the maximum
boundaries were fixed on multiple seismic line junctions. thickness above this is 2.3 s TWT (about 6.9 km), both
evident on 10GA-YU3. Therefore a minimum estimate for
the maximum true thickness of the complex is 10.5 km. In
General seismic observations terms of stratigraphic thickness (that is, in a perpendicular
transect of continuous igneous stratigraphic layering
In the vicinity of the WIC, the Moho appears slightly
in 2D, where layering is not truncated) the maximum
elevated by 0.3 s two-way travel time (TWT, about 1 km),
thickness obtainable from seismic observations is
compared to the northern Youanmi Terrane. However, the
3.7 TWT (~11.5 km).
higher seismic velocities within the WIC are sufficient to
explain this difference and hence the Moho may in fact be The lowermost region is characterized by strongly layered,
flat. Apart from this, there are flat and relatively uniform high-frequency reflections, whereas, above this, it is
reflections below 9 s TWT (~27 km depth) within the seismically less reflective, with more discrete reflective
Moho transition zone. Between 3 and 9 s TWT (4 and horizons. Strong reflectors in the lowermost region are
9 s TWT on 10GA-YU3, about 12 and 27 km depth), likely produced by high impedance contrasts in cyclical,
there are a series of strong reflections dipping distinctly modally layered igneous rocks. High impedance contrasts
~15° to the east. This is interpreted to be the layered, are most likely due to peridotite-gabbro variation as these
intermediate middle crust with a strongly listric geometry have distinctly different densities and velocities. More
(i.e. the Yarraquin Seismic Province). There is a distinctive specifically, cumulates of peridotite, chromitite, pyroxenite
region of this middle crust as shallow as 2.4 s TWT (about and gabbro, are likely interlayered on a decametre scale
7.2 km) just to the north of the complex (seen in line at this level of a mafic–ultramafic intrusion. However,
10GA-YU1). This region deepens to a more typical level the spacing between major reflectors in this zone is on a
for the Youanmi Terrane of 4 s TWT (~12 km) towards the 300–500 m scale, which is indicative of the thickness of
central and southern parts of the complex (see Plate 3 in megacyclical units of an ‘ultramafic zone’. For the upper
Wyche et al., 2014). region, the lithology contrast between pyroxenitic and
anorthositic layers on a ~300 m vertical scale (as observed
In one seismic zone with no distinctive features and low at surface) is consistent with the abundance and magnitude
general reflectivity (e.g. 10GA-YU1 CDP15500–16000 at of these more discrete reflections, corresponding with
2–3s TWT), upper crustal granitic rocks are interpreted to lower zone rocks exposed at the surface.
overlie the strongly anisotropic middle crust and underlie
the WIC. Some of these granitic rocks are interpreted The regions of middle and upper zones on the seismic
to be in sheared contact with the ultramafic zone at the lines appear similar, owing to the presence of high-density
base of the complex, and attributed to the Tuckanarra and contrasts in the lithologies of these zones (i.e. anorthosite,
Big Bell suites, as these units can be traced to surface gabbro, and magnetitite, with the latter much more
outcrops. Some of these granitic rocks at depth, however, common in the upper zone). The upper zone appears to
may include older suites (<2813 Ma) which acted as be well constrained in the upper seismic region where
a basement to the complex. Some of these basement very flat-lying reflections in 10GA-YU2 and 10GA-YU3
granites may belong to the Mount Kenneth Suite which, are evident. This is consistent with the dip of layering
when exposed at the surface, are typically deformed and as observed at the surface. It is also consistent with the
interlayered with felsic supracrustal rocks of the Norie interpretation that this layer intruded as a crosscutting
Group. In addition, there are several instances where the sheet into at least the middle zone of the western lobe
seismic data show crosscutting granitic plutons of the Bald of the complex and possibly into its lower zone and
Rock Supersuite, some of which can be related to surface the various zones of the eastern lobe. This truncation is
exposures of the supersuite. evident in the 10GA-YU2 and 10GA-YU3 (Fig. 7b,c,e
and f).
Using information from all of the Youanmi seismic lines,
the general form of the WIC is a shallow, funnel-shaped Each seismic line shows a chaotic, poorly stratified region
cone, with inward-dipping reflections. The seismic corresponding with unit 9 (Table 5; interpreted to be
character is twofold with regions of very strongly layered the roof zone of the complex and the overlying Kantie
and reflective features, and regions of less strongly Murdana Volcanics Member), which crosscuts all other
reflective, but still layered, features. In each of the seismic zones of the complex, again consistent with surface
lines, these layered features in both seismic regions, observations. These contacts appear to dip at a shallow
typically follow the outer (funnel-shaped) form of the angle (<10°).
14
Table 6. Magnetic susceptibilities and densities for each unit showing standard deviations used for unimodal and bimodal
distributions. Parentheses figures are the percentages defined for the proportion of Mode 1 or Mode 2.
Magnetic susceptibility (SI units) Density (g/cm3)

Mode 1 Mode 1 Mode 2 Mode 2 Mode 1 Mode 1 Mode 2 Mode 2
Unit
mean st. dev. mean st. dev. mean st. dev. mean st. dev.
10 0.005 0.0005 2.60 0.4
9 0.08 0.008 3.00 0.3
8 0.12 (66) 0.04 0.06 (34) 0.004 3.10 (40) 0.2 2.80 (60) 0.2
7 0.03 (70) 0.006 0.06 (30) 0.0095 2.70 0.2
6 0.12 (30) 0.04 0.06 (70) 0.004 3.10 (50) 0.2 2.80 (50) 0.2
5 0.03 (70) 0.006 0.06 (30) 0.0095 2.80 0.2
4 0.04 (40) 0.01 0.06 (60) 0.02 3.00 0.3
3 0.01 0.002 2.65 0.1
2 0.01 0.001 2.67 0.15
1 0 0 3.00 0.2
Table 7. Modelled shear zones in the ‘fault matrix’
Stops on 
F01_ F03_ F02_
ChallaShearZone WindiEastShearZone’ YarlooShearZone
Fault name
F01_ChallaShearZone –
F03_WindiEastShearZone TRUE – TRUE
F02_YarlooShearZone TRUE –
Observations from 10GA-YU1 exposures of modally layered gabbroic cumulates are

known to have similar dips of 15–30° across this region,
(CDP 15300–16497) which are compatible with these reflections. Towards
As the line traverses the complex from the north, it the southern end of the line (closer to the centre of the
progresses up-stratigraphy through the lower zone and, at complex), these dips shallow to almost horizontal, as
the southernmost part of the line, it crosses into surface shown by the reflections in the lower zone towards the
exposures of the Kantie Murdana Volcanics Member (see southern limit of the line.
Fig. 7a,d). This line terminates at CDP5880 of 10GA-YU2.
At the northern contact of the WIC on this seismic line,
At CDP15300, the west-dipping Yarloo Shear Zone has there is an intersection of the complex with a highly
been inferred using aeromagnetic interpretation. Exposure reflective and gently north-dipping unit. Features such
is poor in the area, however, and so this feature is not well as this are extensive on this line further north and on
described. Granitic rocks in the vicinity have a steeply 10GA-YU2 and are interpreted as Proterozoic sills of the
dipping, northerly striking foliation, which strikes at a Warakurna Supersuite (Ivanic et al., 2014). It is inferred
small angle to the seismic line. Thus, they are expressed that this sill correlates with the Mount Holmes Gabbro
only as shallow-dipping reflections on this section. The sill which outcrops to the east of the WIC and is readily
shear zone is interpreted to be within metagranitic rocks visible in seismic line 10GA-YU2 (easternmost part of
of the c. 2700 Ma Big Bell Suite. Fig. 7b,e). In this instance, it is apparent that the high
impedance contrast of the sill within surrounding granitic
Felsic schists assigned to the Norie Group, Yaloginda
rocks transitions immediately once within the WIC into
Formation, including metasedimentary rocks (A-NOy-md
an impedance contrast of close to zero. Hence, it appears
on Fig. 7c), which mantle the WIC, are not reflective but
that the WIC crosscuts the younger sill, whereas the sill
are interpreted to occur in a small wedge immediately to
is effectively ‘seismically truncated’ rather than being
the north of the complex between shear zones. Rocks of
geologically truncated. This feature is considered to be
this unit are only exposed in a small outcrop (200 x 200 m)
below the resolution of the 3D model and is therefore not
north of the complex (Zibra et al., 2013).
included.
Relatively shallow-dipping reflections characterize
the lower zone of the complex. Corresponding surface
15
Ivanic and Brett
10GA-YU2 CDP 5880

