An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region. Modelling new geophysical data

Transcription

1 DAH CUNNAMULLA, % Line 1 A) Airborne Electromagnetics (AEM) along L nt ame e n i oa L Culg Hungerford Granite SE Location of Olepoloko Fault Brewarrina Granite? 300m xle yf au BREWARRINA lt BOURKE, % Line 3 50 e Littl Mou nta B) Broadband Magnetotellurics (BBMT) alon, % 0m ult in Fa uma th-e Lou one ar Z he rra S 500m DTB Inferred from GABWRA Olepoloko Fault 146 E 100 Record ity Line 2017/ ecat Ü Km 1000m 0km 100km C) Basement Geology An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region n or settlement Modelling new geophysical data Chris B. Folkes APPLYING GEOSCIENCE TO AUSTRALIA S MOST IMPORTANT CHALLENGES

2 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region Modelling new geophysical data GEOSCIENCE AUSTRALIA RECORD 2017/01 Folkes, C. B

3 Department of Industry, Innovation and Science Minister for Resources and Northern Australia: Senator the Hon Matthew Canavan Assistant Minister for Industry, Innovation and Science: The Hon Craig Laundy MP Secretary: Ms Glenys Beauchamp PSM Geoscience Australia Chief Executive Officer: Dr Chris Pigram This paper is published with the permission of the CEO, Geoscience Australia Commonwealth of Australia (Geoscience Australia) 2017 With the exception of the Commonwealth Coat of Arms and where otherwise noted, this product is provided under a Creative Commons Attribution 4.0 International Licence. ( Geoscience Australia has tried to make the information in this product as accurate as possible. However, it does not guarantee that the information is totally accurate or complete. Therefore, you should not solely rely on this information when making a commercial decision. Geoscience Australia is committed to providing web accessible content wherever possible. If you are having difficulties with accessing this document please clientservices@ga.gov.au. ISSN X (PDF) ISBN (PDF) ecat Bibliographic reference: Folkes, C. B., An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region: modelling new geophysical data. Record 2017/01. Geoscience Australia, Canberra. Version: 1701

4 Contents Executive Summary Introduction Motivation for this project and objectives Geological Background Overview of the southern Thomson region Geology of the Southern Thomson region Methods used Gravity Airborne Electromagnetics Magnetotellurics Seismic Interpreted basement geology Drillholes and waterbores High-resolution ground geophysics Rock Density Forward Modelling Results and Discussion Comparison of MT, AEM and other datasets Line Line Other MT traverses Summary Forward Gravity Modelling Southern section of Line Whole of Line Line Lines 6 and 7a (combined) Cover thickness variations tested by gravity modelling Conclusions and recommendations Acknowledgements References...39 Appendix A Basement lithology descriptions...41 Appendix B Gravity modelling and inversion details...44 Appendix C Comparison of AEM and AMT conductivity sections...46 C.1 Line 7b...46 C.2 Line C.3 Line 3 (section with AMT)...48 Appendix D Depth to Basement estimates from other datasets...49 Appendix E Available density data...50 Appendix F Reinterpretation of seismic images and new gravity modelling...53 Appendix G Cover thickness variations tested by gravity modelling...55 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region iii

5 G.1 Southern Section of Line G.2 Whole of Line iv An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

6 Executive Summary The under-cover geology of the southern Thomson Orogen in north-western New South Wales and south-western Queensland is largely unknown due to the extensive, up to 600 m thick Cenozoic and Mesozoic cover. This cover (mainly consisting of Eromanga Basin and Lake Eyre Basin rocks) results in very restricted basement outcrop, with a subsequent lack of understanding of subsurface lithologies, structures and the potential for the location of economic resources. As a result, this area was selected for a regional, multi-disciplinary project (the Southern Thomson Project) by Geoscience Australia and its State partners the Geological Survey of New South Wales and the Geological Survey of Queensland. The Project reflects the focus of the UNCOVER Initiative (Australian Academy of Science 2012) that aims to form the basis for Australia s potential future discovery and development of new economic mineral resources. The Southern Thomson Project involves many geoscientific disciplines including geophysics, geochronology, geochemistry and stratigraphy to better understand the region and promote mineral exploration by reducing exploration risk. This report focuses on some of the initial reconnaissance and pre-drilling geophysical data collected in 2014 in particular gravity data, AEM (Airborne Electromagnetics) and MT (Magnetotellurics) along two regional north-south traverses, and a shorter east-west traverse in the northern part of the region. The major aim of this study is to compare AEM and MT electrical conductivity data acquired along these traverses, and integrate them with interpretation of available deep seismic reflection data to generate a series of 2D geological models, which can be tested via forward gravity modelling and subsequent density inversions. This integrative approach allows for a more robust understanding of the crustal architecture and cover thickness (or depth to basement) variations in the Southern Thomson region. The main findings of this report are: Cover thicknesses of 0 to >500 m were initially estimated along various traverses through a combination of AEM and MT data interpretation as well as existing data from drill holes and water bores. Most datasets yield broadly similar results in terms of relative cover thickness variations, although AEM cannot be reliably used when cover thickness is greater than ~150 m due to limitations in the Depth Of Investigation (DOI), and Broadband MT (BBMT) tends to overestimate cover thickness where it is known to be less than 50 m. Cover thickness estimates using MT methods (especially AMT Audio-frequency Magnetotellurics) agree with other datasets such as existing drill holes/water bores, GABWRA (Great Artesian Basin Water Resource Assessment; Ransley and Smerdon 2012) depth to basement results, and targeted high-resolution ground geophysical surveys (Goodwin et al. in prep). On this regional scale, AMT likely provides the most suitable resolution for estimating cover thicknesses of m. Cover thicknesses estimated by AEM and MT conductivity sections have been tested by forward gravity modelling and produce better matches with the observed gravity responses compared to an averaged, uniform cover thickness. This observation shows cover thickness variations can produce discernible variations in gravity responses and need to be taken into account in gravity modelling. Further, this supports the use of a combined approach in using AEM, MT and gravity models to asses cover thickness variations over a broad region. Several alternative interpretations of deep seismic reflection data along the southern part of one of the regional MT traverses (Line 3) were performed to assess crustal architecture. These were tested by forward gravity modelling with subsequent inversions (allowing modelled bodies density An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 1

7 to vary) producing a close match between the observed and modelled gravity responses with reasonable geological densities of crustal units given the limited known and/or inferred rock properties in this region. Two-dimensional (2D) cross-section models along each line were generated by integrating the recent interpretation of basement geology (Purdy et al. 2014) with AEM and MT conductivity sections. These models were tested via forward gravity modelling and subsequent inversions (allowing modelled bodies density to vary). This approach showed that the most accurate model was a thickened crust north of the Olepoloko Fault (the Southern Thomson region). Many of the 2D forward models produced reasonable matches between the observed and calculated (modelled) gravity responses with respect to the large scale crustal architecture and location of prominent resistive bodies (inferred as felsic igneous intrusions) observed in MT conductivity sections. However, gravity inversions sometimes produced unrealistic densities of crustal units given the (limited) known rock properties in this region. Despite these limitations, the simplistic 2D forward models provide a good starting point for future refinement as more geological, geophysical, geochemical and petrophysical data become available. 2 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

8 1 Introduction 1.1 Motivation for this project and objectives The UNCOVER Initiative (Australian Academy of Science 2012) recognises that near-surface mineral resources are becoming increasingly difficult to discover with classical exploration techniques such as saprolite and soil sampling due to the often substantial thickness of near-surface highly-weathered bedrock and sedimentary basin material covering ~80% of the Australian landmass (Roach et al. 2016). There are four themes of the UNCOVER Initiative: 1. Characterising Australia s cover 2. Investigating Australia s lithospheric architecture 3. Resolving the 4D geodynamic and metallogenic evolution of Australia 4. Characterising and detecting distal footprints of mineralisation The Southern Thomson Orogen has been identified as one of the primary focus areas for the UNCOVER Initiative by Geoscience Australia (GA). The geology of this region in northern New South Wales (NSW) and southern Queensland is poorly understood. Basement geology is rarely exposed and there are generally many tens to a few hundred metres of overlying unconsolidated or indurated Cenozoic and Mesozoic cover, largely consisting of rocks of the Eromanga Basin or the Lake Eyre Basin. The Southern Thomson Project is a collaboration between GA, the Geological Survey of New South Wales (GSNSW) and the Geological Survey of Queensland (GSQ). The Project aims to collect new geophysical, geological, geochemical and geochronological data to improve the geological understanding of the area, promote mineral exploration in the region by reducing exploration risk, and inform future land and water resource management decision making. The data collation will also guide targeting for stratigraphic drilling, and will provide answers to scientific problems regarding the assembly of the Tasmanides, the age of the Thomson Orogen compared to the adjoining Lachlan Orogen, and the potential mineral systems that may occur within the southern Thomson Orogen. As part of the Southern Thomson Project, new geophysical data, including airborne electromagnetics (AEM), ground gravity, magnetotelluric (MT), including audio frequency magnetotelluric (AMT) and broadband magnetotelluric (BBMT) data were collected in the study area along coincident traverses. The traverses were designed to map the Thomson-Lachlan orogen boundary and to map near-surface basement in the Eulo, Queensland area. This work aims to take the newly acquired gravity data and perform 2D forward modelling with subsequent density inversions to investigate: The large-scale crustal architecture of this region The depth of cover and nature of the basement in this region. The models are informed by integrating the MT data with the AEM data of Roach (ed. 2015), existing deep seismic reflection data (Glen et al. 2013), the basement interpretation map of Purdy et al. (2014) and the Great Artesian Basin Water Resources Assessment (GABWRA; Ransley and Smerdon 2012) depth to basement estimates. This integrative approach allows for the construction of models that can An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 3

9 be tested with forward gravity modelling and provides new constraints on the large-scale crustal architecture and cover thickness variations in the Southern Thomson region. 1.2 Geological Background Overview of the southern Thomson region The southern Thomson region in southern Queensland and north-western New South Wales (Figure 1.1) remains an enigma in regards to its large-scale crustal architecture, due to the lack of available basement outcrops coupled with thick overlying cover sediments. Much of the geological information has come from interpretation of aeromagnetic and gravity imagery, supplemented by sparse data from oil wells and water bores. Gravity imagery shows an east-west trending gravity ridge in northern NSW (the east-west zone ; marked by A in Figure 1.2) that possesses a distinct geophysical character and an orientation that contrasts with the north-northwest trends commonly observed further south in the Lachlan Orogen. Some authors have proposed that this east-west zone represents the southern margin of a different terrane the Thomson Orogen that is distinctly different from the Lachlan Orogen to the south (e.g. Glen 2005, Glen et al. 2006; Figure 1.2). Figure 1.1 Map of the Tasmanides of eastern Australia showing the main orogenic belts and other elements (from Fergusson and Henderson 2015). Abbreviations: Delamerian Orogen (GSZ = Grampians-Stavely Zone, GZ = Glenelg Zone, KG = Kanmantoo Group, KB = Koonenberry Belt), Lachlan Orogen (BZ = Bendigo Zone, MZ = Melbourne Zone, SZ = Stawell Zone), Precambrian cratons (CC = Curnamona Craton), TFZ = Tamar Fracture Zone. The rectangle indicates the location of Figure An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

10 Figure 1.2 Pseudo colour RGB composite variable density spherical cap Bouguer anomaly gravity image (taken from Glen et al. 2013). Orange/red colours indicate gravity highs (> 70 gravity units); blue colours indicate gravity lows (< -300 gravity units). Red lines show the existing deep seismic reflection survey lines (Line 1 = TL1; Line 2 = TL2; Line 3 = TL3). The dashed lines show the approximate boundary of the Thomson, Lachlan (and Delamerian) Orogens according to Glen et al. (2013). A marks the east-west gravity ridge and B marks the gravity response of the Warraweena volcanics. Within this context, various authors (e.g. Withnall et al. 1996; Fergusson and Henderson 2013; Glen et al. 2013) identified differences between the northern Lachlan and southern Thomson regions prior to the mid-silurian (~430 Ma). Evidence for this includes igneous rocks with different crystallisation ages and affinities between the two regions, and seismic reflection data suggesting different crustal thicknesses with markedly different structures and layering to the north and south of the Olepoloko Fault (the proposed boundary between the two orogens). It is generally agreed that after the mid- Silurian these two regions have a shared history. Conversely, other authors (e.g. Murray 1986; Burton 2010) have suggested that the southern Thomson Orogen is in fact an extension of the Lachlan Orogen that persists at least 500 km further north into Queensland. This latter study questions the need for the distinction of a separate Thomson Orogen. The evidence cited for this includes a similarity in magnetic and gravity data to the north and south of the Olepoloko Fault and the east-west zone, and a similarity in other ages between rocks in the Lachlan Orogen and further north into central Queensland Geology of the Southern Thomson region For a detailed outline of the basement geology and a description of basement units, the reader is directed to Burton (2010), Glen et al. (2013), and Roach (ed. 2015) a summary of these are An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 5

