Fig. 2. Crustal structure models. On the left is the mean seismic refraction model from Drummond 1983, and on the
right is the gravity and magnetic anomaly interpretation.
deeper cover using the general correlation of broad, magnetic lows on granitoid complexes, and
narrow linear magnetic highs over greenstone belts.
Where the boundary outcrops, geological ob- servations show that it is generally very steep. The
average horizontal position of the boundary be- tween granitoid complex and greenstones, at a
depth of about one half way down the boundary, is given by the centre of the maximum gradient of
the gravity anomaly. This maximum gravity gra- dient has a horizontal extent of about 10 km, so
the position of the centre of this maximum gradi- ent can be mapped to within about 3 km over the
whole land area because the gravity station spac- ing is 5 and 11 km.
Fig. 3 shows the position of the granitoid com- plex boundary in outcrop and subcrop, and the
inferred position of the granitoid complex margin at depth. The relative position of these two lines
gives the average direction of dip of the granitoid complex margin. The granitoid complex margins
dip under the granitoid complex inwards, and under the greenstones outwards, for approxi-
mately equal horizontal distances, so the average dip of the granitoid complex margins to the base
is near vertical. The smoother shape of the grani- toid complex margin at depth relative to the
margin of the granitoid complex outcropsubcrop, is due to the maximum gravity gradient showing
the average position of the granitoid complex margin at depth, due to both the averaging effect
of gravity anomalies, and the low resolution of the gravity survey. The complexity of the grani-
toid complex margin at depth is undefined from the gravity anomalies, and it is likely to be similar
to that at the surface.
6. Depth extent of upper crustal structures
Estimates of the depth extent of upper crustal structures can be obtained from the shape of the
tails of the magnetic and gravity anomalies, and the amplitude of the gravity anomalies.
Within the Pilbara Craton there are high-ampli- tude, long-wavelength magnetic anomalies occur-
ring above some of the greenstone belt synclines. near sea level in the north part of the Pilbara
Craton, and about 500 m on the southern margin of the Pilbara Craton. The Pilbara crust is rela-
tively thin, with a large increase in velocity at the base of the crust.
The gravity anomalies Fig. 1a show only slight changes in average value across the craton.
These slight changes are mainly a decrease from the margin to the centre of the craton. The mag-
netic anomaly before removal of the regional, shows only a gradual change in level to the north
and Northeast. The above geophysical informa- tion, from gravity, magnetic and seismic surveys,
shows the Pilbara upper crust appearing to be a single unit, with no regions with different
properties.
5. Margins between granitoid complexes and greenstones
The position of the boundary between granitoid complexes and greenstones has been mapped us-
ing traditional geological methods over the central band of the Pilbara Craton where granitegreen-
stone is exposed Hickman, 1983. This boundary has been mapped under shallow cover using the
texture of the short-wavelength magnetic anoma- lies which reflects shallow geology, and under
The better developed anomalies have an amplitude of 1000 – 3600 nT, a width at one half amplitude of
about 9 km, and a strike length of about 20 km Fig. 4. The depth to the top of the causative
bodies has been estimated using the Vacquier et al. 1951 method, and by modelling 2D sections; it
averages 1.5 km in the west, and 2.0 km in the east of the Pilbara Craton. Given depths to the top,
and anomaly widths about 9 km, 2D modelling shows that the width of a body with vertical sides
has to be similar to the anomaly width at half amplitude. The gravity anomalies are interpreted
above to indicate steep boundaries between grani- toid complex and greenstone belts; hence for sim-
plicity the type of body modelled is a vertical prism, and the observed anomalies were converted
to approximately symmetrical anomalies by reduc- tion-to-the-pole using a magnetic inclination of
−
70 to − 80°. The true magnetic inclination of the area is − 56°, but as discussed below the
induced and remanent magnetisation is in the plane of the geological formation bedding. For
tabular, vertical sided models with vertical mag- netisation the depth to the base of the bodies can
be determined approximately from the distance of the flanking lows from the steep gradients of the
anomaly. For a body of this type, this distance is 3.5 km for a base at 5 km, and 6.5 km for a base
at 14 km Fig. 5. In the area of the largest broad anomalies Fig. 4 the distance out of these lows
averages about 6 km, consistent with the depth to the base of the body of about 14 km. Interpreta-
tion of these large magnetic anomalies was also carried out by 2.5D modelling of profiles using the
software package ModelVision by Encom. Each profile of reduced-to-the-pole anomaly was mod-
elled by a tabular, vertical sided body, with an appropriate depth to the top and width. The
apparent susceptibility and depth to the base of the body was then varied until the amplitude, and
the tails on both sides of the anomaly, were approximately correct. Fig. 6 gives the interpreta-
tion for the largest anomalies. Importantly, for the relatively wide and deep anomalies in both the
eastern and western parts of the Pilbara craton, profile modelling showed that the depth to the
base of the anomalous bodies is about 14 km. The mean apparent susceptibilities of these tabular
bodies was generally about 0.1 – 0.2 SI. This is discussed below.
Fig. 3. Extent of granites. The trace of the granitoid complex boundary in outcrop and subcrop is shown by the thin continuous line. The maximum gravity gradient is shown by the discontinuous line. The separation of these lines is a measure of the dip of the
granite margin. Granite interpreted to be relatively thick base at 14 km are shaded with a dotted outline, other granites not shaded generally have a base at about 8 – 12 km. Parallel lines give margin of the Pilbara Craton from gravity anomalies. Y, C, and
P indicate the Yule, Carlindi and Pippingarra Granitoid complexes.