a) 10GA-YU1
14000 14500 15000 15500 16000 16497 CDP
0 0
5
TWT (seconds)
Depth (km)
3
10
5 15
b) 10GA-YU2 10GA-YU3 CDP 9308 10GA-YU1 CDP 16497

4000 4500 5000 5500 6000 6500
0 0
WL
1
5
TWT (seconds)
Depth (km)
3
10
5 15
10GA-YU2 CDP 4422

c) 10GA-YU39000 9500 10000 10500 11000 11290
0 0
WL
1
5
TWT (seconds)
Depth (km)
3
10
5 15
TJI45_1 08.12.15
10 km
Fault
Reflector trend line
Geological boundary
Shepards Discordant
Zone
Figure 7. (a) Migrated seismic data from 10GA-YU1 showing interpreted structural features; (b) migrated seismic data from
10GA-YU2 showing interpreted structural features (note ‘WL’ = western lobe defined above the Shephards Discordant
Zone; (c) migrated seismic data from 10GA-YU3 showing interpreted structural features (note ‘WL’ = western lobe
defined above the SDZ). Interpretation of the seismic lines showing lithologies (GSWA codes from 1:100 000 maps)
for (d) 10GA-YU1; (e) 10GA-YU2 and (f) 10GA-YU3.
16
10GA-YU2 CDP 5880

d) 10GA-YU1
14000 14500 15000 15500 16000 16497 CDP
0 0
Ya r 5
TWT (seconds)
2 loo
Depth (km)
Shea
r
3 Zone
10
5 15
e) 10GA-YU2 10GA-YU3 CDP 9308 10GA-YU1 CDP 16497

4000 4500 5000 5500 6000 6500
0 0
ne 5
TWT (seconds)
2 Zo
Depth (km)
r
Cundimurra Sh
ear Zone o o Shea
3 Ya r l
10
5 15
10GA-YU2 CDP 4422

