Three-dimensional gravity models are a useful part of improving the geological understanding of large areas in various geological settings. Such models can assist seismic interpretation, particularly in areas of poor seismic coverage. In general, forward modelling and inversion are conducted until a single model is derived that fits well to the observed gravity field. However, the value of such a model is limited because it shows only one possible solution that depends on a fixed set of underlying assumptions. These underlying assumptions are not always clear to the interpreter and an arguably more useful approach is to prepare multiple models that test various scenarios under a range of different assumptions. The misfit between observed and calculated gravity for these various models helps to highlight flaws in the assumptions behind a particular choice of physical parameters or model geometry. Identifying these flaws helps to guide improvements in the geological understanding of the area. We present case studies for sedimentary basins off western Africa and western Australia. The flawed models have been used to rethink assumptions related to the geology, crustal structure and isostatic state associated with the basins, and also to identify areas where seismic interpretation might need to be revised. The result is a more reliable interpretation in which key uncertainties are more clearly evident.

Stations on the Australian continent receive a rich mixture of ambient seismic noise from the surrounding oceans and the numerous small earthquakes in the earthquake belts to the north in Indonesia, and east in Tonga-Kermadec, as well as more distant source zones. The noise field at a seismic station contains information about the structure in the vicinity of the site, and this can be exploited by applying an autocorrelation procedure to the continuous records. By creating stacked autocorrelograms of the ground motion at a single station, information on crust properties can be extracted in the form of a signal that includes the crustal reflection response convolved with the autocorrelation of the combined effect of source excitation and the instrument response. After applying suitable high pass filtering the reflection component can be extracted to reveal the most prominent reflectors in the lower crust, which often correspond to the reflection at the Moho. Because the reflection signal is stacked from arrivals from a wide range of slownesses, the reflection response is somewhat diffuse, but still sufficient to provide useful constraints on the local crust beneath a seismic station. Continuous vertical component records from 223 stations (permanent and temporary) across the continent have been processed using autocorrelograms of running windows 6 hours long with subsequent stacking. A distinctive pulse with a time offset between 8 and 30 s from zero is found in the autocorrelation results, with frequency content between 1.5 and 4 Hz suggesting P-wave multiples trapped in the crust. Synthetic modelling, with control of multiple phases, shows that a local Ppmp phase can be recovered with the autocorrelation approach. This approach can be used for crustal property extraction using just vertical component records, and effective results can be obtained with temporary deployments of just a few months.

As a result of work undertaken by Geoscience Australia during the Australian Government's Energy Security Program (2006-2011), data-poor and little-known frontier basins around Australia's continental margin are receiving increased scientific and exploration attention. Marine and airborne geophysical surveys conducted by Geoscience Australia along the eastern, southern and southwest margins of the Australian continent have yielded new aeromagnetic data, relatively closely-spaced ship-track magnetic and gravity data, industry-standard seismic reflection data and swath bathymetry data. Geoscience Australia's strategy for integrated geophysical interpretation and modelling includes: depth-to-basement determination using spectral and analytic-signal techniques applied to magnetic data; enhancement of aeromagnetic data to facilitate onshore-offshore geological interpretation; use of 3D forward and stochastic inverse modelling of gravity data to guide seismic interpretation of sediment thickness and basement structure; 3D inverse modelling of magnetic and gravity data to constrain the physical properties of the crust; and use of levelled ship-track magnetic and gravity data integrated with onshore data for multi-scale edge-detection analysis to guide interpretations of basement structure. However, Geoscience Australia's efforts to understand frontier basins are not without challenges. Our work highlights the lack of constraints on sub-basin crustal structure that leads to significant ambiguity when determining maximum sediment thickness and basement architecture. These deficiencies indicate a need for seismic refraction surveys that focus on sub-basin crustal structure. Refraction surveys should be complemented by airborne magnetic and gravity surveys that link onshore and offshore areas, and regional 2D seismic reflection surveys designed for deep sedimentary basins.

New compilations of levelled marine and onshore gravity and magnetic data are facilitating structural and geological interpretations of the offshore northern Perth Basin. Multi-scale edge detection helps the mapping of structural trends within the basin and complements interpretations based on seismic reflection data. Together with edge detection, magnetic source polygons determined from tilt angle aid in extrapolating exposed basement under sedimentary basins and, therefore, assist in the mapping of basement terranes. Three-dimensional gravity modelling of crustal structure indicates deeper Moho beneath the onshore and inboard parts of the Perth Basin and that crustal thinning is pronounced only under the outboard parts of the basin (Zeewcyk Sub-basin).