11 presented here. The north-dipping, east-west trending Olepoloko Fault in the south-western part of the region is interpreted as the surface expression of the Thomson Orogen-Lachlan Orogen boundary, reflecting a post-devonian reactivation of an older crustal-scale structure. This boundary becomes northeast to east-northeast trending further east along the Little Mountain Fault and forms the northern part of the Louth-Eumarra Shear Zone (Figure 1.3). Basement geology in this region (defined as Paleozoic rocks that underlie the Jurassic to Cretaceous Eromanga Basin sequences; Roach ed. 2015), is primarily composed of felsic intrusive, clastic sedimentary and metasedimentary rocks. There is also a lesser proportion of mafic to intermediate volcanic and ultramafic basement rocks throughout the region. Figure 1.3B shows the main basement lithologies in the region as interpreted by Purdy et al. (2014) with a detailed list and their descriptions in Appendix A. The south-western part of the study area consists of a shear-dominated domain with the tectonic grain oriented approximately east-west. This area contains the north-dipping Olepoloko and Mount Oxley faults. The central-western part of this region is dominated by various granite intrusions (e.g. the Hungerford, Granite Springs, and Currawinya Granites), partly terminated by northwest trending faults. The northern and north-eastern part of the region is characterised by a large fold structure with axial planes oriented to the northeast. The central-eastern region is dominated by the Culgoa lineament a broad northerly-dipping shear zone characterised by a distinct aeromagnetic signature arising from magnetite-rich sediments and ultramafic rocks. This region also contains the north-easterly trending Warraweena Volcanics and Paka Tank Trough (Mulga Downs Group) sediments consisting primarily of fluviatile sandstone. The main age constraints of rocks in the region have come from samples of the ubiquitous granitoid magmatism, consisting of dominant I-type and subordinate S-type geochemical affinity. Granites and granodiorites are dominant, forming large plutons that underlie wide parts of the central and western part of this region. The Eulo Ridge basement high (see Figure 5.26 of Roach ed. 2015) contains a large proportion of these granitoid bodies (Roach ed. 2015), which form the few basement outcrops in this area. Magmatic ages (from dominantly U-Pb geochronology) suggest that the main period of intrusive activity in the southern Thomson Orogen occurred from ~420 Ma to 430 Ma, with most magmatism in the northern Lachlan Oregon being slightly younger at <420 Ma. 1.3 Methods used The new geophysical datasets (gravity, AEM, AMT and BBMT; see sections to 1.3.3) were collected along two regional traverses in 2014 (Figure 1.3; Figure 1.4). The two traverses were: Line 1 an eastern traverse starting east of Byrock (NSW) running north to Brewarrina, northwest to the Queensland border at Barringun, then continuing north-west towards Thargomindah (also known as the Thomson East-West AEM Traverse of Roach ed. 2015) Line 3 a western traverse running north from Tilpa (NSW) to the Queensland border at Hungerford and continuing north to Eulo (also known as the Thomson North-South AEM Traverse of Roach ed. 2015). Several shorter traverses were also acquired in the northern part of the study area, as shown in Figure 1.4. The geophysical data and other datasets from the southern Thomson region are listed and detailed below. Table 1.1 summarises the geophysical data available for each traverse. 6 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

12 Table 1.1 The various available geophysical datasets for the major lines of newly acquired gravity, AEM and MT data. The AEM E-W traverses refer to those in Roach (ed. 2015). Line Ground Gravity Deep Reflection Seismic Audio MT Broadband MT Airborne EM 1 Yes No No Yes Yes 2 Yes No No Yes No 3 Yes Partial - Southern part only Partial - (small overlap in centre of line) Yes Yes 5 Yes No Yes No Indirectly (AEM E-W traverses) 6 & 7a Partial (overlaps with NW part of Line 1) No Yes Partial (overlaps with NW part of Line 1) Indirectly (AEM E-W traverses) 7b Partial No Yes No Indirectly (AEM E-W traverses) 8 Partial No Yes No Indirectly (AEM E-W traverses) Gravity Ground gravity measurements were conducted by Atlas Geophysics Pty Ltd with a nominal station spacing of 333 m (Allpike 2014), using a Scintrex CG-5 gravity meter to collect the gravity data. These data were imported into Microsoft Excel software where a complete Bouguer anomaly was calculated at each station by adding the terrain correction to the spherical cap Bouguer anomaly (see Appendix B for the location of the original gravity data and an explanation of how it was manipulated). The full gravity dataset is available as an electronic supplementary file to this Record. These data were then imported into ModelVision version software to perform the forward modelling. ModelVision allows the construction of different shaped polygons (with limited strike extent) of varying densities and for these to be forward modelled against the observed gravity anomaly data. The software also allows the background density and regional gravity variations to fluctuate with an inversion tool Airborne Electromagnetics Airborne electromagnetic data were collected along the two primary traverses (Lines 1 and 3), as well as in a series of east-west flight lines in the northern section of the study region in an area from south of Hungerford to north of Eulo (Figure 1.4). Nominal station spacing for each AEM sounding was 25 m. The AEM technique provides a high resolution image of the electrical properties of the near-surface of the Earth and is useful in estimating cover thickness above basement rock (Roach ed. 2015). In the 2014 AEM survey the average Depth of Investigation (DOI) was calculated as ~166 m for the entire region, varying according to surface electrical conductivity conditions. Inversion modelling of these AEM data produced a Depth to Basement (DTB) model in this region (Figure 1.5B). See Roach (ed. 2015) for a report detailing this dataset and the modelling performed. Although not the main focus of this report, a selection of these modelled east-west conductivity sections are included throughout as they run sub-parallel with many of the lines with gravity and MT data in the northern part of the region. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 7

13 Figure 1.3 A) An aeromagnetic base map image (stretched greyscale 1 st Vertical Derivative Reduced to Pole Total Magnetic Intensity) overlain by the two main gravity traverses (Lines 1 and 3). Blue colours denote low gravity values, orange and red colours denote gravity highs. B) Shows the same region but with the basement interpretation map of Purdy et al. (2014). See Appendix A for a list of basement lithologies and their known/inferred features. 8 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

14 Figure 1.4 Gravity, AEM and MT datasets collected in 2014 for the Southern Thomson Project. The base image is a map of stretched greyscale 1 st Vertical Derivative Reduced to Pole Total Magnetic Intensity. The BBMT collected along Lines 1 and 3 mostly coincides with gravity data collected along these lines. Variations in station locations are due to land access issues. Red stars show the locations of high-resolution ground geophysics (Goodwin et al. in prep). The thick red line denotes the location of the Olepoloko Fault, suggested by Glen et al. (2013) to be the boundary between the Lachlan and Thomson Orogens. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 9

15 1.3.3 Magnetotellurics Two types of MT data have been acquired, which vary with respect to station spacing and penetration depth (e.g. Wang et al. 2017). Broadband MT has a coarse resolution and can image the crust and uppermost mantle. In contrast, AMT is a higher resolution method used with closer station spacing and is more useful in imaging the top few kilometres of the crust (such as imaging depth to basement). Nominal station spacing was 1 km for AMT and 5 km for BBMT with the modelling and generation of conductivity sections presented in Wang et al. (2017). For the majority of this region, BBMT is the only available deep crustal geophysical dataset and has been used for interpreting crustal structures in conjunction with forward gravity modelling. Audio MT conductivity sections were produced along survey lines that intersected Lines 1 and 3 in the northern part of the region as shown in Figure 1.4. The main traverses investigated that contained AMT data were Lines 6 and 7a as these were the longest and had coincident gravity data along the majority of their length. Other AMT conductivity sections are presented in Appendix C Seismic Deep crustal seismic reflection data are available along the southern portion of Line 3 (seismic lines 05GA-TL01 & 05GA-TL02; Figure 1.4). In this Record, the names of these seismic lines are abbreviated to TL1 and TL2. Glen et al. (2013) initially interpreted these data and used the results as an input for forward gravity modelling. These data are reinterpreted in this record, and in combination with the recent basement interpretation of Purdy et al. (2014), are tested against forward gravity modelling. Section provides the results of these interpretations Interpreted basement geology Purdy et al. (2014) compiled an interpretation of basement geology (Figure 1.3B) based on aeromagnetic data (Figure 1.3A) from the Southern Thomson Project study area. This dataset was important in providing an input for the forward gravity modelling. Appendix A provides a detailed list of these basement lithologies and their descriptions Drillholes and waterbores Three DTB models using drillholes and waterbores were obtained for the Southern Thomson region: A Geological Survey of Queensland (GSQ) DTB model (Figure 1.5A) derived from sparse borehole basement intersections and depth to magnetic basement modelling (Simpson and Cant 2013). This dataset uses the Lower Devonian as the top of Paleozoic basement. The Great Artesian Basin Water Resource Assessment (GABWRA) constructed from basement intersections in sparse water bore data and previous interpretations (Ransley and Smerdon 2012). A Basement Elevation Model (BEM) uses the base of the Jurassic as the top of the basement (Layer 10), and a DTB model (Figure 1.5B) was derived by subtracting the GABWRA BEM from the SRTM (Shuttle Radar Topographic Mission) 1-second DEM (Digital Elevation Model) (Ransley and Smerdon 2012; Roach ed. 2015). Groundwater borehole DTB data from the Wanaaring survey from Adams (2011; Figure 1.5A). 10 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

16 These datasets provide independent estimations of the DTB over much of the southern Thomson region (Figure 1.5). Appendix D details how DTB values were extracted from these datasets along the main traverses and integrated with the other available geophysical datasets High-resolution ground geophysics Cover thickness was calculated from high-resolution ground geophysical techniques including passive seismic, refraction seismic and AMT (Goodwin et al. in prep) performed at various localities that overlap or are close to (mostly within 5 km of) the gravity, AEM and MT traverses (Figure 1.4). This dataset provided independent DTB estimations against which the other datasets could be tested Rock Density There are relatively few density measurements for cover and basement rocks in the Southern Thomson region. The main source of existing density values is the Australian Geoscience Information Network (AUSGIN; These are supplemented by data from Adams (2011). Both datasets are presented in Appendix E. Where possible, these density values helped to constrain the various forward models. Wet bulk density measurements are the preferred density values used for this study as they most closely resemble the rocks in their natural state (accounting for porosity and environmental fluids) in the crust that are responsible for producing the observed gravity values. These datasets show that the Eromanga Basin cover has a (wet bulk) density of 2.1 to 2.2 g/cm 3 and the various basement rocks have (wet bulk) densities of 2.4 to 2.9 g/cm Forward Modelling Two-dimensional (2D) forward models were constructed in ModelVision version using the comparison and integration of the conductivity sections and/or interpretations from the different geophysical datasets described above. In particular, cover thickness and crustal architecture were examined using this process. The modelled bodies were given strikes of 50 km, to allow the projection of these 2D models into 3D space (creating 2.5D models). The models were extended at least 30 km beyond the extent of the acquired gravity data boundaries to reduce edge effects of the modelling process during the forward gravity modelling. Density values were attributed to modelled bodies where known (as described in Section 1.3.8). Modelled bodies with unknown densities were given notional (but geologically reasonable) values that were then modified through the inbuilt inversion function in ModelVision to produce the closest match with the observed gravity response whilst reducing geological ambiguity as described in Section 2.2. The rationale behind the construction of individual forward models is described in the relevant sections throughout this Record. Appendix B presents a more detailed description of the methods for forward modelling and gravity inversions (allowing modelled bodies densities to vary). The finalised ModelVision files for each forward model are available as an electronic supplementary dataset to this Record. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 11

17 Figure 1.5 Images showing depth to basement (DTB) estimates in the working area. The base image is a map of stretched greyscale 1 st Vertical Derivative Total Magnetic Intensity Reduced to Pole. A) GSQ DTB model (Simpson and Cant 2013), and the Wanaaring survey DTB (Adams 2011) using groundwater boreholes; B) GABWRA (Ransley and Smerdon 2012) and AEM (Roach ed. 2015) modelling. Survey lines as per Figure An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