Fig. 4. Map of large magnetic anomalies under the greenstone belts of the Pilbara Craton. Contours give magnetic anomalies
reduced to the pole assuming a magnetic inclination of − 80°, and using a contour interval of 250 nT. Thick black lines gives
outcropsubcrop contact between the granites and greenstone belt. The white dotted line gives the steepest gravity gradient,
showing the smoothed position of the margin of the granites at depth.
ture at 14 km Fig. 7. Although this and the other modelled section are consistent with a depth
extent of upper crustal structure being about 14 km, the gravity modelling is unconvincing, be-
cause of the poor control of 5 km spacing data on gradients, the relative subtlety of the differences in
calculated anomaly shape between upper crustal structures 5 and 14 km deep, and the unrealistic
simplicity for the constant density differences for such large horizontal and vertical distances.
The amplitude of the gravity anomalies gives information on the depth extent of the upper
crustal structures, provided that the average den- sity contrast between a granitoid complex and a
greenstone belt can be estimated. I have adopted the following Yilgarn Craton mean densities for
the major rock types: granitoid 2.65 t m
− 3
, sedi- ments 2.72 t m
− 3
, felsic volcanic rock 2.74 t m
− 3
, maficultramafic rocks 2.92 t m
− 3
House, 1996. In the area of the larger gravity anomalies in the
eastern part of the Pilbara Craton, greenstone belts are composed of maficultramafic rocks plus
Fig. 5. Modelling of large magnetic anomalies by a 2.5D vertical prism. For the same anomaly width, anomaly ampli-
tude, and depth to the top of the body, a body with a base at 5 km has a minimum anomaly closer to the body, than a body
with a base at 14 km. Arrows show the position of the minimum.
The depth extent of a vertical contact can, in theory, be determined from the shape of the 2D
gravity profile across the contact Fig. 7. Two profiles were modelled, one in the east, and one in
the west part of the Pilbara Craton. The gravity profiles were taken where observed gravity was
better determined 5 km observation grid, where the greenstone belt and granitoid complex were
relatively wide so that both tails could be ob- served, and the granitegreenstone contact was
near linear. The granitoid complex was modelled in three dimensions, as a 3D prism with a plan
shape of the maximum gravity gradient, a vertical margin, a thickness of 14 km, and a constant
horizontal density contrast with the surrounding greenstone belts. The observed and calculated
profiles across a modelled boundary are consistent to within the accuracy of the 5 km station spacing
for a model with the base of upper crustal struc-
Fig. 6. Large magnetic anomalies over the greenstone belts crosses, with their modelled anomalies using tabular bodies
lines. The bodies are not shown. For each anomaly the body is a 2.5D vertical prism, with a top at 2 – 4 km, and a base at
14 km. A sloping straight line shows an inferred regional field. The centres of the profiles are at 21°05S 119°15E, 20°55S
120°20E, 20°55S 119°45E, 21°20S 117°45E.
0.27 t m
− 3
greenstone belts with 100 maficul- tramafic rocks, and is likely to be about 0.21 t
m
− 3
greenstone belts with 70 maficultramafic rocks and 30 sediment and felsic volcanic rock.
The probable limits are 15 – 45 sediment and felsic volcanic rock, so the uncertainty of the
density contrast is + 20 – 12. 3D gravity mod- elling was carried out over those greenstone belts
with the largest gravity anomaly contrasts be- tween granitoid complexes and greenstone belt.
The greenstone belt was represented by a 3D prism with vertical sides, and a plan shape that is
similar to the maximum gravity gradient. The two largest observed anomalies were about 650 mm
s
− 2
amplitude, over a body about 30 km wide, but about 15 km from one end of the body. The
modelling showed that for these two larger anomalies, the greenstone belt thickness would be
about 14 km for a density contrast of 0.21 t m
− 3
. The interpretation of metamorphic assemblages
gives some geological support for greenstone belt rocks extending down to a considerable depth in
the past. In the Warrawooma Syncline occur both greenschist facies rocks and kyanite bearing
schists that have been buried to at least 6 kbar Collins and van Kranendonk, 1999, and this is
interpreted as indicating that the greenstone belts were subvertical to about 20 km below the then
surface, and that slivers of the belts were rapidly uplifted Collins et al., 1998.
Fig. 7. Model of a gravity gradient between a granitoid complex and greenstone belt. Observed profile is continuous
line, modelled profile is dashed line. Profile location is 21°24S 117°06E. In this location the greenstone belt is likely to be
mainly sediments.
some sediments and felsic volcanic rocks, so the density contrast between granitoid complex and
greenstone belts must be significantly less than
Fig. 8. Distribution of banded iron formations and shear zones. Black areas; high-amplitude, wide magnetic anomalies interpreted to be due to banded-iron formation material deep in the upper crust. Thin line; trace of the granitoid complex boundary in outcrop
and subcrop. Thick lines; extent of major shear zones crossing the west and north Pilbara Craton, mapped from gravity and magnetic anomalies, except AA the Scholl Shear Zone mapped by geology. Parallel lines; margin of the Pilbara Craton from gravity
anomalies.
Using the above arguments and magnetic, grav- ity, and geological data, I conclude that the upper
crustal structures extend down to about 14 km, the base of the upper crust indicated by seismic
refraction interpretation.
The minimum geographical extent of the green- stone belt synclines that extend well down into
the upper crust is shown by the extent of high-am- plitude long-wavelength magnetic anomalies Fig.
8, and by the axes of the gravity highs. These indicators allow deep synclines to be mapped in
areas of cover, and, importantly, show that deep synclines occur in both the eastern and western
part of the Pilbara Craton. The deep synclines are displaced by major shear zones in the west
and north. The similar amplitude of the gravity anomalies
throughout the
craton is
consis- tent with the granite and greenstone structures
extending to the base of the crust throughout the craton.
7. Cause of the large magnetic anomalies