f) 10GA-YU3 9000 9500 10000 10500 11000 11290
0 0
oo
Ch
SZ
and
all
1 Ya
aS
rlo
em
Cu o
Z
SZ 5
TWT (seconds)
Wy
2 nd
Depth (km)
im
ur
ra
3 Sh
ea 10
r
Zo
4 ne
5 15
TJI45_2 08.12.15
10 km
Norie Group
Regolith
metasedimentary rocks
Warakurna Suite: Proterozoic roof zone
dykes and sills High reflectivity crust
Igneous Complex
upper zone
Bald Rock Supersuite monzogranites Low reflectivity crust
Windimurra
middle zone
Tuckanarra Suite metamonzogranites Fault
lower zone
Big Bell Suite metagranodiorites Reflector trend line
Kantie Murdana ultramafic zone
Volcanics Member Geological boundary
border zone Shephards Discordant
Zone
Figure 7. continued
17
Ivanic and Brett
GRAVITY MAGNETICS
a) b)
Observed
39 547
28 192
17 –163
5 –519
–5 –875
–16 –1229
mGals nT
c) d)
Forward modelled
48 814
31 584
14 353
–1 123
–18 –107
–35 –337
mGals nT
e) f)
Inversion misﬁt
17 205
10 –30
3 –266
–3 –503
–10 –739
–17 –975
mGals nT
TJI61 19.05.15
18
Figure 8. (left) a) Bouguer gravity anomaly (mgals) image of Observations from 10GA-YU3
the Windimurra Igneous Complex clipped to the area
relevant for the inversion volume; b) aeromagnetic (CDP 8600–11290)
(nT) image of the Windimurra Igneous Complex;
c) calculated gravity field (mgals) as a result of Seismic line 10GA-YU3 provides an east–west cross-
the forward modelling of 3D Model 3; d) calculated section across the complex from CDP 8800 to CDP
magnetic field (nT) generated as a result of the 10800 (see Fig. 7c,f). In the west, the line intersects the
forward modelling of 3D Model 3; e) gravity field sheared western contact of the complex along the Challa
misfit yielded by the inversion; f) magnetic field Shear Zone and continues through to CDP 9308, where
misfit yielded by the inversion. Note the probable it intersects 10GA-YU2. The line then crosses the SDZ,
regions of remanence located in the northwest
and the south-central parts of the complex. Also
2 km north of the Windimurra vanadium pit, and continues
note that the east–west trending Proterozoic dyke out past the eastern extent of the complex, across an
in the north of the complex is not accounted for unexposed contact with metasedimentary rocks within the
in the 3D modelling. Wyemandoo Shear Zone. The line finishes in post-tectonic
granitic rocks of the Walganna Suite.
The SDZ has an apparent dip which varies from close to
horizontal in the central part of the complex to moderately
east-dipping in the western part of the complex. At CDP
Observations from 10GA-YU2 10000, this zone is cut by a brittle fault, to the east of
which the zone dips to the west at approximately 30°
(CDP 4000–6900) apparent dip. Seismic trend lines interpreted within the
In map view, seismic line 10GA-YU2 starts in the centre complex have a shallow dip and form a broadly inward-
of the western lobe of the complex, proceeds towards the dipping structure, which is likely related to the primary
northwestern edge, and, as the line bends around to the igneous layering geometry.
east, traverses the middle of the lower zone, and proceeds
out to the eastern edge of the complex. At CDP 4422 the The ultramafic zone is interpreted to underlie the lower
line intersects 10GA-YU3 and at CDP 5880 it intersects zone and form a slight keel-shaped region under the centre
the end of 10GA-YU1 (Fig. 7b,e). of the complex down to approximately 3.3s TWT. This is
a region of apparently strongly layered ultramafic rocks
The dips in the ultramafic zone and lower zone support whose reflectors are partly continuous with other parts
an interpretation of an overall bowl-shaped vessel for of the interpreted ultramafic zone. However, immediately
the WIC, with steeper dips at the margins (c. 20°) and adjacent to this region at shallower depths, there are
subhorizontal reflections in the interior. Since the line lower reflectivity units present. These may be due to
approaches the margin of the complex in two places, this the presence of a region of strongly serpentinized (and
bowl shape is duplicated in this section. hence more isotropic) ultramafic rocks. Alternatively, the
lower reflectivity may be due to structural complexity
The ultramafic zone is imaged along the longest strike or granitic intrusions, all of which could account for
length of all the seismic lines and it is considered that the the discontinuous nature of this region. Our preferred
data from this seismic line is representative of the typical interpretation (Fig. 7f) is that younger granitic rocks
thickness of this zone at about 2.1 km. Therefore, in intruded at the base of the ultramafic zone; however, it is
contrast to 10GA-YU3 (next section), it is likely that this still considered a region of high uncertainty. In addition,
line does not image any of the thicker ‘root zone’ region. there is potential for this deepest part of the ultramafic
zone to be a remnant of a pipe- or funnel-shaped ‘root
The SDZ is interpreted as a south-dipping horizon towards zone’ (c.f. 10GA-YU2 where the ultramafic zone appears
the southern end of the line, cutting into lower-zone to have a relatively consistent thickness).
stratigraphy and overlain by middle- and upper-zone
material. Very shallow dipping (apparent dip of 0–5°)
reflections from CDP 4000 to CDP 4500 are interpreted Summary of seismic interpretation
to represent the magnetitite-rich base of the upper zone,
which can be seen to truncate moderately dipping middle Observations from the three seismic lines provided first
zone material (with an apparent dip of 5–15°). order constraints on the thicknesses of various features of
the WIC (Table 8).
A maximum thickness of 1 km in the line of section
is estimated for a shallow, bowl-shaped region above The following is a list summarizing the features
reflective, layered gabbroic rocks at approximately discovered from seismic interpretation:
CDP 5800. This is interpreted to be composed of 500 m
thickness of the roof zone of the WIC, and 500 m 1. The deepest part of the complex identified is in the
thickness of the overlying Kantie Murdana Volcanics central part of its intersection by the 10GA-YU3 seismic
Member. line, approximately 10.5 km below the present erosion
level. This estimate from seismic data is a minimum
The Proterozoic Mount Holmes Gabbro sill is interpreted estimate (utilizing a velocity of 6000 ms-1) since it is
at shallow levels to the east of the complex, and dips 5° likely that the velocity is >6250 ms-1 and hence it may
to the northeast, as noted in surface outcrops adjacent to be as deep as 11.5 km. This estimate might not represent
the seismic line. This feature is considered to be below the the typical thickness of the layered sequence as it may
resolution of the 3D model and is therefore not included. contain a remnant of a pipe-like ‘root zone’.
19
Ivanic and Brett
2. The typical thickness of the layered part of the 5. Strongly reflective and layered material underlies the
complex away from this potential root zone is entirety of the WIC. This is likely representative of the
estimated at 10.5 km. ultramafic zone, which cannot be traced to the surface
using the seismic data. Regions along the northern
3. Seismic data supports the hypothesis that the WIC had margin of the complex (e.g. on line 10GA-YU1),
an initial shallow conical form. The complex appears however, are identified as having the shallowest parts
relatively intact and coherent in its main part, with of this zone at about 600 m depth. The western margin
local modification by strike-slip shear zones, which also has a region of shallow (<1 km depth) ultramafic
have likely steepened layering, at the margins of the zone material identified at CDP 8800 on 10GA-YU3.
complex. In addition, brittle faulting and crosscutting
granitic plutons have affected the complex. 6. The Challa Shear Zone dips approximately 70° to the
east-northeast. This dip decreases below 3 km depth to
4. A two-stage stratigraphy consisting of an eastern approximately 45°.
and western lobe is permissible given the seismic
data; however, it is accepted that it is not clear. Here, 7. The interpretation of the SDZ as marking the base of
we consider the eastern lobe to represent the initial the western lobe of the complex remains plausible as
intrusive form, evolving from an ultramafic zone it is consistent with seismic data.
through lower, middle and upper zones. A second 8. An intrusion of a flat slab of ~1 km-thick upper
lobe (i.e. the western lobe) consistently overlies the zone material in the western lobe (in the vicinity of
eastern lobe when its zones are extrapolated to depth the Canegrass Ti–V–Fe prospect) is consistent with
along reflector trend lines. In general, the reflectors seismic data in 10GA-YU2 and 10GA-YU3.
are subparallel to the underlying eastern lobe. At the
crucial part of 10GA-YU3 (CDP 9500 and 1s TWT) 9. A strongly reflective, west-dipping slab (c. 30°)
it appears that there is a faulted contact, hence no of upper zone material (including the unit at the
primary truncation of eastern lobe by western lobe is Windimurra vanadium pit) in the eastern lobe is
observed in seismic data even though it is apparent at apparent from seismic data in 10GA-YU3, down to at
the surface. least 3 km depth.
Table 8. Summary of thickness constraints for the WIC, derived from observations on the Youanmi seismic
lines. SDZ = Shephards Discordant Zone, KMVM = Kantie Murdana Volcanics Member. Results
are conservative estimates since an average seismic velocity of 6000 ms-1 was used for Archean
crust in the published seismic cross-sections. Jones et al. (2012) explain that in the vicinity of the
WIC, the velocity is likely closer to 6500 ms-1, yields thicknesses 8.3% higher (i.e. those shown
in the final column estimates). Estimates are not made for features that are too thin such as the
border zone or the SDZ. Particular estimates are not made for features with minimal information
such as the ultramafic and lower zones of the western lobe. Note: zones here are fully subdivided
and do not directly match the units in the 3D geomodel.
Estimated
Max. vertical Max. cumulative Maximum typical
Zone thickness perpendicular depth-to- Primary source thickness (km)
(km) thickness (km) base (km) where
v = 6.5 kms-1
KMVM 0.5 0.5 0.7 10GA-YU2 0.5
Roof 0.3 0.4 0.8 10GA-YU2 0.3
Western lobe 4.1 5.2 4.1 10GA-YU3 -
Upper 0.9 1.0 1.4 10GA-YU3 1.0
Western lobe
Middle 2.3 2.8 3.1 10GA-YU2 2.2

Lower 1.8 2.0 4.0 10GA-YU3 -
Ultramafic - - - - -
SDZ - - 3.8 10GA-YU3 -
Eastern lobe 7.9 9.7 10.5 various 9.5
Eastern lobe
Upper 0.7 0.8 2.6 10GA-YU3 0.6