18 2 Results and Discussion Estimation of the thickness and character of the cover sequence (in this case, dominated by the Eromanga Basin sediments) is an important factor to be considered before drilling down to basement rocks and/or assessing areas for potential economic resources. Estimations of DTB over a large region such as the southern Thomson Orogen are therefore important when considering future investigations such as potential stratigraphic and exploration drilling targets. 2.1 Comparison of MT, AEM and other datasets The geophysical datasets acquired by GA in 2014 for the Southern Thomson Project provide an opportunity to directly compare and integrate data from different geophysical techniques and drill hole information. This was possible due to the design of the acquisition campaign, where different geophysical techniques were often collected along the same traverses. This style of acquisition yields benefits over the collection of isolated flight lines or transects containing a single geophysical technique, as it provides several constraints for the same traverse and hence can make the geological interpretation and the related forward modelling more realistic Line 3 Figure 2.1 presents a comparison of AEM and BBMT conductivity sections along Line 3. There is a general agreement between the AEM and BBMT sections regarding the relative variations and/or patterns of cover thickness. Cover thins where the dark blue/purple resistive bodies (presumably felsic intrusions) are close to the surface, or outcrop directly (in the case of the Hungerford Granite). Cover thickness is estimated to be thicker (up to 600 m in the case of the BBMT conductivity section) in the area south of the Hungerford Granite outcrop and north of the Olepoloko Fault (Figure 2.1). This area is not imaged by AEM because of low signal penetration. The DTB line from the GABWRA dataset (and AEM modelling of Roach ed. 2015) provides a reasonable match with the base of the conductive cover from the BBMT conductivity section. The BBMT conductivity section (and GABWRA data) tend to overestimate the cover thickness (compared to the AEM data), especially in areas where basement is known to be close to the surface or outcropping, as in the case of the Hungerford and Eulo granites. This is likely due to the wide station spacing of the BBMT data (~5 km) resulting in a coarse conductivity model derived by this method. In areas of cover thickness greater than 150 m, the AEM section along Line 3 tends to underestimate cover thickness (in comparison to the BBMT section). This is due to limitations in the DOI of this technique, which in this area is < 150 m unless resistive bedrock is very close to the surface (see Roach ed. 2015). However, the overall pattern of relative cover thickness in the AEM section matches well with other datasets, such as in the region north of the Olepoloko Fault where the Hungerford Granite crops out at the surface (Figure 2.1). The relatively high variation in cover thickness north of the Hungerford Granite outcrop that is observed in the AEM modelling of Roach (ed. 2015), is also present in the BBMT conductivity section and the GABWRA DTB dataset. The DTB values from groundwater bores of the Wanaaring survey from Adams (2011; with horizontal black bars) and high-resolution ground geophysical surveys (Goodwin et al. in prep) located on or near Line 3 (vertical black bars with circles) are generally consistent with cover thicknesses inferred from the BBMT conductivity section Figure 2.1B). An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 13

19 Figure 2.1 Comparison of A) AEM and B) BBMT conductivity sections along Line 3. Vertical to horizontal ratio = In both cases, blue and purple colours refer to electrically resistive bodies (e.g. felsic igneous bodies) and red and orange colours refer to electrically conductive bodies (e.g. Mesozoic and Cenozoic sedimentary basin cover). The sub-horizontal line overlain on the BBMT section represents the DTB taken from the GABWRA dataset (black line; Ransley and Smerdon 2012) and augmented with the AEM DTB interpretation (white line) of Roach (ed. 2015). Vertical black lines with horizontal bars at their base indicate the DTB from groundwater bores on or near Line 3 from Adams (2011). Vertical black bars (with black circles) denote the maximum cover thickness (with average) estimations based on high-resolution ground geophysical surveys on or near Line 3 (Goodwin et al. in prep). For both datasets, solid vertical lines = surveys within 2 km of Line 3; dashed vertical lines = surveys within 2 10 km of Line 3. Black stars indicate the location of known basement outcrops. Inverted triangles above the BBMT section indicate BBMT station locations. C) Interpretation of basement geology along Line 3 from Purdy et al. (2014). Black arrows above the bottom strip denote the location of inferred faults Line 1 Figure 2.2 shows a comparison of AEM and BBMT conductivity sections and the GABWRA dataset along Line 1. Compared to Line 3, the results show a similar, but not quite so clear, correlation of cover thickness inferred from AEM and BBMT conductivity sections. The DTB estimates inferred from the GABWRA dataset and BBMT conductivity section show a good overall correlation, with thin cover (less than 100 m) above resistive areas on the southern and northern margins of the transect (possibly related to basement highs due to igneous intrusive units e.g. the Granite Springs Granite in the north-western-most part of the transect). Cover is inferred to be thicker (DTB up to 600 m) throughout the central part of the traverse, overlying the extensive Nebine Metamorphics sequence. This thick cover is not imaged by the AEM conductivity section along Line 1, due to the limited DOI of < 150 m depth as observed along Line 3. Interestingly, both the AEM and BBMT conductivity sections indicate shallow resistive bodies at the southern margin of the transect that extend close to the surface below the Girilambone Group. These could be related to the series of dykes as indicated in the basement interpretations from aeromagnetic data and/or the other highly magnetic igneous intrusions located immediately to the south of this transect (Figure 1.3B; Purdy et al. 2014). 14 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

20 Figure 2.2 Comparison of A) AEM and B) BBMT conductivity sections along Line 1. Vertical to horizontal ratio = In both cases, blue and purple colours refer to electrically resistive bodies (e.g. felsic igneous bodies) and red and orange colours refer to electrically conductive bodies (e.g. Mesozoic and Cenozoic sedimentary basin cover). The sub-horizontal line overlain on the BBMT section represents the DTB taken from the GABWRA dataset (black line; Ransley and Smerdon 2012) and augmented with the AEM DTB interpretation (white line) of Roach (ed. 2015). Vertical black lines (with black circles) denote the maximum cover thickness (with average) estimations based on high-resolution ground geophysical surveys within 7 km of Line 1 (Goodwin et al. in prep). The black star indicates the location of known basement outcrop. Inverted triangles above the BBMT section indicate BBMT station locations. C) Interpretation of basement geology along Line 1 from Purdy et al. (2014). Black arrows above the bottom strip denote the location of inferred faults Other MT traverses Other MT traverses in the region provide indications of crustal architecture and absolute and relative cover thickness variations. For example, Lines 3, 5, 6 and 7a (combined), 7b and 8 contain AMT datasets that provide higher resolutions than the BBMT method and are potentially more useful in observing cover thickness variations in the upper 1 km of the crust. Figure 2.3 shows a comparison of AEM and AMT conductivity sections for the combined Lines 6 and 7a. There is no AEM data directly along these lines so, the AEM data in this figure are a series of eastwest 2D cross-section traverses taken from Roach (ed. 2015), with the spatial locations of the AMT stations overlain on these traverses. There are some obvious similarities between the two datasets in terms of crustal resistivity values and cover thickness. The highly resistive body (dark blue/purple colours) in the centre-left part of the AMT conductivity section is also present in the AEM traverses and likely represents the Granite Springs Granite (GSG) that is exposed in this locality. The moderately resistive (light blue colours) areas at depths >600 m in the western part of the AMT conductivity section are also present in the series of AEM traverses. These features could be due to the frequent elongate domains in the Werewilka Formation thought to represent metavolcanic intervals (see the basement geology map in Figure 1.3B). An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 15

21 Figure 2.3 Comparison of A) stacked east-west AEM conductivity sections from Roach (ed. 2015) and B) an AMT conductivity section along the combined Lines 6 and 7a (note that A shows the AMT stations locations along Lines 6 and 7a). Vertical to horizontal ratio = In both cases, blue and purple colours refer to electrically resistive bodies (e.g. felsic igneous bodies) and red and orange colours refer to electrically conductive bodies (e.g. Mesozoic and Cenozoic sedimentary basin cover). The sub-horizontal line overlain on the AMT section represents the DTB taken from the GABWRA dataset (black line; Ransley and Smerdon 2012) and augmented with the AEM DTB interpretation (white line) of Roach (ed. 2015). Vertical black lines (with black circles) denote the maximum cover thickness (with average) estimations based on high-resolution ground geophysical surveys on or near Lines 6 and 7a (Goodwin et al. in prep). For both datasets, solid vertical lines = surveys within 100 m of Lines 6 and 7a; dashed vertical lines = surveys within 2 7 km of Lines 6 and 7a. The black star indicates the location of known basement outcrops. Filled circles above the AMT section indicate AMT station locations. C) Interpretation of basement geology along Lines 6 and 7a from Purdy et al. (2014). 16 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

22 DTB estimates from the GABWRA dataset and AEM modelling along Lines 6 and 7a are also shown in Figure 2.3B. The general patterns of cover thickness are similar between these datasets and the AMT conductivity section. This is most obvious with the thinning of highly conductive cover towards the highly resistive body in the central part of the AMT conductivity section (presumably related to the outcrop of GSG), and thickening of cover to the margins of the section. The DTB inferred from AEM modelling (Roach ed. 2015) also shows a good correlation with the cover thickness estimated from the AMT conductivity section above the resistive GSG body. However, in general, the DTB estimates from the GABWRA dataset are much shallower than observed with the AMT conductivity section, and do not show any of the finer scale variations in cover thickness present with the AMT dataset. Cover thickness estimations from high-resolution ground geophysical surveys on or near Lines 6 & 7a (vertical black lines) are in closer agreement with thicknesses inferred from the AMT conductivity section than the GABWRA dataset. Figure 2.4 shows a comparison of AEM and AMT conductivity sections along Line 5. The highly resistive body to the centre-left (western side) of the AMT conductivity section (related to the Currawinya and Granite Springs granites) is not easily observed in the AEM traverses, likely due to the AMT stations not lying directly along these traverses and/or the DOI limitations of AEM for depths >150 m. The right-hand (eastern) side of the AMT conductivity section shows the top of a deeper (>800 m), resistive unit (~37 km along the AMT section) that is also observed in the AEM traverses at shallower depths (<150 m). This likely reflects the numerous igneous intrusive units in this area (e.g. the Cumbooromon Bore Granite; Figure 1.3B). There is an interesting feature centred at ~700 m depth, ~27 km along the AMT section that has a lower conductivity than the surrounding crust. However, this feature coincides with a gap in the AMT station coverage and accordingly should be viewed with caution. Figure 2.4B also shows a comparison of the AMT conductivity section with the DTB estimates from the AEM modelling (Roach ed. 2015) and GABWRA dataset along Line 5. As with the other traverses, the general patterns of cover thickness are similar between the different datasets, with thicker cover predicted in the west of the traverse and a shallower DTB towards the eastern end of the traverse. Appendix C shows a series of figures comparing AEM and AMT conductivity sections with GABWRA DTB modelling for Lines 7b, 8 and a short section of Line 3 around the town of Hungerford. These figures show similar results to the other lines, notably that: Electrically conductive cover is observed to thin above shallow resistive bodies (igneous intrusions) The AEM conductivity sections have difficulty in showing DTB > ~150 m The GABWRA and GSQ datasets often underestimate DTB compared to the AMT method Summary In summary, there is generally good agreement when comparing DTB and/or cover thickness of the different models and conductivity sections along the same traverses at these scales (10 s to 100 s km across the surface). However there are significant variations in the resolutions (mainly due to station spacing) and limitations in the DOI of the different techniques: Airborne electromagnetic data are very useful in inferring detailed DTB variations where the cover is less than ~150 m thick, but it cannot be reliably used when cover thickness is > ~150 m due to limitations in DOI (Roach ed. 2015). Broadband MT data are potentially very useful in delineating highly conductive cover from more resistant basement outcrops (especially igneous intrusive units) on regional scales (a few 100 km). However, data tend to overestimate cover thickness where it is less than 50 m, and due to the An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 17

