The crustal geophysical signature of a world-class magmatic mineral system

World-class magmatic mineral systems are characterised by fluid/melt originating in the deep crust and mantle. However, processes that entrain and focus fluids from a deep-source region to a kilometre-scale deposit through the crust are unclear. A magnetotelluric (MT) and reflection seismic program across the margin of the Gawler Craton, Australia yield a distinct signature for a 1590 Ma event associated with emplacement of iron-oxide copper gold uranium (IOCG-U) deposits. Two- and three-dimensional MT modelling images a 50 km wide lower-crustal region of resistivity <10 Ωm along an accreted Proterozoic belt. The least resistive (~1 Ωm) part terminates at the brittle-ductile transition at ~15 km, directly beneath a rifted sedimentary basin. Above the brittle-ductile transition, three narrow low-resistivity zones (~100 Ωm) branch to the surface. The least resistive zone is remarkably aligned with the world-class IOCG-U Olympic Dam deposit and the other two with significant known IOCG-U mineral occurrences. These zones are spatially correlated with narrow regions of low seismic reflectivity in the upper crust, and the deeper lower-crust conductor is almost seismically transparent. We argue this whole-of-crust imaging encapsulates deep mineral system and maps pathways of metalliferous fluids from crust and mantle sources to emplacement at discrete locations.

OD1 1,2 . In contrast with the original processing 2 of line 03GA-OD1 where workflow was designed to provide a consistent image of the whole crust, this processing technique emphasises variation in signal strength by preserving the temporal relative amplitude ratio, which has the effect of sharpening the output image, particularly in the upper crust.
Wise et al. 1 compared the results of the differing processing methodologies, and concluded that greater reflectivity contrasts in the upper crust are revealed by this approach, permitting interpretation of steeply dipping seismic structures and weakly-reflective zones which may reflect possible fossil fluid pathways.

MT data processing
Magnetotellurics is natural-source EM method used to explore the resistivity distribution of Earth on scales of tens of metres to hundreds of kilometres 3 . In the MT method, orthogonal components of inducing horizontal magnetic fields (H) and resulting horizontal electric fields (E) are measured simultaneously as a function of time to determine the Earth's electrical impedance as a function of frequency. The fields are related, in the frequency domain, by the impedance tensor (Z) given by E=ZH. Apparent resistivity as a function of frequency f is given by = 1 0 | | 2 , where Z is an impedance element, µo is magnetic permeability of free space, ω is angular frequency = 2 . The skin-depth (approximate depth of investigation in kilometres) is ( ) ≈ 0.5√ where is the apparent resistivity, or the average resistivity of an equivalent half-space, and T is the period of induction. The complex impedance tensor can be written in terms of its real (X) and imaginary (Y) parts as Z = X + i Y, from which the MT phase tensor is defined by the relation = − and is not affected by galvanic distortion 4,5 .
The Olympic Dam transect A-A' in Figure 1 consisted of a 200 km transect of 110 broadband sites with a 1 to 2 km spacing and 30 long-period instruments at 5 km intervals 6 . An additional thirty broadband sites along B-B' were deployed to form a northeast-southwest transect just north of Olympic Dam at 2 km intervals. The MT data were processed using a robust, remotereference code 7 , resulting in response functions over the period range of 10 1 -10 4 s for longperiod sites and 10 -3 -10 3 s for broadband sites.
Strike analyses using the azimuth of phase tensor 4 and invariants of impedance tensor 8 approaches revealed a dominant geo-electric strike of N115°E at periods >1 s, as shown in Figure S1. At short periods (<1 s), there is no well-defined strike as induction is predominantly in the sedimentary cover which is 1D. We note that the longer-period geoelectric strike is consistent with the strike of the long-wavelength Bouguer gravity trends in Figure S2, and although there is a 90 degree ambiguity from the tensor alone, the geology indicates that this is the correct modelling orientation.

MT modelling
All MT responses were rotated to this geoelectric orientation and the modelled transect was N25°E (at right angles to the geoelectric strike). In this rotated frame, the orientation of the impedance component with magnetic field parallel to geoelectric strike is denoted the TM mode, and the impedance component with the electric field parallel to geo-electric strike is the TE mode. Phase tensor skew angles are less than five degrees for almost all sites at periods up to 10 2 s, consistent with data being responsive to a 2D regional resistivity structure ( Figure S2, S3). Even at longer periods of 10 3 s ( Figure S3) most sites are 2D, with a few regions of 3D response.
On the basis that the 110 sites were 2D to at least 10 2 s and most to 10 3 s, we inverted profiles A-A' and B-B' using the WinGlink program 9 . Many inversions were carried out, systematically testing different starting models and smoothing parameters to assess the robustness of the features shown in Figure 2. It was found that smoothing parameter () value of between 1 and MT data and the model response resulting from 2D inversions of the preferred models for selected sites is presented in Figure S4. At periods >10 2 s, the TM mode is better fitted compared to TE mode as the TM mode is less affected by resistivity variations due to offtransect 3D effects.   spaced approximately 10 km apart to with total RMS of 1.8. At periods >100 s, the TM mode is better fit as this mode is less affected by resistivity variations due to 3D effects.
The inversion also allowed for static shift on both modes of the apparent resistivity to be determined as an independent variable. Figure S5 shows that static shifts are small, with, less than half-an-order of magnitude spread across all sites. This is expected as the survey area is in a thick sedimentary cover that has relatively uniform low-resistivity.  Figure S6 shows a low-resistivity zone C3 below 25 km consistent with the 2D and 3D inversions. However, C3 is more resistive that the C3 region under A-A', and it appears that this reflects the limited spatial extent of the conductor in the NW-SE orientation.