Middle 0.9 1.0 3.0 10GA-YU3 0.9
Lower 6.1 6.3 6.2 10GA-YU2 5.7
Ultramafic 3.2 3.2 10.5 10GA-YU3 2.3
Border zone - - - - -
WIC as a whole 10.1 13.7 10.5 10GA-YU3 11.4
20
10. The Kantie Murdana Volcanics Member is just Model 2 — addition of shear zones
distinguishable from underlying units and appears to
overlie up to 1 km thickness of gabbroic rocks of the The discrepancies listed above were readily resolved in
roof zone of the complex (as visible in high-resolution Model 2, first by the introduction of the Challa Shear
seismic data on 10GA-YU2 presented in Wyche et al., Zone at the western margin of outcrop of the complex.
2014). Hence, both of these units overlie and truncate This fault was allowed to displace the complex entirely
the seismically layered rocks of the eastern lobe of the above surface to the west of the fault. The result of this
complex. was that a large area of high gravity to the west of the
fault (the Muleryon Hill area) was occupied by granitic
rocks in the model, therefore changes were made so
that the ultramafic zone was not entirely displaced and
Results from 3D modelling hence allowed to outcrop under the Challa salt lake area
(including the known exposures of ultramafic rocks
of Muleryon Hill). Second, shear zones along the east
Initial results and refinement (unnamed shear zone) and northwest (Yarloo Shear Zone)
of the complex were introduced using surface constraints
Model 1 — simple geology model on their orientation in order to remove outcropping
ultramafic zone rocks along these marginal areas. Third,
An initial model was created which prioritized structural a ‘post-tectonic granite’ (Bald Rock Supersuite) was
data from the well-preserved (i.e. away from shear zones) introduced at the southern margin in order to truncate the
parts of the WIC with primary igneous layering. Large- zones of the complex as shown in Fig. 2. The resultant
scale geometric relations in seismic data from the various 3D map produced was Model 2 which accounted for the
zones of the complex were also incorporated, and it was internal and external form of the complex given surface
possible to create a model which extrapolates these data and seismic constraints (resultant volumes shown in Table
in 3D without the introduction of structural complexity. 9). Forward modelling (see later) suggested modifications
Hence, results of a simplified geological 3D model prior were required to account for the geometry of the central
to the introduction of faults yielded Model 1, which part of the complex
fitted selected and well-constrained map and seismic
data in the core of the complex only. In the periphery
of the complex there were discrepancies relative to Forward modelling and the final 3D
surface mapping constraints and volume estimates from
the subsequent 3D models as follows: (1) The western model
lobe of the complex (i.e. that which lies above the SDZ)
extends ~30 km further west of the Challa Shear Zone Model 3 — forward modelled
and occupies approximately three times the volume
indicated from successive models (Table 9); (2) The A smaller volume within the GeoModeller project was
lower zone extends west and occupies approximately 2.5 used for forward modelling and inversion (Figs 3 and
times the volume of successive models. Similarly, the 8a,b; Table 10). This enabled focus on the WIC and
ultramafic zone is approximately 1.6 times the volume avoidance of high-density and high-magnetic regions in
and extends further west; (3) The ultramafic zone projects the vicinity of Mount Magnet and Youanmi townsites that
to the surface with an outcrop approximately 4 km wide were not modelled in 3D. It also reduced the time taken
concentrically around the eastern lobe of the complex; for each forward model/inversion run. The approach
(4) The total volume modelled for the WIC without adopted here was to use available geometric and physical
faults is approximately double that of successive models, property constraints (as previously described) to produce
indicating that the initial volume of the complex may an initial forward model. Minor adjustments to the
have been in the vicinity of 26 000 km3. Admittedly, this physical properties were required for poorly constrained
volume estimate is quite poorly constrained; however, we or unknown units such as the ultramafic zone and
consider it the best estimate possible with the present 3D basement granites in order to yield a first order fit for
data. It also represents an estimate of the original shape of magnetic and gravity survey data. Following this, better
the intrusion. fits were achieved by modifying the model geometry.
Table 9. The volumes of the various modelled units in Table 10. Maximum and minimum values
Models 1 and 2. Note: only Model 3 is shown in the in the coordinate system MGA
3D geomodel zone 50 for the modelled volume
Model 1 Model 2 in metres
Unit
volumes (km3) volume (km3)
S02_KantieMV_RoofZone 63 Minimum Maximum
S03_UpperMiddleZonesWest 357
X 605 000 665 000
S04_ShephardsDiscordantZone’ 4 050 2 360
S05_UpperMiddleZonesEast 11 400 740 Y 6 822 000 6 934 000
S06_LowerZoneEast’ 6 252
Z -20 000 450
S07_Ultramafic 10 800 3 803
TOTAL 26 250 13 155
21
Ivanic and Brett
This was changed in simple ways and in selected This is the result of one major change, that basement
regions with higher degrees of freedom (i.e. away from granitic rocks of low density were required to replace
known geometric constraints from surface or seismic a significant proportion of the ultramafic zone of the
observations). The result of this iterative process was to complex. This is relatively straightforward to reconcile
achieve a reasonable forward-modelled fit to potential with seismic observations in that granitic sills parallel
field data given the resolution and other limitations of the to layering of the ultramafic zone would be consistent
3D model. with the strongly layered seismic character of this zone.
There was also a shift to reduce the proportion of other
The forward model result of Model 2 at 2 x 2 x 1 km high-density lithologies of the middle and upper zones of
resolution produced a sharp high-density area in the both the eastern and western lobes, which may be due to
central region of the complex whereby the upper zones overestimates of their densities.
for both lobes of the complex overlay each other. The
geometry of the eastern lobe was therefore simplified at
depth so that middle and upper zones were more gradually
Table 11. The volumes of the various modelled units in
curved. This resulted in a geometry whereby these zones Model 3
did not underlie the Kantie Murdana Volcanics Member
to any significant extent. The western lobe also attained Model 3 Model 3 –
Unit
a simpler geometry with a slight reduction in its volume. volume (km3) Model 2 (%)
The modifications were in a region that was relatively V01_BaldRockSupersuite 1 258
unconstrained by seismic data, which also had poor V02_KantieMV_RoofZone 108
exposure; therefore, it was deemed a feasible adjustment
V03_UpperMiddleZonesWest 386 +7%
to make.
V04_LowerZoneWest 2 170 -9%
Model 3 resulted from the adjustments described above V05_UpperMiddleZonesEast 589 -26%
and minor modifications in the thickness of various units
V06_LowerZoneEast 6 415 +3%
in order to produce a good fit to potential field data. The
results are shown in Figure 8c and d, Table 11, indicating V07_UltramaficZone 4 162 +9%
that the large reduction in middle and upper zones of V08_Metagranites_Basement 39 509
the eastern lobe (i.e. ‘MiddleZone’) was accommodated V09_YarraquinSeismicProvince 82 827
primarily by increased lower zone material. This fit
(i.e. relatively low degree of mismatch) was deemed
sufficient given the inherent limitations of the 3D model
in general, especially in terms of resolution (as discussed Table 12. Inversion parameters for runs 1–4
earlier).
P(lith
Inversion Voxel size P(voxel in
Table 11 shows the final volumes for Model 3 and that the run
Voxels #
(km)
change)
model)%
lower zone of the eastern lobe accounts for approximately %
half the volume of the entire complex. A conservative 1 30 x 56 x 20 2x2x1 50
Uncon-
estimate for the volume of the WIC as a whole is therefore strained
c. 13 000 km3. 2 30 x 56 x 20 2x2x1 0 100
3 60 x 112 x 41 1 x 1 x 0.5 25 84
4 60 x 112 x 41 1 x 1 x 0.5 50 70
Inversion
In order to refine and test the implications of Model 3,
a geophysical inversion was run to provide a statistical
Table 13. The volumes of the various modelled units in
appraisal of the success or failure of aspects of the implied
Model 3 compared to inversion results
geometry. 3D GeoModeller uses a stochastic inversion
process to alter either the modelled lithological boundaries Model 3 Model 3
Model 3
or properties within specified constraints with a set 15% inversion
inverted inverted
chance of a property or geometry change. This process Unit – Model – Model 3
volume
3 initial initial (vol.
requires a starting model (Model 3) which can closely (km3)
(km3) % change)
reproduce the observed geophysical responses. Joint V01_BaldRockSupersuite 1 027 -131 -10%
inversions were run on Model 3 utilizing detrended gravity
V02_KantieMV_RoofZone 92 -16 -15%
and aeromagnetic grids using the parameters outlined
in Table 12. Voxel sizes of between 2 and 0.5 km were V03_UpperMiddleZonesWest 335 -50 -13%
used as the resolution of the model was 1 km; however, it V04_LowerZoneWest 2 310 +143 +6%
was found that more accurate results in the region of the V05_UpperMiddleZonesEast 178 -411 -70%
relatively thin Kantie Murdana Volcanics Member required V06_LowerZoneEast 6 700 -285 -4%
a smaller voxel size of 0.5 km.
V07_UltramaficZone 1 624 -2538 -60%
Table 13 shows that some significant lithological voxel V08_Metagranites_Basement 44 773 5264 +13%
changes (i.e. substitution of lithologies of different V09_YarraquinSeismicProvince 80 382 -2445 -3%
densities) were required by the final inversion run.
22
Discussion an estimate for the brittle ductile transition at 2810 Ma of