23 wide station spacing (~5 km), produce coarse conductivity models that obscure some of the finerscale cover thickness variations. Because of the deep penetration depth of this technique, it is potentially very useful in imaging deeper parts of the crust and upper mantle. Audio MT data provide the best resolution for estimating cover thicknesses of m on a regional spatial scale of 10s to 100s of km, as they agree with other independent techniques for assessing DTB variations (e.g. high-resolution ground geophysical surveys), producing higher resolution conductivity sections than the BBMT data and more accurate cover thickness estimations where cover is known to be less than 50 m thick. Figure 2.4 Comparison of A) stacked AEM conductivity sections and B) an AMT conductivity section along Line 5 (note that A shows the AMT stations locations along Line 5). Vertical to horizontal ratio = In both cases, blue and purple colours refer to electrically resistive bodies (e.g. felsic igneous bodies) and red and orange colours refer to electrically conductive bodies (e.g. Mesozoic and Cenozoic sedimentary basin cover). The subhorizontal line overlain on the BBMT section represents the DTB taken from the GABWRA dataset (black line; Ransley and Smerdon 2012) and augmented with the AEM DTB interpretation (white line) of Roach (ed. 2015). The vertical black line (with a black circle) denotes the maximum cover thickness (with average) estimations based on a high-resolution ground geophysical survey ~7 km north of Line 5 (Goodwin et al. in prep). Filled circles above the AMT section indicate AMT station locations. C) Interpretation of basement geology along Line 5 from Purdy et al. (2014). 18 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

24 2.2 Forward Gravity Modelling The BBMT and AMT conductivity sections were used to inform the construction of 2D geological models that were tested via forward gravity modelling along Lines 1, 3, and the combined Line 6 and 7a. In particular, variations in cover thickness, large scale crustal architecture, and the presence and shape of the major resistive igneous bodies extending beneath the cover indicated in the conductivity sections were investigated. Forward gravity modelling (and subsequent inversions) aimed to: 1. Test the cover thickness variations predicted from AMT, BBMT and AEM conductivity sections 2. Investigate a realistic simplified large-scale crustal architecture for the region 3. Model the highly resistive bodies from AMT, BBMT and AEM conductivity sections that suggest igneous intrusions 4. Verify the basement interpretation of Purdy et al. (2014). As with many other geophysical modelling methods, this type of modelling contains an inherent problem of non-uniqueness (i.e. multiple models can produce calculated responses that match the observed data). In general, unrealistic models are those that are constrained by few (if any) geological or other geophysical datasets. In this Record, a combination of different geophysical and geological datasets and techniques have been used to identify any inferred geological commonalities, which is the basis for producing common forward models that satisfy a range of input parameters. This approach aimed to reduce geological ambiguity by: Producing cross sections that accurately reflect the interpreted lithological boundaries of Purdy et al. (2014) and available seismic interpretations, without introducing unnecessarily complex structures and/or lithology divisions Producing models that conform to interpretations of lithological thicknesses and divisions of the MT conductivity sections Producing models with realistic densities for lithological divisions based on known data and average textbook values for similar rock types Minimising modelled density variations between the same or very similar geological units. Changing the regional gravity response and/or background density may potentially alter the modelled densities of the cover and different rock units. To minimise the impact of this, throughout the new forward modelling, the background density was fixed at 2.78 g/cm 3 (close to the average full-crustal value in Christensen and Mooney 1995), except for Lines 6 and 7a where a value of 2.72 g/cm 3 was used to reflect the imaging of the upper 10 km of the crust using an AMT conductivity section along this traverse. Regional gravity responses were calculated from the Compute from Data function within ModelVision (see Appendix B). Density values were attributed to modelled bodies where known (as described in Section 1.3.8). Modelled bodies with unknown densities were given notional (but geologically reasonable) values which were then modified through the inbuilt inversion function in ModelVision to produce the closest match with the observed gravity response whilst reducing geological ambiguity as described above. Appendix B presents a more detailed description of the methods for forward modelling and gravity inversions (allowing modelled bodies densities to vary). The finalised ModelVision files for each forward model are available as an electronic supplementary dataset to this Record. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 19

25 2.2.1 Southern section of Line Previous modelling and interpretations by Glen et al. (2013) The southern part of Line 3 had an existing geophysical dataset and previous interpretations that could be tested. The existing dataset consisted of a deep crustal seismic reflection survey that coincided with the new gravity data acquired along the southern part of Line 3 (seismic lines TL1 and TL2 that ran either side of the Olepoloko Fault) see Figure 1.4. These seismic surveys had previously been interpreted by Glen et al. (2013) and were used as the basis of their forward gravity modelling of this area (Figure 2.5). Unfortunately, there was little information regarding the modelling methods or the models themselves presented in Glen et al. (2013), so it was decided to recreate their forward model in ModelVision. The subsurface structures, lithologies and densities in Figure 2.5 were used as inputs in recreating the forward gravity model of this data. When digitising the seismic interpretation, the top of the model was set to 0 m elevation with every 1 second of two-way travel time representing 3 km depth. The closest match using the Glen et al. (2013) data with the recreated forward gravity modelling in ModelVision is presented in Figure 2.6 and required a regional gravity field to be applied (a 1 st order polynomial) with a background density of 2.74 g/cm 3 (this is a comparatively low value when modelling the majority of the lithosphere (e.g. Christensen and Mooney 1995 quote an average full-crustal density value of 2.83 g/cm 3 ). The recreated modelling shows a good, although not identical match between modelled and observed gravity data. The StatWatch function in ModelVision provides a quantification of the match between the calculated (modelled) and the observed gravity responses using two metrics: RMS (root mean-squared) normalised by dynamic range (lower values indicate a better match between observed and modelled gravity responses) Correlation coefficient, where a value closer to unity (1) indicates a better match. For this model, the two parameters were RMS/range = and correlation coefficient = However, two different seismic interpretations contained in Glen et al. (2013) (Figure 2.7), and the recent basement interpretation geology map (Purdy et al. (2014) forced alternative interpretations of the deep crustal reflection seismic survey to be explored. Figure 2.5 Forward modelling of gravity data along the southern section of Line 3 by Glen et al. (2013). Image from Glen et al. (2013). Densities are in g/cm 3. The dashed box indicates the extent of seismic lines TL1 and TL2 (from Glen et al. 2013) that overlap with the new gravity data along Line 3 in this report (see Figure 1.4 for the location of the TL1 and TL2 seismic lines). 20 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

26 Figure 2.6 A) Recreated forward gravity modelling of the Glen et al. (2013) interpretation of the crust and upper mantle (B) beneath the southern section of Line 3 for which deep seismic reflection data is available. Densities of lithologies are identical to those in Figure 2.5 and are in g/cm 3. C) Basement interpretation from Purdy et al. (2014) Reinterpretation of seismic images and new gravity modelling Structures to the south of the Olepoloko Fault are well constrained based on seismic data. The Moho is easily identified, overlain by a layered crust with seismically distinct layers of the Lachlan Orogen dipping at a shallow angle to the north (Figure 2.7A). On the northern side of the Olepoloko Fault, however, the interpretation of the crustal architecture is less clear (e.g. Figure 2.7B and Figure 2.7C). The Moho is deeper (up to 50 km), with lower crustal packages dipping towards the south and a broad region in the middle crust with many internal variations that appear to arise from smaller-scale structural and lithological divisions. Various reinterpretations of the seismic data were performed (Figure 2.8) and assessed through forward modelling. In all of the interpretations, the interpreted structures and lithological boundaries are similar in the south of the section (left side), whereas the interpretations of the structures that make up the Olepoloko Fault as well as structures north of this fault vary. Forward models were constructed using the interpreted boundaries of lithologies from the three seismic interpretations shown in Figure 2.8 and the basement interpretation map of Purdy et al. (2014). Inversions were subsequently run that allowed the variation of unknown modelled bodies densities. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 21

27 Figure 2.7 A) seismic reflection image of TL1 and TL2 corresponding to the southern section of Line 3; B) seismic interpretation by Glen et al. (2013); C) alternative seismic interpretation by Glen et al. (2013) of the same section. Note the different interpreted dip angles of the Olepoloko Fault at depth between the two interpretations in B) and C). Although the models produce similar results, seismic interpretation 3 (Figure 2.8C) produces the closest match to the observed gravity response (Figure 2.9; lowest RMS/range = and the highest correlation coefficient = ). This suggests it may be more viable than the other models. The results show good agreements of the calculated gravity responses from the modelling with the observed gravity values. The inversion process produced densities for units (Figure 2.9B) that are reasonable based on the limited available rock property information for this region (AUSGIN; and Adams 2011) and average textbook values for different parts of the crust and upper mantle. The forward modelling results of seismic interpretations 1 and 2 are shown in Appendix F. 22 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

28 Figure 2.8 Different interpretations of the deep seismic reflection data of the southern section of line 3 (corresponds to seismic lines TL1 and TL2 of Glen et al. 2013). Vertical to horizontal ratio = One second of twoway travel time represents 3 km depth. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 23

29 Figure 2.9 A) Modelled gravity response based on the forward modelling of seismic interpretation 3 (from Figure 2.8C) in B). Densities (in g/cm 3 ) were varied using an inversion process in ModelVision; the blue dashed line shows the same forward model but with the densities of all units rounded to one decimal place; B) forward model based on seismic interpretation three. Vertical to horizontal ratio = 0.40; C) lithological boundaries along Line 3 from Purdy et al. (2014) Model three (from seismic interpretation three) consists of a thin crust (30-40 km) south of the Olepoloko Fault, split into a series of sub-horizontal layers dipping shallowly to the north near the fault extent at depth (Figure 2.9B). The crust is interpreted to be thicker (45-50 km) north of the Olepoloko Fault with a series of southerly-dipping units. This interpretation is similar to that of Glen et al. (2013) and Dunstan et al. (2016). The uppermost crustal unit (immediately below the cover) was split into lithologies with inferred faults using interpretations of Purdy et al. (2014) and Hegarty and Doublier (2015). There is a small amount of variation in the shallow basement densities (e.g. the Gumbalara Province shows the maximum variation from 2.68 to 2.74 g/cm 3 ) that could be explained by natural density variation in these units and/or artificial variation as a result of the inversion tool in ModelVision trying to find the best fit to the observed data. The forward models attribute average densities for each polygon, but this does not account for the large degree of complexity in natural geological systems (e.g. increasing density with depth within the same polygon). Whilst the polygons appear to be a similar lithology from the seismic interpretations, the modelling suggests they have varying densities or they could be different geological units. 24 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

30 To test the validity of this forward model, the densities of each of the polygons used in the model were rounded to one decimal place (the dashed line in Figure 2.9A). This helped to reduce the potential problem of different polygons that contained the same lithology but with different modelled densities. This forward model is clearly not an ideal match (RMS/range = ; Correlation Coefficient = ), but there is a general similarity in the pattern of the modelled gravity response and the observed gravity data, suggesting that the general structure and distribution of the crustal and upper mantle units in this model is realistic. In summary, the forward model based on seismic interpretation 3 (Figure 2.8C) produces the closest match with the observed gravity data (Figure 2.9). The reasonable densities calculated for the modelled units with unknown densities via the inversion process, and the match between the location of modelled units near the surface and those on the basement interpretation map, support this conclusion and thus this is the preferred 2D (or 2.5D) model based on interpretation of the seismic data Comparison of deep seismic surveys and Broadband MT The challenges in using different geophysical techniques that image different physical properties to infer subsurface geology and structures are highlighted in Figure 2.10, which shows a BBMT conductivity section and a deep crustal seismic reflection survey in the same region the southern section of Line 3. The Olepoloko Fault can be traced at depth on the seismic image, whereas it is more cryptic on the BBMT conductivity section. The Moho to the south of the Olepoloko Fault (i.e. part of the Northern Lachlan Oregon) is distinct on the seismic reflection image (Figure 2.10B). On the BBMT section (Figure 2.10A), the transition from the resistive region (>1000 Ωm) to the more conductive region (< 1000 Ωm) at ~30 km depth correlates with the position of the Moho, although it becomes less clear towards the Olepoloko Fault. The difficulties in using MT to reveal conductivity contrasts across the crust-mantle interface are well known, with the lower crust often producing a shielding effect on the resistivity of the upper mantle (e.g. Chave and Jones 2012). The BBMT section contains a highly conductive region (< 20 Ωm) from ~2-15 km depth (denoted by the red and yellow colours) that is not obvious or distinct in the seismic reflection image. This conductive region could be related to saline or freshwater-dominated fluids in the vicinity of the Olepoloko Fault, or alternatively along other smaller faults north of this major fault (as inferred from the seismic interpretations; Figure 2.8) As MT data are recording variations in electrical conductivity of the Earth s sub-surface, modelling will inevitably produce 2D cross-sections very different to the seismic method that measures the acoustic impedance of crustal materials. As more data from other geophysical methods as well as direct sampling from drilling becomes available for this region, it may consolidate interpretations as inputs into the modelling process can be constrained with more confidence. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 25

31 Figure 2.10 A comparison of A) the BBMT conductivity section and B) seismic reflection image for the southern section of Line 3 (western traverse; corresponds to seismic lines TL1 and TL2 of Glen et al. 2013). Vertical to horizontal ratio = The position of the Olepoloko Fault and inferred depth of the Moho are shown for reference. Inverted triangles above the BBMT section indicate BBMT station locations. For depth conversion of the reflection seismic, one second of two-way travel time is equal to 3 km depth Whole of Line 3 Due to the ambiguity of the seismic interpretations for the southern portion of Line 3 (especially north of the interpreted Olepoloko Fault), it was decided to simplify the structures and lithological boundaries for the middle to lower crust when performing forward modelling along the entire line for the following reasons: There is no deep seismic data for the region that comprises the central and northern parts of Line 3. The nearest deep-crustal seismic survey is the Eromanga-Brisbane Geoscience Transect that 26 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

32 runs in an east-west direction > 150 km to the north of the study region (Finlayson et al. 1990; Figure 2.11). AEM data are available along the majority of Line 3, but the conductivity sections extend to only ~400 m depth. BBMT data have been acquired along the majority of Line 3 and show features on a coarser (and deeper) scale than both AMT and AEM conductivity sections. It is therefore useful for larger, crustal-scale interpretations. The physical properties (e.g. density) of rocks that make up the majority of lithologies in the region (especially along the northern half of Line 3) are largely unknown. Estimations based on the inferred rock types from basement interpretations can be made from an average textbook value. A visual examination of gravity anomalies in the study area can produce interpretations of relative density variations between major lithologies, and running inversions of forward models may be able to help elucidate the likely range of density values for these lithologies. Figure 2.11 Interpretation of the southern Thomson Orogen by Finlayson et al. (1990) and Glen et al. (2013). Image from Glen et al. (2013). In light of this (and the available seismic interpretation for the southern part of Line 3), a simplified forward model was generated consisting of thickened crust north of the Olepoloko Fault composed of lower (27 to ~45 km depth), middle (12 to 27 km depth) and upper (2 to 12 km depth) layers (Figure 2.12D). Above this, the top two kilometres of the crust was modelled using the available data of: Basement interpretation (Purdy et al. 2014) Airborne Electromagnetic conductivity sections (shows depths up to ~400 m, but most reliable at depths <150 m) Broadband MT conductivity sections. The boundaries of major lithological units along Line 3 (from Purdy et al. 2014) were used as an input when drawing polygons to model these units in the sub-surface. Due to the dimensions at this scale, extreme vertical exaggeration means the boundaries between basement units are near-vertical on the forward model. For example in Figure 2.12E, the vertical to horizontal ratio (V:H) is The depth extent of cover was inferred from the MT conductivity sections along this line (see Section 2.1). Where outcrops are known to occur at the surface, cover thickness was set to zero, and away from these areas, cover thickness was forward modelled using a combination of the available AEM and MT conductivity sections (based on Figure 2.1). The BBMT conductivity section along this line shows regions in dark blue or purple (> 1000 Ωm) that indicate highly resistive lithologies. In many areas, these correspond to felsic igneous intrusions according to basement outcrops and the interpretation by Purdy et al. (2014) the Eulo and Hungerford granites are two of the best examples along this transect (Figure 2.1). From this inference, these highly resistive regions in the BBMT conductivity section have been used as approximate outlines of felsic igneous intrusions for the purposes of the forward model. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 27

33 Figure 2.12 A summary of the forward gravity modelling along Line 3 with a simplified (thick) southern Thomson crust north of the Olepoloko Fault. A) A comparison of calculated (blue line) and observed (black line) gravity data. B) Basement interpretation according to Purdy et al. (2014) along Line 3. C) BBMT conductivity section. Inverted triangles above the BBMT section indicate BBMT station locations. Vertical to horizontal ratio = D) and E) Forward models at different depth scales the red box in D) shows the position of the cross section in E) (vertical to horizontal ratio =0.036). Densities are shown in g/cm 3. The model and subsequent inversion (Figure 2.12) produces a generally good match between the modelled (blue line) and observed (black line) gravity data with a RMS/range = and a 28 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

34 correlation coefficient = Inversion produced a wide range of densities for lithologies with an unknown density (Figure 2.12D). For large-scale crustal units (Figure 2.12D), modelled densities are reasonable (e.g. upper mantle density = 3.21 g/cm 3, middle to lower crust = 2.86 g/cm 3 ). However, there are a number of problems relating to unrealistic densities produced by this forward model. These include: Unrealistic densities of some shallow basement units below the conductive cover (e.g. high-mag intrusive = 2.48 g/cm 3 ; Intermediate Igneous body = 3.4 g/cm 3 ). Unrealistic densities for basement units ~120 km along the transect (from the south). This relates to the surface position of the Louth Group and associated shallow intrusions interpreted from aeromagnetic imagery. The location of these units correspond with the pronounced negative gravity anomaly in this region (-100 gravity units) compared to the gravity highs (+100 gravity units) on either side (e.g. Figure 2.12A). One explanation for the gravity pattern in this region could be the presence of more extensive and/or deeper distributions of low-density (<2.6 g/cm 3 ) felsic intrusions than the forward gravity modelling currently predicts. Although this is not evident in the BBMT conductivity section, the ubiquitous presence of inferred igneous intrusions on and around where the Louth Group is thought to be present from the aeromagnetic imagery (Figure 1.3B; Purdy et al. 2014) may point to this hypothesis being valid. The presence of shallow intrusions in the forward modelling of the northern part of the southern section of Line 3, inferred from the seismic interpretation (Figure 2.8C; Section ) supports this hypothesis. The forward model of the northern half of the transect (~200 km from the southern margin of Line 3; Figure 2.12A) produces gravity responses with the poorest match to the observed data. This is likely due to the uncertainty of basement outcrop type and location, or uncertainty in the MT conductivity section in this complex part in this region. Interestingly, for the major intrusive bodies interpreted from the BBMT conductivity section, most produce modelled gravity responses that match with the observed gravity response, with corresponding reasonable densities (e.g. Hungerford Granite = 2.58 g/cm 3 ; Eulo Granite = 2.62 g/cm 3 ; Unknown Intrusion = 2.66 g/cm 3 ; Figure 2.12E). These results suggest that whilst there are other inaccuracies with this forward model, the locations of large igneous intrusive bodies predicted by the BBMT conductivity section seem realistic Line 1 This transect starts south of Gongolgon in NSW, heading north through the town of Brewarrina and then north-west towards Barringun on the border with Queensland. It then carries on in a north-west direction until it intersects Line 3 south-west of Eulo before heading WNW towards Thargomindah (Figure 1.3 and Figure 1.4). Many of the features of the Lachlan and southern Thomson lithospheric architecture along Line 1 are likely to be similar to those along Line 3. For example, the crust north of the Olepoloko Fault along Line 3 is interpreted to be thicker than the crust just south of the Fault this interpretation has been applied to the forward modelling of Line 1. The depth of cover along Line 1 has been inferred from the BBMT conductivity sections (Figure 2.2B), similar to the approach taken for Line 3. The Moho was modelled at a depth of 35 km for the northern Lachlan Orogen south of the Olepoloko Fault and 40 km for the southern Thomson orogen north of the Olepoloko Fault. Above the Moho, the crust is split into simplified horizontal layers consisting of: The lower crust extending up to km depth An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 29

35 The upper crust extending up to ~1.5 km depth. The uppermost part of the crust below the cover sediments was then split up into the major lithological units based on the basement interpretation of Purdy et al. (2014). These boundaries were then extended vertically downwards to ~1.5 km depth to reflect the vertical exaggeration (e.g. V:H = for Figure 2.13E) of this scale of the cross section, similar to the approach taken for Line 3. The BBMT conductivity section along this line shows some highly resistive bodies (> 1000 Ωm; purple and dark blue colours) at depths of up to ~30 km (Figure 2.13D). Some of these bodies correspond to intrusions extending to shallow depths such as the Brewarrina Granite and the Granite Springs Granite (the latter outcrops in the northern part of the section). In contrast, the two deeper, more extensive highly resistive regions in the centre of the transect, have no near-surface expression. These were good targets to include in the forward gravity modelling and subsequent inversion. Figure 2.13A shows the results of the forward gravity modelling and subsequent inversion (allowing unknown modelled bodies densities to vary) based on the interpreted crustal and upper mantle lithologies and structures (Figure 2.13C). Figure 2.13B shows the distribution of basement lithologies along Line 1 according to Purdy et al. (2014). The fit of the modelled gravity data to the observed gravity data gave a RMS/range = and a Correlation Coefficient = This is a relatively good match, although when the modelled rock densities are examined, there are some improbable and variable densities generated for some of the basement units such as the Werewilka Formation (2.77 to 2.99 g/cm 3 ) and the Gumbalara Province (2.14 to 2.86 g/cm 3 ; Figure 2.13E). This suggests that the model needs adjusting in one or more of the following ways: The large-scale modelled crustal layers are overly simplistic. It is likely the current horizontal layering is not applicable over the large scales covered in the modelling. These divisions in the crust are taken from the forward modelling along Line 3 and may not directly correlate with Line 1. The interpreted boundaries of the basement lithologies may be incorrect and need updating. The forward model is simplistic; for example, the interpreted basement lithologies from Purdy et al. (2014) are extended vertically downwards to ~1.5 km depth. In reality, the orientations of these boundaries are likely to be more variable. Faults, folds and shear zones in the region (other than the Olepoloko Fault) have not been included in the modelling. For example, Line 1 crosses the Culgoa Lineament (Hegarty and Doublier 2015) and the Louth-Eumarra Shear Zone (Dunstan et al. 2016). These are major structural features inferred from aeromagnetic imagery. Including structures such as these in forward models will likely affect any modelled gravity response. The shape and depth extent of many of the inferred igneous bodies in the BBMT conductivity section may not be a reflection of their true forms and these electrically resistive units can also produce a screening or shielding effect on the underlying crust (Berdichevsky and Dmitriev 2008). Owing to the nature of processing MT data, there is a lower confidence in these resistive bodies at depth as they extend downward (e.g. Chave and Jones 2012). Many of the modelled intrusive igneous bodies inferred from the BBMT conductivity section produce a good overall match with the observed gravity response and return reasonable densities (Figure 2.13C). This includes the Brewarrina Granite intrusion (towards the southern end of the transect) with a density of 2.65 g/cm 3 and the two larger unknown igneous intrusions in the centre of the transect with densities of 2.63 g/cm 3. These likely reflect felsic intrusions and could be related to the numerous granitic bodies interpreted to be a pervasive feature of the geology in this region by Purdy et al. (2014; Figure 1.3B). 30 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

36 As with the forward gravity modelling for Line 3, the forward model along Line 1 provides a good first indication of the large-scale crustal architecture of this region, and should be refined or improved upon when more data become available. Figure 2.13 A summary of the forward gravity modelling along Line 1 with a simplified (thick) southern Thomson crust north of the major Olepoloko Fault. A) The results of the forward modelling; B) The major lithological divisions according to the basement interpretation map (Purdy et al. 2014); C) Forward model of the crust and uppermost mantle (Vertical to horizontal ratio = 0.13). The red rectangle outlines the enlarged profile of the upper crust shown in E); D) BBMT conductivity section along Line 1. Inverted triangles above the section indicate station locations; E) Enlarged view of the forward model to a depth of 2.5 km (vertical to horizontal ratio = ). Lithology densities are in g/cm 3. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 31