about 11 km. This estimate is probably reasonable, since
the elevation of the upper contact is also constrained by
the presence of the Kantie Murdana Volcanics Member
Implications of the first order data above the central parts of the complex.
and constraints
Discordance
Thickness estimates The subdivision of the WIC into an eastern and western
The stratigraphic thickness estimated for the combined lobe along the SDZ is consistent with surface observations
eastern and western lobes of the WIC is approximately of magmatic discordance between the two lobes (Ahmat,
11 km, which is thicker than that estimated for the 1986). The lack of deformation features in this zone, in
8 km- thick Rustenberg Layered Series of the Bushveld addition, strongly argues for a nondeformational origin.
Igneous Complex (Cawthorn and Walraven, 1998), The interpretation presented here of the SDZ extending
previously the thickest such sequence of layered gabbroic across to the west at depth and entirely subdividing the
rocks identified. This 11 km estimate may be too high complex is also consistent with seismic interpretations
for several reasons. (1) If the eastern and western lobes and 3D constraints. The implication of this is that, rather
are considered separate but overlapping intrusions, then than being an isolated region of discordance of a structural
11 km represents a combined thickness. However, the or magmatic nature (Ahmat, 1986), it is implicitly a large
eastern lobe would still show a vertical thickness of and significant subdivision of the complex, reflecting the
between 9 and 10 km, which is still extremely thick. The later timing (within the error of geochronological data)
revised stratigraphic column is presented in Figure 9 and crosscutting nature of the western lobe.
and is the result, primarily of seismic data (Ivanic et al., In this interpretation, the SDZ is positioned whereby it
2014); (2) Structural repetition may cause some enhanced represents the uppermost part of the eastern lobe in 3D.
thickening of the central parts of the complex. Very small- Along its eastern contact, there is little or no discordance
scale reverse faults have been noted in the Windimurra with igneous stratigraphy of the eastern lobe, whereas on
vanadium pit in the eastern lobe of the complex; however, the western margins of the SDZ, as one follows from the
these are not thought to be responsible for large-scale southern extent of the complex to the north, a traverse is
duplication and the thickening associated with them is made up the igneous stratigraphy of the western lobe from
thought to be minor; (3) A downwarping of the entire the central part of the lower zone to the upper part of the
complex could artificially enhance the thickness in the middle zone (Fig. 2; i.e. several kilometres of discordance
centre of the complex. Some evidence of this has been compared to the eastern lobe). Then to the northwest,
noted — for example, a synformal axial trace in the the SDZ is itself truncated by the roof zone and Kantie
northern part of the eastern lobe (Ivanic, 2012b) but this Murdana Volcanics Member. Finally, to the west, it re-
is unlikely to contribute more than a few hundred metres emerges at the surface and represents the lowermost upper
of enhanced thickness; (4) It is possible that some of the zone of the western lobe, where it directly overlies the
highly reflective material at the base of the ultramafic zone lower zone of the eastern lobe. This geometry runs into
in 10GA-YU3 is part of a feeder funnel at the base of the the crosscutting Challa Shear Zone at the western margin
complex. Inclusion of this in the total thickness is deemed of the WIC. At depth, these constraints are maintained
reasonable as there are contiguous layered rocks within underneath the entire western lobe where it is underlain
and above this region; (5) Minor addition of gabbroic by variable thicknesses of ultramafic and lower zone
material into the layered sequence by sills assigned to the sequences of the eastern lobe.
Warakurna Supersuite may have contributed up to 100 m
in thickness locally. Therefore, even allowing for these Small-scale discordance as observed in 10 m to 1 km
points, the thickness of the complex as a whole is still features in the field (Ahmat, 1986; Ivanic et al., 2014)
likely to be in the region of 11 km. Furthermore, several were not considered in this interpretation and on a macro-
factors may be reducing this estimate of the thickness such scale, the eastern lobe is not considered to show the type
as normal faulting, erosion from the top and intrusion of of discordances found in the western lobe. Progressing
granites from beneath. up-stratigraphy into the western lobe, its middle zone as a
whole shows a transgressive discordant geometry, whereby
This vast thickness suggests that there may have been at the surface it cuts across and entirely truncates the lower
something fundamentally different about the WIC, zone of the western lobe to the east (complete truncation
compared to proximal Archean intrusions and subsequent occurs about 5 km south of the Windimurra vanadium pit).
examples of Proterozoic–Phanerozoic intrusions. We
speculate that it was the superimposed, multipulsed nature The crosscutting nature of the upper zone of the western
of the developing cumulate sequence that allowed for a lobe (Nebel et al., 2013b) is exemplified in the seismic
dominance of vertical accumulation over lateral spread. and 3D interpretation presented here (Figs 7, 9 and 10;
Appendix 3D model). This zone is interpreted to cut down
Thermomechanical modelling of the dynamics of magma into all underlying igneous stratigraphic units, for example
emplacement suggest a transition from vertical to into the lower zone of the eastern lobe (on 10GA-YU2 at
horizontal flow for large intrusions at the brittle–ductile CDP 4570; Fig. 7e). At the surface, the middle zone of
transition (Brown, 2007 and references therein). Therefore, the western lobe is entirely overstepped by the upper zone
the depth of the base of the complex potentially provides (Fig. 2) both at its northern and southern limits.
23
Ivanic and Brett
observed
zone
Ultramafic zone
minerals
The two-fold subdivision of the complex brings into
magnetite
question the presence of an ultramafic zone for the western
ap./ilm.
olivine
western eastern lobe, which has not previously been defined. In the 3D
lobe lobe model the ultramafic zone is not separated into eastern and
11 western portions and is interpreted to underlie the entire
complex. One possibility is that the ultramafic zone of the
upper
zone
western lobe is distinct from that of the eastern lobe and
underlies the southwestern part of the complex, where it
10
is not exposed and has not been imaged by the southerly
extensions of the seismic lines. In general, the details
middle zone
of the form of the western lobe still remain relatively

9 unconstrained in 3D owing first to the lobe’s significant
truncation by the Challa Shear Zone to the west; hence,
this lobe’s ultramafic zone may have been removed.
Second, there is a lack of seismic data further south and
8
third, there is a lack of rock exposure especially around the
Challa Salt Lake system.
7
A significant ultramafic zone is likely to underlie the entire
complex with a thickness of 1–2 km and possibly as much
as 3 km. The presence of this basal component of the WIC
chromite
Vertical thickness (km)
effectively accounts for the ‘missing ultramafic zone’

6 (Ahmat, 1986; Ivanic et al., 2010), and therefore mitigates
against the possibility (Bunting, 2004) that the Narndee
lower zone
Igneous Complex represents the ultramafic portion of

5
the WIC. The bulk composition of the WIC is, therefore,
much less Ca- and Al- rich than previously anticipated
??
(Ahmat, 1986) and closer to a tholeiitic/komatiitic basalt
composition.
4
There also remains the possibility that remnants of the
original feeder system exist underneath the central part
lithologies inferred from seismic data
of the complex as highlighted in the 10GA-YU3 seismic

3 line in 2D. This region is likely intruded by granitic rocks,
rendering its precise form untraceable in 3D. Reflectors
in 10GA-YU3 do, however, hint at a funnel-shaped form
which may well become more significant out of the plane
2
of this line towards the south.
ultramafic zone
0
BIF layers
Kantie Murdanna Volcanics Gabbronorite
Member: rhyolitic volcanics
Roof zone porphyritic gabbro Norite-olivine gabbronorite Figure 9. (left) Revised igneous stratigraphic column for the
Predominantly leucocratic Pyroxenite Windimurra Igneous Complex (see Fig. 6), showing
rocks
true thickness for all major components of the
Border zone Troctcolite
amphibolitized dolerite complex as derived from observations made at
Western lobe upper zone
Wehrlite the surface and from 3D modelling. The Complex
sheet
is divided into an eastern and western lobe along
Magnetitite interlayered with Peridotite including dunite-
leucogabbro harzburgite-lherzolite the Shephards Discordant Zone (dashed correlation
Anorthosite
lines, indicating an offset in stratigraphy). The
Magnetitite layers
ultramafic zone is interpreted from seismic data
Gabbro
(Corner Well Gabbro) Chromitite layers alone. Note also the major oxide horizons of
TJI41a 21.08.15
chromite-rich and magnetite-rich rock types.
24
Implications of the inversion These magnetitites (likely still hot as no clear chilled
margin exists) may have acted as an impenetrable barrier
results to this incursion; however, these appear cut across in the
The results of the geophysical inversion yielded some north of the complex; Stage 5: fractional crystallization
implications, which are primarily that the large-scale and continued pulsed inflation of the western lobe leading
geometry, in the upper parts of the complex at least, to emplacement of its middle zone with an erosive base;
is consistent with geophysical data. However, specific (Fig. 10c; Stages 6 and 8) Stage 6: emplacement of
excursions from the initial geometry of the 3D model were the upper zone of the western lobe as a single pulse of
implied by the geophysical inversion (Table 13), which extremely Fe-enriched magma with extreme discordance
resulted in the replacement of high-density lithologies to all underlying units. Northwest–southeast trending
with lower density ones. There may be several reasons density currents likely created the channel features of
for this tendency. First, the densities of the ultramafic thicker magnetitites present in the Canegrass region.
and magnetite-rich rocks may have been overestimated. These apparently scoured deeper into the middle zone in
For example, as the ultramafic zone has not been directly this vicinity probably owing to their more central position
sampled, it may contain a significant proportion of as a result of down-dip density currents sourced from
pyroxenite or gabbroic rocks, reducing its average density. the southeast; Stage 7 (not shown on figure): intrusion
Likewise, the upper zone of the western lobe may have been of the Corner Well Gabbro pipes probably sourced from
assigned higher density and magnetic susceptibility values a final phase of low-volume primitive melt underlying
owing to utility of diamond drillhole data, which primarily the complex; Stage 8: extrusion of the overlying Kantie
samples the magnetite-rich lower part of the upper zone. Murdana Volcanics Member and emplacement of the roof
zone gabbroic plutons, which intrude and chill at the base
Second, there may be a significant intrusion of granitic of the volcanics; (Fig 10d; Stage 9) Stage 9: post 2813
rocks into the ultramafic zone as indicated by the inversion Ma events. These are dominantly granitic intrusions;
voxels in this region. These granitic rocks may belong to deformation along the Challa Shear Zone and Wyemandoo
the Big Bell or Tuckanarra Suites, or alternatively they Shear Zone; brittle internal deformation of the WIC;
may be an extension of the Walganna Suite rocks inferred Proterozoic dykes and sills (not shown on figure); erosion
to intrude at the base of the complex in 10GA-YU3 to present-day level.
(Fig. 7f). In addition, significant serpentinization may
also account for a density reduction of the ultramafic zone With this genetic model as a base, future work may be able
especially in proximity to major shear zones. to test the validity of several of these proposed stages for
the evolution of the WIC. Given that this is, in the most
Third, there may be a geometric inaccuracy in the model; part, a deformed and metamorphosed intrusion, future
however, this is unlikely in the upper zone as it is well work will require more detailed geophysical analysis and
constrained by surface and seismic observations. Lacking further isotopic work in order to document the nuances
further constraints, the form of the ultramafic zone may of the finer scaled layering geometry and timing of
be more irregular than that estimated from seismic data emplacement for the various units.
alone and it could possess a more restricted funnel region
compared with the 3D model.
Conclusions
Our assessment of these implications is that we have likely
overestimated the density of the upper zones of the eastern A preliminary 3D model for the complex was
and western lobes. In addition, we infer that the density generated using GeoModeller, and incorporated
of the ultramafic zone may also have been overestimated; map data and geophysical survey data as well as
however, it is likely that our estimation of the volume of preliminary interpretation of the three seismic lines
granitic intrusions into this zone was underestimated. (see data: 3D geomodel). The model yielded a 3D form
for the zones of the intrusion that is consistent with
surface and seismic data. The model confirmed the broad
Evidence for a genetic model funnel-shaped form and the inward-dipping nature of the
From the geometric constraints described above, the main subdivisions of the complex.
overall emplacement sequence is envisaged as follows
(Fig. 10a; Stages 1–2): Stage 1: incipient emplacement 1. The WIC is the world’s thickest mafic–ultramafic
including formation of chilled margins represented layered intrusion reported at 10.5 km. This estimate
by the border zone dolerites and potentially a ‘root from seismic data is a minimum estimate (utilizing a
zone’ as a funnel-shaped feeder system at depths in velocity of 6000 ms-1) since it is likely that the velocity
excess of 10 km below the present-day erosion level; is >6250 ms-1 and hence it may be as thick as 11.5 km.
Stage 2: emplacement of the layered sequence
2. A two-fold stratigraphy for the complex with an
of ultramafic, lower, middle and upper zones of the
eastern and western lobe is required to accommodate
bowl-shaped eastern lobe, a multipulsed fractional
3D constraints presented here. The eastern lobe
crystallization sequence; (Fig. 10b; Stages 3–5)
represents the initial intrusive form evolving from an
Stage 3: west-down tilting of the eastern lobe, possibly
ultramafic zone through the lower, middle and upper
~10°; Stage 4: emplacement of the magma of the western
zone. Subsequently, the western lobe intruded in the
lobe. Initially an ultramafic postulated to have developed
south west, and cut across the lower, middle and upper
at depth with the lower zone up against magnetitites
zones of the eastern lobe.
of the upper zone of the eastern lobe along the SDZ.
25
Ivanic and Brett
SW NE 3. 3D inversions show that a significant ultramafic zone,

a) approximately 3 km thick at its maximum, likely
uplift, doming and erosion of 2819 Norie Group basaltic flows
3 km
underlies the entire complex. The bulk composition is,
0 km therefore, much less Ca- and Al- rich than previously
anticipated and closer to a tholeiitic/komatiitic basalt
composition. Contiguous parts of the ultramafic zone
granitic upper crust
20 km
of the complex are modelled to be within 500 m of the
brittle-ductile transition surface along its northwestern margins.
–12 km
gneissic middle crust
4. Possibly as much as half of the original western lobe of
feeder system – multiple the complex was faulted off to the west by the Challa
pulse inflation
Shear Zone according to our Model 1. The original
fractionating volume of the complex as a whole is estimated at
t1 = 2813 Ma staging chamber
about 26 000 km3. The location of this missing volume
b) is unknown, but likely eroded.
chamber disruption ± shift of magmatism ± tilting
3 km
5. The ultramafic zone likely extended to the surface
0 km around most of the complex prior to deformation. It is
* also likely to occur as a lens along a north northwest-
trending, 20 km strike length of the Challa Shear Zone,
resulting in the extended gravity high west of the main
–12 km outcrop of the complex. Therefore, the rocks of the
Muleryon Hill area are thought to belong to the WIC
proto-caldera formation rather than the Narndee Igneous Complex.
t2 6. Identification of the thick, subsurface ultramafic
~ t1 + 10 ky staging chamber collapse,
shift and replenishment zone underlying the lower zone of the eastern lobe
of the complex is a new target for Ni–Cr–PGE
c) upper zone emplacement with erosional base mineralization.
3 km
7. Extension of the SDZ to the west of the complex
0 km
provides a 3D geometry consistent with geophysical
* data. This zone is cut by the upper zone of the western
future site of lobe in the north. It is also truncated in the north by the
Wyemandoo roof zone and the overlying Kantie Murdana Volcanics
Shear Zone
future site of
–12 km
Challa Shear Zone Member.
upper zone intruded as megapulse