37 2.2.4 Lines 6 and 7a (combined) Lines 6 and 7a run in an east-west direction in the northern part of the study area, overlapping with the north-western part of Line 1 (Figure 1.4). The depth of cover along these lines has been inferred from the AMT conductivity section (Figure 2.3), similar to the approach taken for Lines 1 and 3 (using BBMT conductivity sections). As there is no deep crustal seismic or BBMT conductivity section along Lines 6 and 7a, the crust was forward modelled to only 20 km depth (the extent of the upper crust ). This is reflected by the selection of a lower background density in this forward model (2.72 g/cm 3 ) than previous models (2.78 g/cm 3 ). Below the cover, the major lithological units based on the basement interpretation of Purdy et al. (2014) were distinguished. These boundaries were then extended vertically downwards to ~1.5 km due to the vertical exaggeration of this scale of the cross section (e.g. V:H = for Figure 2.14D), similar to the approach taken for Lines 1 and 3. The AMT conductivity section along these lines shows some more resistive bodies (>1000 Ωm; purple and dark blue colours) than the surrounding crust at depths of up to ~8 km (Figure 2.14E) and these are interpreted to correspond to igneous intrusive bodies. The major body near the centre of the section extends all the way towards the surface and corresponds to the known outcrop of the Granite Springs Granite (Purdy et al. 2014). In contrast, the resistive bodies to the east and west of the central granite do not seem to have any near-surface expression and hence represent good targets to test with forward gravity modelling and subsequent inversions. Figure 2.14A shows the results of the forward gravity modelling and subsequent inversion (allowing lithology densities to vary) based on the interpreted crustal units (Figure 2.14C). Figure 2.14B shows the distribution of basement lithologies along Lines 6 and 7a according to Purdy et al. (2014). The fit of the modelled gravity data to the observed gravity data gave a RMS/range = and a Correlation Coefficient = This is a relatively good match, although many of the basement units and inferred igneous intrusions return some variable densities after the forward modelling and inversion process. For example, the Werewilka Formation returns a variable density range of 2.47 to 2.63 g/cm 3, the cover sediments are predicted to have a higher than expected density of 2.31 g/cm 3 and the Nebine Metamorphics have a density of just 2.52 g/cm 3 (Figure 2.14D). Possible explanations for this could include: An incorrect background density selected for this region. A forward model that only extends to 20 km depth (due to the absence of BBMT for these lines, forward modelling is restricted to the upper portion of the crust). The interpreted boundaries of the basement lithologies may be incorrect and need updating. Further, as with the BBMT sections for Lines 1 and 3, the depth extent of many of the inferred igneous bodies is not well constrained using the AMT method, combined with these highly resistive bodies potentially producing a shielding or screening effect on the underlying crust (Berdichevsky and Dmitriev 2008; Chave and Jones 2012). However, even with these caveats, the overall match between the observed and calculated (modelled) gravity responses is good, with many lithologies returning realistic densities, providing a reasonable initial forward model able to be modified and further constrained by the collection of future geological, geophysical, geochemical, and petrophysical data. 32 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

38 Figure 2.14 A summary of the forward gravity modelling along the combined Lines 6 & 7a. A) The results of the forward modelling; B) The major lithological divisions according to the basement interpretation map (Purdy et al. 2014); C) Forward model of the crust down to 20 km depth (vertical to horizontal ratio = 0.16). The red rectangle outlines the enlarged profile of the upper crust shown in D); D) Enlarged view of the forward model to a depth of 8 km (Vertical to horizontal ratio = 0.064). Lithology densities are in g/cm 3 ; E) AMT conductivity section along Lines 6 and 7a. Inverted triangles above the section indicate AMT station locations Cover thickness variations tested by gravity modelling Lines 6 and 7a (combined) One of the aims of this Record was to examine whether variations in cover thickness (depth to basement) influence the gravity modelling. Lines 6 and 7a (combined) were used to test whether variations in cover thickness (depth to basement) produce discernible gravity variations in the forward modelling and/or whether this modelling can be used to predict cover thickness. The AMT conductivity An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 33

39 section along these lines (Figure 2.15A) produces a good estimation of cover thickness (corresponding to the highly conductive red region; <30 Ωm), particularly where it is known to be less than 50 m and/or where basement is known to outcrop. This is demonstrated in the centre-left of the AMT section (~40-55 km from the western margins of the section) where the resistive (blue) body extends to the surface and corresponds with the location of the Granite Springs Granite (Purdy et al. 2014). The forward model (with identical densities of basement units) of Lines 6 and 7a has been used (Figure 2.15) with only cover thickness varying. Figure 2.15B shows the results of forward modelling when the cover thickness varies according to the basement-cover interface interpreted from the AMT conductivity section. Figure 2.15C shows the results of the forward modelling when an averaged uniform cover thickness of 350 m is used. The correlation statistics show that there is a better match (a lower RMS/range and a higher correlation coefficient) of modelled and observed gravity data when a more accurate cover thickness is used (Figure 2.15; Table 2.1). Note that the densities of all units are the same in both models; it is only the thickness of cover that varies. The data show that using an accurate cover thickness constrained by an independent geophysical dataset (for instance, MT) produces a better match with the observed gravity data than when an averaged, uniform cover thickness is used. This can be seen with the Granite Springs Granite body that is known to outcrop along this Line. At this location, the modelled gravity response is much closer to the observed gravity response when an accurate modelled cover thickness is used (Figure 2.15B) than if an average cover thickness is applied across the model (Figure 2.15C). This is an important observation that shows cover thickness variations can produce discernible variations in gravity responses and need to be taken into account in gravity modelling. It also suggests that MT conductivity sections can reflect the natural variations in cover thickness/depth to basement. Table 2.1 Correlation statistics between calculated gravity data and observed gravity data of two different models of cover thickness along Lines 6 and 7a. Model 1, with varying cover thickness constrained by the AMT conductivity section produces a better match with the observed gravity data (lower RMS/Range and a higher correlation coefficient). Model RMS/Range Correlation Coefficient 1) Line 6 and 7a cover thickness constrained by AMT conductivity section 2) Line 6 and 7a uniform cover thickness of 350 m An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

40 Figure 2.15 Comparison of the agreement between calculated and observed gravity responses of two different models to represent cover thickness along Lines 6 and 7a. A) AMT conductivity section showing conductive (red) cover thickness varying along the section. Inverted triangles above the section indicate AMT station locations (vertical to horizontal ratio = ); B) Model 1, with varying cover thickness constrained by the AMT conductivity section; C) Model 2 with a uniform cover thickness (350 m). Lithology densities are shown in g/cm 3. Model 1 (a varying cover thickness) produces a better fit (a lower RMS/Range and a higher correlation coefficient); D) The major lithological divisions according to the basement interpretation map (Purdy et al. 2014) Other lines Forward models with varying cover thicknesses were also generated along the southern section of Line 3 (where seismic data is available), and the whole of Line 3. These data are shown in Appendix G. These lines show similar results to Lines 6 and 7a above, most notably that using an accurate cover thickness constrained by an independent geophysical dataset (MT) produces a better match with the observed gravity data than when an averaged, uniform cover thickness is used. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 35

41 3 Conclusions and recommendations New coincident gravity and electrical conductivity (AEM, MT and BBMT) data were collected along long traverses in the Southern Thomson region to more accurately assess cover thickness (or DTB) and refine forward gravity models. These electrical conductivity methods show great promise for estimating cover thickness, with cover sediments and rocks in overlying basins being identified by generally relatively high electrical conductivities. During this research, cover thickness estimations interpreted using MT methods (especially AMT) often agreed with other datasets such as existing mineral drill holes or water bores used for the GABWRA (Ransley and Smerdon 2012) DTB interpretation. The findings of this research confirm that within the region of the southern Thomson Orogen, AEM cannot be reliably used when cover thickness is > ~150 m due to limitations imposed by the excessive electrically conductive cover affecting the DOI, and BBMT tends to overestimate cover thickness where it less than 50 m because it is collected at too wide a station spacing to allow for highresolution modelling of cover conductivity. The interpretations of cover thickness along the traverses studied in this research are confirmed independently by cover thickness estimations obtained using high-resolution ground geophysical surveys (Goodwin et al. in prep). Audio MT likely provides the most suitable resolution for estimating cover thicknesses of m on the regional scale assessed in this research. The complimentary nature of the three electrical conductivity datasets means that the AEM can be used to assess cover thickness to within its DOI limits. In addition, with thickening cover, AMT data, and then BBMT data, can be used to assess cover thicknesses up to the scale of major sedimentary basins, e.g. ~10 km thick, and produce much more refined forward gravity models. These cover thickness estimations from this combined approach were tested with forward gravity modelling, which shows that a better match between the observed and calculated (modelled) gravity responses can be obtained using interpreted true cover thickness when compared to an averaged, uniform cover thickness. Seismic data reinterpretation along the southern part of Line 3, included in forward gravity modelling of this region showed that it is possible to produce a very close match between the observed and modelled gravity responses. A subsequent gravity inversion (allowing modelled bodies density to vary) produced densities of most crustal and upper mantle units that were reasonable given the known (and inferred) rock properties in this region. The AEM and MT conductivity sections along with interpreted basement geology (Purdy et al. 2014) added to the near-surface detail of the simplified 2D cross sections depicting different crustal architectures along a variety of traverses (Lines 1, 3, 6 and 7a). For all lines, the addition of realistic cover thicknesses to forward models produced gravity responses with a generally close match to the observed gravity data. However, in all models, there were varying numbers of crustal lithologies that returned unrealistic densities after the inversion process, suggesting the models were inaccurate in places or overly simplistic. Further, the forward modelling generated 2D (or 2.5D) cross-section models. The main limitation of 2D models is that is it difficult to project them away from isolated traverse lines to interpret regional geological and cover thickness variations on a regional scale. Combining numerous 2D models in 3D space or constructing 3D models based on the current available data and modelling should be a focus for the next stage of this work. Despite these limitations, the forward models provide a good initial evaluation of the crustal architecture and variations in cover thickness in the southern Thomson Orogen. Models with a simplified, thickened crust (~45 km) north of the Olepoloko Fault produce a close match of observed 36 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

42 and calculated gravity responses coupled with the most reasonable densities for crustal and upper mantle units. Further, the felsic intrusions inferred at depth from the highly resistive bodies in the MT conductivity sections produce accurate gravity responses and reasonable post-inversion densities. More accurate and realistic forward models should be generated through an iterative process as more data are acquired through targeted high-resolution pre-drilling ground geophysical surveys, stratigraphic drilling and analyses of the subsequent drill core material (e.g. geochronology, geochemistry, and petrophysics). In addition, more detailed modelling and inversions could be performed on selected key sections of the gravity datasets. This could be achieved by modelling individual anomalies in areas with more complex structures such as the Louth-Eumarra Shear Zone (Dunstan et al. 2016) and the Culgoa Shear Zone (Hegarty and Doublier 2015). Joint inversions of coincident gravity data with available MT and magnetic data may also help elucidate more complex regions. There are numerous advantages to conducting different geophysical surveys along the same traverses or in the same regions. Firstly, different geophysical methods (e.g. MT, seismic, gravity, and magnetics) are measuring different physical properties of the crust and can reveal different features and characteristics. This allows a comparison and integration of these methods, which can be augmented by geophysical modelling to provide a more robust understanding of crustal architecture and the nature of shallow cover sequences than if only one method is available or is relied upon. Secondly, when used in combination, these different geophysical methods can validate each other if the modelled outputs are similar, or alternatively, if they show something different, generate new interpretations of why these differences exist. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 37

43 4 Acknowledgements The author acknowledges the traditional owners and landholders within the Southern Thomson Project area, without whose cooperation, these datasets and this study could not have been completed. The gravity, MT and AEM surveys were planned with the help of the State Geological Surveys of New South Wales and Queensland. Ian Roach, Michael Doublier, James Goodwin and Peter Milligan are thanked for their reviews of this Record. Jingming Duan and Liejun Wang are thanked for providing copies of the AMT and BBMT 2D conductivity sections. James Goodwin is thanked for assistance with ModelVision and manipulation of the processed gravity data. Cameron Adams is thanked for providing access to additional density measurement data from northwest NSW. This record is published with the permission of the CEO, Geoscience Australia. 38 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