8. A modelled shallow-dipping, low-curvature slab of
1 km thick upper zone material from the western lobe
t3
~ t1 +100 ky staging chamber fractionated in the vicinity of the Canegrass Ti–V–Fe prospect is
to Fe-rich compositions consistent with geophysical data.
d)
3 km
Current 9. In the vicinity of the Windimurra vanadium pit, a
erosion
surface
modelled moderately west-dipping slab of upper
0 km zone material from the eastern lobe is consistent with
geophysical data.
10. The Kantie Murdana Volcanics Member is modelled at
0.5 km thick, overlying up to 1 km thickness of relatively
–12 km Strike-slip
shearing at margins dense gabbroic rocks of the roof zone of the complex.
Granitic intrusions
t4
= t1 + 100 Myr (2700-2660 Ma / cratonization by 2600 Ma)
Figure 10. (left) Schematic genetic model for the formation
faults upper zone of the Windimurra Igneous Complex (see text for
shear zones discussion). a) Initial stage of intrusion into the
middle zone
late granitic suites Norie Group (t1), formation of the eastern lobe;
lower zone
Kantie Murdanna b) formation of the western lobe (t2); c) formation of
Volcanics Member the upper zone of the eastern lobe, Mount Kenneth
*
ultramafic zone
Mount Kenneth Suite and Kantie Murdana Volcanics Member (t3);
Suite tonalites border zone
d) formation of granitic rocks of the Big Bell Suite,
roof contact zone
magmatic feeder Tuckanarra Suite and Bald Rock Supersuite and
erosive turbidity currents
at base of upper zone of deformation and erosional features (t4). White
western lobe staging chamber
Shephards Discordant
asterisks in b) and c) denote possible site of
Norie Group greenstone
Zone host rock ultramafic zone of western lobe now removed by
TJI43a 02.11.15 shearing.
26
Acknowledgements Ivanic, TJ, Korsch, RJ, Wyche, S, Jones, LEA, Zibra, I, Blewett, RS,
Jones, T, Milligan, P, Costelloe, RD, Van Kranendonk, MJ,
Doublier, MP, Hall, CE, Romano, SS, Pawley, MJ, Gessner, K,
The Youanmi seismic interpretation workshops held in
Patison, N, Kennet, BLN and Chen, SF 2014, Preliminary
2013 involved diverse discussions concerning many facets interpretation of the 2010 Youanmi deep seismic reflection lines
of the geophysics and geology in the vicinity of the WIC. and magnetotelluric data for the Windimurra Igneous Complex,
Leonie Jones provided seismic processing refinements in Youanmi and southern Carnarvon seismic and magnetotelluric
in the vicinity of the WIC. The following are thanked in (MT) workshop 2013 compiled by S Wyche, TJ Ivanic and I Zibra:
particular for contributions that helped shape our views Geological Survey of Western Australia, Perth, Record 2013/6,
presented here: Richard Blewett, Stephen Wyche, Russell p. 97–111.
Korsch and Brian Kennett. In addition, John Bunting Ivanic, TJ, Wingate, MTD, Kirkland, CL, Van Kranendonk, MJ and
and Graham Kennedy are thanked for useful discussions Wyche, S 2010, Age and significance of voluminous mafic–ultramafic
on the formation of the Windimurra Igneous Complex. magmatic events in the Murchison Domain, Yilgarn Craton:
Craig O’Neil (Macquarie University) is acknowledged for Australian Journal of Earth Sciences, v. 57, p. 597–614.
analysis of rock properties shown in Table 3. Jones, LEA, Ivanic, TJ and Costelloe, RD 2012, Seismic reflection
imaging of the mafic–ultramafic Windimurra Igneous Complex,
This manuscript has benefited from helpful discussions Yilgarn Craton, Western Australia, in Extended abstracts volume:
with Klaus Gessner and David Howard, and thorough 22nd Australian Society of Exploration Geophysicists international
reviews from Heather Howard and Ruth Murdie at the conference and exhibition, Brisbane, Australia, 26 February 2012,
Geological Survey of Western Australia. p. 1–4.
Korsch, RJ, Blewett, RS, Wyche, S, Zibra, I, Ivanic, TJ, Doublier, MJ,
Romano, SS, Pawley, MJ, Johnson, SP, Van Kranendonk, MJ,
References Jones, LEA, Kositcin, N, Gessner, K, Hall, CE, Chen, SF, Patison, N,

Kennett, BLN, Jones, T, Goodwin, JA, Milligan, P and Costelloe, RD
Ahmat, AL 1986, Petrology, structure, regional geology and age of the 2014, Geodynamic implications of the Youanmi and southern
gabbroic Windimurra complex, Western Australia: The University of Carnarvon deep seismic relection surveys: a ~1300 km traverse from
Western Australia, Perth, PhD thesis (unpublished), 279p. the Pinjarra Orogen to the eastern Yilgarn craton, in Youanmi and
southern Carnarvon seismic and magnetotelluric (MT) workshop
Ahmat, AL and De Laeter, JR 1982, Rb–Sr isotopic evidence for 2013 compiled by S Wyche, TJ Ivanic and I Zibra: Geological Survey
Archaean–Proterozoic crustal evolution of part of the central Yilgarn of Western Australia, Perth, Record 2013/6, p. 147–166.
Block, Western Australia: constraints on the age and source of the
anorthositic Windimurra Gabbroid: Journal of the Geological Society Nebel, O, Arculus, RJ, Ivanic, TJ and Nebel-Jacobsen, YJ 2013a, Lu–Hf
of Australia, v. 29, p. 177–190. isotopic memory of plume–lithosphere interaction in the source of
layered mafic intrusions, Windimurra Igneous Complex, Yilgarn
Brown, M 2007, Crustal melting and melt extraction, ascent and Craton, Australia: Earth and Planetary Science Letters, v. 380,
emplacement in orogens: mechanisms and consequences: Journal of p. 151–161.
the Geological Society, v. 164, p. 1–22.
Nebel, O, Arculus, RJ, Ivanic, TJ, Rapp, R and Wills, KJA 2013b, Upper
Bunting, JA 2004, The nickel–PGE potential of the Narndee and Zone of the Archean Windimurra layered mafic intrusion, Western
Windimurra intrusions; Apex Minerals NL: Geological Survey of Australia: insights into fractional crystallisation in a large magma
Western Australia, Statutory mineral exploration report, A69643 chamber: Journal of Mineralogy and Geochemistry, v. 191, no. 1,
(unpublished). p. 83–107.
Cawthorn, RG and Walraven, F 1998, Emplacement and crystallization Nebel, O, Mavrogenes, JA, Arculus, RJ, Ivanic, TJ and Langford, R 2010,
time for the Bushveld complex: Journal of Petrology, v. 39, no. 9, Geochemical depth-profiling of late-stage melts from the ~2.8 Ga
p. 1669–1687, doi:10.1093/petroj/39.9.1669. Windimurra Igneous Complex, Western Australia, in Abstracts: Fifth
Costelloe, RD and Jones, LEA 2014, Youanmi seismic survey 2010: International Archean Symposium, Perth, 5 September 2010, 4p.
acquisition and processing, in Youanmi and southern Carnarvon Nelson, DR 2001, 169003: vesicular rhyolite, Carron Hill;
seismic and magnetotelluric (MT) workshop 2013 compiled by Geochronology Record 170: Geological Survey of Western Australia,
S Wyche, TJ Ivanic and I Zibra: Geological Survey of Western 4p.
Australia, Perth, Record 2013/6, p. 1–6.
Parks, J 1998, Weld Range platinum group element deposit, in Geology
Gill, M 2011, The petrogenesis and FeTiO accumulation of the Youanmi of Australian and Papua New Guinean mineral deposits edited by
Igneous Complex (Yilgarn Craton), Western Australia: University of DA Berkman and DH Mackenzie: The Australian Institute of Mining
Tasmania, Honours thesis (unpublished). and Metallurgy, Monograph 22, p. 279–286.
Ivanic, TJ 2009, Madoonga, WA Sheet 2444: Geological Survey of Scowen, PAH 1991, The geology and geochemistry of the Narndee
Western Australia, 1:100 000 Geological Series. intrusion: Australian National University, Canberra, Australian
Ivanic, TJ 2011, Coolamaninu, WA Sheet 2540: Geological Survey of Capital Territory, PhD thesis (unpublished), 214p.
Western Australia, 1:100 000 Geological Series. Van Kranendonk, MJ, Ivanic, TJ, Wingate, MTD, Kirkland, CL and
Ivanic, TJ 2012a, Challa, WA Sheet 2541: Geological Survey of Western Wyche, S 2013, Long-lived, autochthonous development of the
Australia, 1:100 000 Geological Series. Archean Murchison Domain, and implications for Yilgarn Craton
tectonics: Precambrian Research, v. 229, p. 49–92.
Ivanic, TJ 2012b, Windimurra, WA Sheet 2641: Geological Survey of
Western Australia, 1:100 000 Geological Series. Wang, Q 1998, Geochronology of the granite–greenstone terranes in
the Murchison and Southern Cross Provinces of the Yilgarn Craton,
Ivanic, TJ 2014, Youanmi, WA Sheet 2640: Geological Survey of Western Western Australia: Australian National University, Canberra, PhD
Australia, 1:100 000 Geological Series. thesis (unpublished), 186p.
27
Ivanic and Brett
Wingate, MTD, Kirkland, CL and Ivanic, TJ 2012, 194747: metagabbro,