44 5 References Adams C Gravity Modelling of the Thomson Orogen, Northwest New South Wales. Honours Thesis. pp Department of Earth and Planetary Sciences, Macquarie University, Australia. Allpike R Southern Thomson Gravity Survey. Atlas Geophysics Report Number R Australian Academy of Science 2012 Searching the Deep Earth. Online: Berdichevsky M.N. & Dmitriev V.I Models and methods of magnetotellurics. Springer-Verlag, Berlin Heidelberg. Burton G New structural model to explain geophysical features in northwestern New South Wales: implications for the tectonic framework of the Tasmanides, Australian Journal of Earth Sciences, vol. 57, no. 1, pp Chave A.D. & Jones A.G The magnetotelluric method - theory and practice. Cambridge University Press. Christensen N. & Mooney W Seismic velocity structure and composition of the continental crust: A global view, Journal of Geophysical Research, vol. 100, no. B7, pp Dunstan S., Rosenbaum G. & Babaahmadi A Structure and kinematics of the Louth-Eumarra Shear Zone (north-central New South Wales, Australia) and implications for the Paleozoic plate tectonic evolution of eastern Australia, Australian Journal of Earth Sciences, vol. 63, no. 1, pp Fergusson C.L. & Henderson R. A Chapter 3 Thomson Orogen. In Jell P. A. ed. The Geology of Queensland. pp Queensland Government. Fergusson C.L. & Henderson R.A Early Palaeozoic continental growth in the Tasmanides of northeast Gondwana and its implications for Rodinia assembly and rifting, Gondwana Research, vol. 28, pp Finlayson D.M., Leven J.H., Wake-Dyster K.D. & Johnstone D.W A crustal image under the basins of southern Queensland along the Eromanga-Brisbane Geoscience Transect. In Finlayson D. ed. The Eromanga-Brisbane Geoscience Transect: a guide to basin development across Phanerozoic Australia in southern Queensland. pp Bureau of Mineral Resources, Geology and Geophysics Bulletin 232. Glen R.A The Tasmanides of eastern Australia. In Vaughan A., Leat P. & Pankhurst R. eds. Terrane Processes at the Margins of Gondwana. pp Special Publication of the Geological Society London. Glen R.A., Korsch R.J., Costelloe R.D., Poudjom Djomani Y. & Mantaring R Preliminary results from the Thomson-Lachlan Deep Seismic Survey, northwest New South Wales. In Lewis P. ed. Mineral exploration geoscience in New South Wales, Extended Abstracts, Mines and Wines Conference, Cessnock, NSW. pp SMEDG, Sydney. Glen R.A., Korsch R.J., Hegarty R., Saeed A., Poudjom Djomani Y., Costelloe R.D. & Belousova E Geodynamic significance of the boundary between the Thomson Orogen and the Lachlan Oregon, northwestern New South Wales and implications for Tasmanide tectonics, Australian Journal of Earth Sciences: An International Geoscience Journal of the Geological Society of Australia, vol. 60, no. 3, pp Goodwin J.A., Jiang W., Buckerfield S., Czarnota K., McAlpine S., Meixner T., Nicol M. & Crowe M. in prep Estimating cover thickness in the Southern Thomson Oregon - Results from the application of seismic refraction, passive seismic and audio-magnetotelluric methods. GA Record 2017/xx. Hegarty R. & Doublier M Defining major structures and their depth extent under cover in the southern Thomson Orogen, NSW. Proceedings of 24th International Geophysical Conference and exhibition, AGES. Perth, Australia. Murray C.G Metallogeny and tectonic development of the Tasman Fold Belt System in Queensland, Ore Geology Reviews, vol. 1, pp An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 39

45 Purdy D., Hegarty R., Doublier M. & Simpson J Interpreting basement geology in the southern Thomson Orogen. Proceedings of the Australian Earth Sciences Convention Newcastle, Australia: Geological Society of Australia. Ransley T.R. & Smerdon B.D Hydrostratigraphy, hydrogeology and system conceptualisation of the Great Artesian Basin. A technical report to the Australian Government from the CSIRO Great Artesian Basin Water Resource Assessment. pp CSIRO Water for a Healthy Country Flagship. Roach I.C. (ed.) 2015 The Southern Thomson Orogen VTEMplus AEM Survey: Using airborne electromagnetics as an UNCOVER application. Record 2015/29, Geoscience Australia, Canberra. Roach I.C., Blewett R., Czarnota K., de Caritat P., McPherson A.A., Meixner, A.J., Neumann N., Schofield A., Thomas M., Wilford J Regolith studies and the UNCOVER Initiative at Geoscience Australia. Fourth Australian Regolith Geoscientists Association Conference February, 2016, Thredbo, NSW. Rock Properties Explorer. Geoscience Australia Rock Properties data delivery website. Online: Simpson J. & Cant R Depth to basement calculation in southern Thomson, Queensland. ASEG Extended Abstracts 1, 1-4. pp Wang L., Chopping R. & Duan J Southern Thomson magnetotelluric (MT) survey report and data modelling. Record 2017/03. Geoscience Australia, Canberra. Withnall I.W., Golding S.D., Rees I.D. & Dobos S.K K-Ar dating of the Anakie Metamorphic Group: evidence for an extension of the Delamerian Orogeny into central Queensland, Australian Journal of Earth Sciences, vol. 43, pp An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

46 Appendix A Basement lithology descriptions Appendix A Figure 1 Basement interpretation map of Purdy et al. (2013) (as per Figure 1.3B). See Appendix A Table 1 for a description of basement lithologies. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 41

47 Appendix A Table 1 A description of the basement lithologies in the southern Thomson region and their colours/symbols on the basement interpretation map (Figure 1.3B; Appendix A Figure 1). Based on information by Purdy et al. (2014). Basement Unit Mulga Downs Group Warratta Group Cobar Supergroup Tongo Formation Gumbalara Province Colour on basement interpretation map (Purdy et al. 2014) Description Non-magnetic massive fluviatile sandstones: mainly flat lying but linear magnetic trends over some bed-dip ridges. Late Early Devonian to Late Devonian (fossil control and intrusive relations) Elongate zones having numerous narrow linear trends of moderate magnetic intensity; formations are interbedded sandstone, siltstone and mudstone, locally pyritic. Late Cambrian to Lower Ordovician (isotopic ages of tuff and intrusive re* (text missing) Metasedimentary sequences of turbidites and shelf facies which are typically non-magnetic. Late Silurian to Early Devonian (fossils and isotopic age control) Domain of uncertain age and association characterised by linear trends of moderate magnetic intensity; possibly magnetised metasedimentary rocks similar to the Nurri Group? No age control. Broad areas of low magnetic intensity interpreted as metasedimentary rocks. No age control (no samples) Louth Group Highly Magnetic Intrusion Non-magnetic intrusion Broad areas of complex trends having low to moderate magnetic intensity; siliclastic metasedimentary rocks, with intercalated mafic horizon flows, sills and dykes. Early Silurian (Provenance MDA) Areas of granites and intermediate intrusives having high magnetic intensity Areas of granites (both S-type and some I-types) having low magnetic intensity Torowoko Province (mafic) Schistose Zones Intermediate igneous Characterised by linear zones of high magnetic intensity sourced at depths of m, along strike from the Tongo formation; possible mafic horizons and/or increased magnetisation associated with fault zones. No age control Curved magnetic trends of low to moderate intensity adjacent to granite basement; mica schist, gneiss and minor granite. No age control Areas of moderate magnetic intensity: possible intermediate plutonic/volcanic unit. No age control Hungerford Granite Werewilka Formation Large, irregular to ovoid area of low magnetic intensity and generally low gravity; poorly defined, may comprise multiple intrusions. Outcrop comprises deformed, medium-grained, moderately porphyritic muscovite, biotite monzogranite.* (text missing) Elongate domains of distinctly stripy magnetic appearance related to narrow parallel zones of alternating high and low magnetic character; interpreted as a metasedimentary sequence. May appear as xenoliths within the Granite Springs Granite 42 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

48 Basement Unit Werewilka Formation (magnetic domains) Nebine Metamorphics Eulo Granite Girilambone Group Tarcoon Suite Colour on basement interpretation map (Purdy et al. 2014) Description Distinct elongate magnetic high domains within the Werewilka formation; may represent metavolcanic intervals within the overall metasedimentary sequence Broad area of low to moderate magnetic intensity, generally featureless but a few areas exhibit trends parallel to regional foliation/fabric. May correlate with drill holes in Nebine Ridge area - GSQ Mitchell 1 - relatively high * (text missing) Small area of low magnetic intensity with minor, slight magnetic high bands interpreted as a multiphase felsic intrusion; outcrop near Eulo comprises grey, medium-grained equigranular biotite granite ±2.5Ma Zone characterised by linear trends of low magnetic intensity reflecting structural fabrics; siliciclastic metasedimentary rocks. Lower to Middle Ordovician (fossil and isotopic control) Highly magnetic intrusion - Areas of granites and intermediate intrusives having high magnetic intensity. 416Ma Carrington Formation Narrow linear horizons of moderate to high magnetic intensity; includes magnetic mafic schist. No age control Brewarrina Granite Warraweena Volcanics Several elongate (faulted) zones of low magnetic intensity and low gravity values. Equigranular medium- to coarse-grained cordieritebiotite-bearing granite, S-type, with accessory apatite, opaques, zircons and common xenoliths. * (text missing) Zone of linear trends of moderate to high magnetic intensity; andesite, basaltic andesite. Poor age control - Proterozoic zircons recorded Culgoa Lineament Strathmore Formation Granite Springs Granite Dyke Narrow linear horizons of moderate to high magnetic intensity; includes magnetite quartzite, mafic schist, intermediate intrusive rocks, and ultramafic rocks. Lower Ordovician (Provenance MDA), Carboniferous (intrusive isotopic age) Broad, featureless area of low magnetic intensity; interpreted as metasediments. Possibly intersected by AOP Strathmore 1 - Interbedded pale green shale and fine-grained quartzose sandstone, poorly developed cleavage is parallel * (text missing) Relatively bland magnetic zone within and distinct from the strong Werewilka formation fabric; Crops out at Granite Springs on Werewilka station as coarsely porphyritic, foliated, biotite, muscovite, alkali feldspar granite to * (text missing) Normally-magnetised dyke, interpreted from aeromagnetic data An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 43

49 Appendix B Gravity modelling and inversion details The forward gravity modelling and subsequent inversions (allowing modelled bodies densities to vary) conducted in this study followed the process below: Ground gravity measurements were conducted by Atlas Geophysics Pty Ltd with a nominal station spacing of 333 m (Allpike 2014) using a Scintrex CG-5 gravity meter to collect the data. Gravity Data were imported along survey Lines 1, 3, 6 and 7a from GADDS (Geophysical Archive Data Delivery System; e). The station coordinates along the survey lines were converted into eastings and northings using a UTM projection using the GDA94 datum and MGA Zone 55. The station coordinates were converted into a straight line with equal spacing between stations. A Complete Bouguer Anomaly value was calculated at the stations along each line by adding the Terrain Correction (TC) to the Spherical Cap Bouguer Anomaly (SCBA) value (both included in the original dataset from GADDS). A complete set of the gravity data is included as an electronic supplementary dataset to this Record. These datasets were imported as ASCII files into ModelVision version in which the forward modelling was undertaken. Regional gravity fields were applied to each dataset by examining the general trend of the gravity anomalies to define a polynomial function calculated from the data using ModelVision. The most recent version of the Regional Gravity map of Australia was used when integrating regional gravity data. Regional gravity fields were applied using the Calculate from Data function in ModelVision and took the following form: Line 1: 1 st order polynomial Southern Section of Line 3: 1 st order polynomial Whole of Line 3: 6 th order polynomial Combined Lines 6 and 7a: 3 rd order polynomial Background images (e.g. seismic interpretation, MT conductivity sections) were imported as.bmp files into ModelVision. These formed the base layers for drawing polygons that represented subsurface rock units and cover sediments that would be tested by forward modelling. The top of the forward models were constructed according to the ground elevation where the (ground) gravity values were taken, except for the modelling along the southern section of Line 3 (with coincident deep seismic reflection data) where the top of the models are set to 0 m elevation. When digitising the seismic interpretation, every 1 second of two-way travel time represents 3 km depth using an average crustal velocity of 6 km/s. Density values, where known, were attributed to the polygons in the forward models (as described in Section 1.3.8, using the AUSGIN Geoscience portal and data contained in Adams 2011). Lithologies/polygons with unknown densities were given notional values (based on estimations of compositions of crustal rocks from averaged, textbook values). 44 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

50 These density values were then modified using the inbuilt inversion tool in ModelVision to produce the closest match with the observed gravity responses whilst reducing geological ambiguity as described in Section 2.2. The inversion process used the following run parameters: Maximum iterations = 40 Maximum data points = 2000 Maximum free parameters = 100 Free parameters (to vary) = property (density), regional level and regional slope Polygons in the forward models were given strikes of 50 km to allow projection of the 2D models into 3D space (creating 2.5D models). These models were extended at least 30 km beyond the extents of the gravity data boundaries to reduce edge effects of the modelling process during the forward gravity modelling. Quantification of the match between calculated (modelled) gravity data and the observed (actual) gravity data was performed in ModelVision using the StatWatch function. The two main metrics used were: RMS normalised by data dynamic range: a lower value indicates a better match Correlation coefficient: a value closer to unity (1) indicates a better match An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 45

51 Appendix C Comparison of AEM and AMT conductivity sections C.1 Line 7b Appendix C Figure 1 A comparison of A) stacked AEM conductivity sections and B) an AMT conductivity section along Line 7b (vertical to horizontal ratio = 0.04). In both cases, blue and purple colours refer to electrically resistive bodies (e.g. felsic igneous intrusions) and red and orange colours refer to electrically conductive bodies (e.g. sediments). Filled circles above the AMT section indicate AMT station locations. C) Shows an interpretation of basement lithologies along Line 7b from Purdy et al. (2014). 46 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

52 C.2 Line 8 Appendix C Figure 2 A comparison of A) stacked AEM conductivity sections and B) an AMT conductivity section along Line 8 (vertical to horizontal ratio = 0.035). In both cases, blue and purple colours refer to electrically resistive bodies (e.g. felsic igneous intrusions) and red and orange colours refer to electrically conductive bodies (e.g. sediments). Filled circles above the AMT section indicate AMT station locations. C) Shows an interpretation of basement lithologies along Line 8 from Purdy et al. (2014). An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 47

53 C.3 Line 3 (section with AMT) Appendix C Figure 3 Comparison of A) AEM and B) AMT conductivity sections along Line 3 (vertical to horizontal ratio = 0.078). In both cases, blue and purple colours refer to electrically resistive bodies (e.g. felsic igneous intrusions) and red and orange colours refer to electrically conductive bodies (e.g. sediments). The vertical black line (with black circle) denotes the maximum cover thickness (with average) estimations based on a highresolution ground geophysical survey within 100m of Line 3 (Goodwin et al. in prep). Inverted triangles above the AMT section indicate AMT station locations. C) Shows an interpretation of basement lithologies along Line 3 from Purdy et al. (2014). 48 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

54 Appendix D Depth to Basement estimates from other datasets Various datasets were used to compare Depth to Basement (DTB) (cover thickness) estimates with the Magnetotelluric conductivity sections. These are presented in Figure 1.5 and comprise: The GABWRA DTB contours dataset (Ransley and Smerdon 2012) The GSQ Drill hole and Waterbore DTB contours dataset (Simpson and Cant 2013) The AEM DTB modelling contained in the report by Roach (ed. 2015) Groundwater borehole and drillhole data from the Wanaaring Survey (Adams 2011) Cover thickness estimations from high-resolution ground geophysical surveys comprised of passive seismic, refraction seismic and AMT techniques (Goodwin et al. in prep). DTB data have been extracted from these datasets and is shown in numerous figures in Section 2.1. The process used to generate this data is outlined below: The GABWRA, AEM and GSQ DTB datasets were imported into ArcGIS. The locations of the MT stations used to generate the BBMT and AMT conductivity sections for selected lines were overlain on the above datasets. At each MT station, the corresponding DTB value was recorded from the DTB datasets DTB from the AEM models was given preference over the GABWRA and GSQ values. DTB values were interpolated where stations were situated between contours on the GSQ and GABWRA datasets. These DTB values were subtracted from the ground elevation to give terrain-corrected DTB values. These DTB values were then plotted and overlain on the MT conductivity sections. DTB values from groundwater boreholes and drillholes (Adams 2011) within 10 km of the MT survey lines were overlain on the MT conductivity sections (vertical black lines with horizontal bars at the base). Maximum and average cover thickness estimations based on high-resolution ground geophysical surveys conducted in late 2015 (Goodwin et al. in prep) were also overlain on the MT conductivity sections (where they were located within 10 km of the MT survey lines). These included the maximum (vertical black lines) and average (black circles) cover thickness values at each location. For the above two datasets, solid lines indicate boreholes/drillholes or ground geophysical surveys within 2 km of the MT survey lines and dashed lines indicate boreholes/drillholes or ground geophysical surveys within 2-10 km of the MT survey lines. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 49

55 Appendix E Available density data Appendix E Table 1 A summary of available density data in the southern Thomson region. Where multiple analyses have been collected at a particular location, the average densities are shown. The sources of samples are listed in Appendix E Table 2. Sample Longitude Latitude Type Unit Lithology # of Analy ses Dry Bulk Density (g/cm 3 ) Grain Mass Density (g/cm 3 ) Wet Bulk Density (g/cm 3 ) Outcrop Amphitheatre Group Outcrop Amphitheatre Group Metased. siliciclastic Sedimentary siliciclastic Weilmorin gle Borehole Coreena Member Sedimentary siliciclastic Weilmorin gle Borehole Doncaster Member Sedimentary siliciclastic Weilmorin gle 1A Borehole Hooray Sandstone Sedimentary siliciclastic Yantabull a Borehole Hooray Sandstone Sedimentary siliciclastic Yantabull a Borehole Wallumbilla Formation Sedimentary siliciclastic Outcrop Igneous Felsic Intrusive Outcrop Igneous Felsic Intrusive Outcrop Girilambone Group Outcrop Merre Conglomerat e Member Outcrop Mount Dijou Volcanic Member Metased. siliciclastic Metased. siliciclastic Igneous Mafic Volcanic Outcrop Nurri Group Metased. siliciclastic Outcrop Igneous Felsic Intrusive Eugene # Borehole Siltstone/shale Louth L Borehole Andesite? Louth L Borehole Siltstone/shale Mt Jack # Borehole Red Sandstone An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

56 Sample Longitude Latitude Type Unit Lithology # of Analy ses Pondie Range # Borehole Red Sandstone Dry Bulk Density (g/cm 3 ) Grain Mass Density (g/cm 3 ) SRRC Borehole Gabbro Wet Bulk Density (g/cm 3 ) Toorale 6/1-3 Toorale 6/ Borehole Andesite Borehole Andesite DM Wanaarin g # Borehole Pale Sandstone? 8 - DM Yantabull a #1 Yarralee #1 East Tibooburr a 1 East Tibooburr a Borehole Gabbro Borehole Siltstone/shale Outcrop Granodiorite Outcrop Granodiorite Appendix E Table 2 Sources for density measurements shown in Appendix E Table 1 Sample Source , Weilmoringle 1, Weilmoringle 1A Hawke DM Weilmoringle No 1 and 1A Completion Report - NSW GS1971_664 Yantabulla , , , , , , Eugene #1 Louth L1 Louth L5 Mt Jack #1 Pondie Range #1 SRRC2 Toorale 6/1-3 Toorale 6/7-1 Hawke Completion report - DM Yantabulla No. 1 - NSW GS1971_665 NSW GS2010/0173 NSW GS1965/174 NSW GS1965/175 NSW WCR069 NSW WCR123 NSW GS2000/129 NSW GS1979/110; GS1979/110 NSW GS1979/110; GS1979/109 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 51

57 Sample DM Wanaaring #1 DM Yantabulla #1 Yarralee #1 Source NSW GS1971/666 NSW GS1971/665; GS1971/001 NSW WCR153; WCR152 East Tibooburra 1 & 2 Adams (2011) Appendix E Figure 1 Map of the southern Thomson region (overlain on a greyscale 1 st Vertical Derivative Total Magnetic Intensity Reduced to Pole) showing the location of available density measurements. Yellow stars indicate data from Adams (2011); green stars indicate data downloaded from the Australian Geoscience Information Network (AUSGIN; Gravity and MT lines as per Figure An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

58 Appendix F Reinterpretation of seismic images and new gravity modelling The following figures are the results of the forward gravity modelling for seismic interpretations 1 and 2 as discussed in Section Appendix F Figure 1 A) Forward modelling of seismic interpretation 1 (from Figure 2.8A). Densities (in g/cm 3 ) were varied using an inversion process in ModelVision; B) lithological boundaries along Line 3 from Purdy et al. (2014). An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 53

59 Appendix F Figure 2 A) Forward modelling of seismic interpretation 2 (from Figure 2.8B). Densities (in g/cm 3 ) were varied using an inversion process in ModelVision; B) lithological boundaries along Line 3 from Purdy et al. (2014). 54 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

60 Appendix G Cover thickness variations tested by gravity modelling G.1 Southern Section of Line 3 The model constrained by seismic interpretation 3 (Figure 2.8C) has been used with only the cover thickness varying. Appendix G Figure 1B shows the result when the cover thickness (density of 2.11 to 2.12 g/cm 3 ) varies according to the MT conductivity section (Appendix G Figure 1A). Appendix G Figure 1C shows the result when an average, uniform cover thickness of 300 m is used. The correlation statistics (Appendix G Table 1) show that there is a better match between modelled and observed gravity data when a variable cover thickness based on the MT conductivity section is used (lower RMS/range; higher correlation coefficient). Note that the densities of all units are the same in both models (as per Figure 2.9B) - only the thickness of cover varies. Appendix G Table 1 Table showing the correlation statistics between calculated (modelled) gravity data and observed gravity data along the southern part of Line 3 of two different models to represent cover thickness (see Appendix G Figure 1 for the different forward models). Model 1, with varying cover thickness constrained by a BBMT conductivity section produces a better fit (lower RMS/Range and a higher correlation coefficient). Model RMS/Range Correlation Coefficient 1) Southern part of Line 3 varying cover thickness constrained by MT modelling 2) Southern part of Line 3 uniform cover thickness of 300m These data show that using an accurate cover thickness constrained by an independent geophysical dataset produces a better match in the gravity forward modelling than when an averaged, uniform cover thickness is used. It also suggests that the MT conductivity sections can reflect the natural variation in cover thickness/depth to basement. An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 55

61 Appendix G Figure 1 Comparison of the agreement between calculated and observed gravity data of two forward models to assess cover thickness. A) BBMT section showing cover thickness varying along the section (vertical to horizontal ratio = ). Inverted triangles above the section indicate BBMT station locations; B) Model 1, with varying cover thickness as interpreted from BBMT data; C) Model 2 with a uniform cover thickness (300 m). Both models have identical densities for all units (cover density = g/cm 3 ). A varying cover thickness produces a better fit (lower RMS/Range and higher correlation coefficient). D) Basement interpretation along Line 3 (based on Purdy et al. (2014). 56 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region

62 G.2 Whole of Line 3 Line 3 was used to test whether variations in cover thickness (DTB) produce discernible gravity variations in the forward modelling and/or whether this modelling can be used to predict cover thickness. A similar process that was applied to Lines 6, 7a and the southern portion of Line 3 was employed using the highly conductive (red) region on the MT conductivity section to constrain cover thickness. The forward model (with identical densities of basement units) of the simplified (thick) Thomson crust has been used. Appendix G Figure 2B shows the results of the forward modelling when an averaged uniform cover thickness of 250 m (densities of 2.13 to 2.31 g/cm 3 ) is used. Appendix G Figure 2C shows the results of forward modelling when the cover thickness varies according to the BBMT conductivity section (Appendix G Figure 2A). The correlation statistics (Appendix G Table 2) show that there is a better match (lower RMS/range; higher correlation coefficient) between modelled and observed gravity data when a variable cover thickness based on the MT conductivity section is used. Note that the densities of all units are the same in both models (as per Figure 2.12); it is only the thickness of cover that varies. As with Lines 6 and 7a and the southern section of Line 3, the data shows that using an accurate cover thickness constrained by an independent geophysical dataset (in this case, BBMT) produces a closer match with the observed gravity data than when an averaged, uniform cover thickness is used. As with the other Lines, it also suggests that the MT conductivity sections can reflect the natural variations in cover thickness/depth to basement. Appendix G Table 2 Table showing the correlation statistics between calculated gravity data and observed gravity data of two different models to represent cover thickness along Line 3. Model 2, with varying cover thickness constrained by the BBMT conductivity section produces a better fit (lower RMS/Range and a higher correlation coefficient). Model RMS/Range Correlation Coefficient 1) Line 3 uniform cover thickness of 250 m ) Line 3 cover thickness constrained by BBMT conductivity section An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region 57

63 Appendix G Figure 2 Comparison of the agreement between calculated and observed gravity data of two different models to assess cover thickness along Line 3. A) BBMT section showing cover thickness varying along the section (vertical to horizontal ratio = ). Inverted triangles above the section indicate BBMT station locations; black stars indicate known surface outcrops; B) Model 1, with a uniform cover thickness (250 m); C) Model 2 with varying cover thickness constrained by the BBMT data. Lithology densities are shown in g/cm 3. All other densities are the same between both models. Model 2 (varying cover thickness) produces a better fit (a lower RMS/Range and a higher correlation coefficient). 58 An integrative approach to investigating crustal architecture and cover thickness in the Southern Thomson region