Malamiter Well; Geochronology Record 1013: Geological Survey of
Western Australia, 4p.
Wyche, S, Ivanic, TJ and Zibra, I (compilers) 2014, Youanmi and
southern Carnarvon seismic and magnetotelluric (MT) workshop:
Geological Survey of Western Australia, Record 2013/6, 180p.
Wyche, S, Kirkland, CL, Riganti, A, Pawley, MJ, Belousova, E and
Wingate, MTD 2012, Isotopic constraints on stratigraphy in the
central and eastern Yilgarn Craton, Western Australia: Australian
Journal of Earth Sciences, v. 59, no. 5 (Archean evolution — Yilgarn
Craton), p. 657–670, doi:10.1080/08120099.2012.697677.
Wyman, DA and Kerrich, R 2012, Geochemical and isotopic
characteristics of Youanmi terrane volcanism: the role of mantle
plumes and subduction tectonics in the western Yilgarn Craton:
Australian Journal of Earth Sciences, v. 59, no. 5 (Archean evolution
— Yilgarn Craton), p. 671–694, doi:10.1080/08120099.2012.702684.
Zibra, I, Ivanic, TJ and Chen, SF 2013, Wynyangoo, WA Sheet 2542:
Geological Survey of Western Australia, 1:100 000 Geological Series.
28
RECORD 2015/12
This Record is published in digital format (PDF) and is available as a free
download from the DMP website at
<www.dmp.wa.gov.au/GSWApublications>.
Further details of geological products produced by the
THE WINDIMURRA IGNEOUS COMPLEX, YILGARN CRATON:

Geological Survey of Western Australia can be obtained by contacting:
Information Centre
AN ARCHEAN LAYERED INTRUSION REVEALED BY

100 Plain Street
EAST PERTH WESTERN AUSTRALIA 6004
SEISMIC DATA AND 3D MODELLING

Phone: +61 8 9222 3459 Fax: +61 8 9222 3444
www.dmp.wa.gov.au/GSWApublications
Ivanic and Brett

The Windimurra Igneous Complex, Yilgarn Craton: An Archean Layered Intrusion Revealed by Seismic Data and 3D Modelling

Uploaded by

Copyright:

Available Formats

The Windimurra Igneous Complex, Yilgarn Craton: An Archean Layered Intrusion Revealed by Seismic Data and 3D Modelling

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Windimurra Igneous Complex, Yilgarn Craton: An Archean Layered Intrusion Revealed by Seismic Data and 3D Modelling

Uploaded by

Copyright:

Available Formats

Government of Western Australia

Department of Mines and Petroleum

THE WINDIMURRA IGNEOUS COMPLEX, YILGARN

THE WINDIMURRA IGNEOUS COMPLEX, YILGARN

DIRECTOR GENERAL, DEPARTMENT OF MINES AND PETROLEUM

EXECUTIVE DIRECTOR, GEOLOGICAL SURVEY OF WESTERN AUSTRALIA

National Library of Australia Card Number and ISBN 978-1-74168-640-1

Published 2015 by Geological Survey of Western Australia

The Windimurra Igneous Complex, Yilgarn Craton:

Introduction the results of a mantle plume affecting a wide region of

118° 122° 118° 119°

Warriedar Suite Polelle Group Seismic line, with CDP points

Meeline Suite c. 2.95 Ga formations

Corner Well Gabbro

c. 2813 Ma Meeline Suite

<2821 Border zone dolerites

w c. 2819 Ma Middle zone

20 2820 – 2810 Ma lithologies (dominantly Norie

30 2813 Geochronology sample site showing age

Seismic line showing CDP points

New gravity and magnetic survey data acquired from

Igneous Igneous Igneous Igneous gabbro Igneous

Vertical thickness (m)

Table 3. Summary of rock density determinations for GSWA samples

Construction of the 3D model 7. ‘EW1’ — A supplementary cross-section for extra

Table 5. Modelled stratigraphic column describing the features merged

Unit order and codename Unit name Comments

Results complex as a whole. The upper part of the complex, in

Magnetic susceptibility (SI units) Density (g/cm3)

Table 7. Modelled shear zones in the ‘fault matrix’

F03_WindiEastShearZone TRUE – TRUE

Observations from 10GA-YU1 exposures of modally layered gabbroic cumulates are

10GA-YU2 CDP 5880

b) 10GA-YU2 10GA-YU3 CDP 9308 10GA-YU1 CDP 16497

10GA-YU2 CDP 4422

Reflector trend line

10GA-YU2 CDP 5880

e) 10GA-YU2 10GA-YU3 CDP 9308 10GA-YU1 CDP 16497

10GA-YU2 CDP 4422

Middle 2.3 2.8 3.1 10GA-YU2 2.2

Upper 0.7 0.8 2.6 10GA-YU3 0.6

Discussion an estimate for the brittle ductile transition at 2810 Ma of

of the form of the western lobe still remain relatively

effectively accounts for the ‘missing ultramafic zone’

Igneous Complex represents the ultramafic portion of

of the complex as highlighted in the 10GA-YU3 seismic

SW NE 3. 3D inversions show that a significant ultramafic zone,

upper zone intruded as megapulse

References Jones, LEA, Kositcin, N, Gessner, K, Hall, CE, Chen, SF, Patison, N,

Wingate, MTD, Kirkland, CL and Ivanic, TJ 2012, 194747: metagabbro,

Further details of geological products produced by the

THE WINDIMURRA IGNEOUS COMPLEX, YILGARN CRATON:

AN ARCHEAN LAYERED INTRUSION REVEALED BY

SEISMIC DATA AND 3D MODELLING

Ivanic and Brett

You might also like