The Australian earth sciences have been recognized as part of Australia's key scientific capability to understand the structure and evolution of the Australian continent. Over the last five years, Geoscience Australia, through its Onshore Energy Security Program (OESP), in conjunction with the State and Territory Geological Surveys, the Predictive Mineral Discovery Cooperative Research Centre (pmd*CRC), the AuScope Earth Imaging (under Australian Government's National Collaborative Research Infrastructure Strategy) and the Australian National Seismic Imaging Resource has acquired over 6,500 line kilometres of new world-class seismic reflection data and over 3,700 kilometres of magnetotelluric (MT) data from more than 640 stations. Geoscience Australia acquires high quality deep seismic reflection data in most of Australia's economically significant geological regions, by collecting at least one deep seismic reflection traverse across the key structures. The acquisition parameters for regional vibroseis surveys have been selected from broad experience in hard rock environments and experimental programs prior to seismic acquisition. Three IVI HEMI-50 or 60 peak force vibrators are used with three 12 s varisweeps with 80 m between vibration points, 40 m group interval, and 20 s listening time to image down to approximately 60 km in depth. Geoscience Australia continues to provide expertise in deep crustal seismic reflection processing and mineral province interpretation to collaborative research programmes which focus on understanding the 3D crustal architecture and mineral systems within `hard-rock' mineral provinces. As part of this program , broadband and long period MT data have been acquired along 12 deep seismic reflection transects across potential mineral provinces and frontier sedimentary basins.

The New Caledonia Trough is a bathymetric depression 200-300 km wide, 2300 km long, and 1.5-3.5 km deep between New Caledonia and New Zealand. In and adjacent to the trough, seismic stratigraphic units, tied to wells, include: Cretaceous rift sediments in faulted basins; Late Cretaceous to Eocene pelagic drape; and ~1.5 km thick Oligocene to Quaternary trough fill that was contemporaneous with Tonga-Kermadec subduction. A positive free-air gravity anomaly of 30 mGal is spatially correlated with the axis of the trough. We model the evolution of the New Caledonia Trough as a two-stage process: (i) trough formation in response to crustal thinning (Cretaceous and/or Eocene); and (ii) post-Eocene trough-fill sedimentation. To best fit gravity data, we find that the effective elastic thickness (Te) of the lithosphere was low (5-10 km) during Phase (i) trough formation and high (20-40 km) during Phase (ii) sedimentation, though we cannot rule out a fairly constant Te of 10 km. The inferred increase in Te with time is consistent with thermal relaxation after Cretaceous rifting, but such a model is not in accord with all seismic-stratigraphic interpretations. If most of the New Caledonia Trough topography was created during Eocene inception of Tonga-Kermadec subduction, then our results place important constraints on the associated lower-crustal detachment process and suggest that failure of the lithosphere did not allow elastic stresses to propagate regionally into the over-riding plate. We conclude that the gravity field places an important constraint on geodynamic models of Tonga-Kermadec subduction initiation.

<p>The Roebuck Basin is considered a new and relatively untested hydrocarbon province in the central North West Shelf of Australia. Inconsistent results from drilling for hydrocarbons highlights the need to better understand the deep structures along this rifted margin that initially formed as an intra-continental, failed rift during Late Permian. Recent wells penetrated the previously unknown Lower-Middle Triassic fluvio-deltaic sedimentary package in the Bedout Sub-basin (inboard part of the Roebuck Basin), including intervals with major oil and gas discoveries. Another two wells, Anhalt 1 and Hannover South 1, only penetrated the top of this succession and they encountered volcanics in the Rowley Sub-basin (outboard part of the Roebuck Basin). Steeply dipping clinoforms observed in the seismic data in the Rowley Sub-basin have been interpreted either as a lava delta complex associated with a failed triple junction; or as a series of back-stepping, Late Permian carbonate ramps and banks, interpreted to have developed on a thermally subsiding rift flank. The implication for prospectivity between the two scenarios is significant. Geoscience Australia undertook a Triassic regional basin analyses, including potential field modelling to validate whether the two proposed models are a plausible solution. A combination of magnetic and gravity 2.5D modelling along nine key regional seismic lines, considered the distribution of potential intrabasinal volcanic rocks and the crustal structure, including Moho depth and depth to top crystalline basement. <p>New seismic interpretation correlated to recent wells, including 2D and 3D seismic reflection surveys was integrated with deep seismic reflection and refraction data resulting in an improved geometry and lithology model that was input into the potential field analyses. The results show that the combined Jurassic and Triassic successions reach up to 16 km deep in the central North West Shelf. The Lower-Middle Triassic sediment package in the Rowley Sub-basin correlates with up to 10 km of dense material (about 2.7 g/cm3 density) and contains magnetic features partially sourced from basalts at the top of the section, as intersected in Anhalt 1 and Hannover South 1. Combined with other causative sources within basement, the basalts correlate with a spatially large positive magnetic anomaly that extends north onto the Scott Plateau and into the Barcoo Sub-basin; in the offshore southwest part of the Browse Basin, where Warrabkook 1 intersected Late Jurassic volcanoclastics at its total depth. The presence of high density and high positive magnetic anomalies in the Lower-Middle Triassic and basement supports the presence of a large igneous province in this area. This interpretation in the outer Rowley Sub-basin downgrades the petroleum prospectivity in this area for this Lower-Middle Triassic interval. Petroleum prospectivity remains in the area due to the overlying sediments containing good source rocks which have been identified to have good to excellent generative potential. <p>The Lower-Middle Triassic sediment package in the adjacent northern Carnarvon Basin has been intersected only on the Lambert Shelf; encountering fluvio-deltaic sediments. In the offshore part of the northern Carnarvon Basin, the nature of this sediment package still remains enigmatic. It correlates with low density sediments (about 2.5 g/cm3 density) that include magnetic bodies on the outboard Exmouth Plateau. The basement and crust show crustal thinning with the presence of a thick layer of interpreted hyper-extended continental crust or exhumed lithospheric mantle. This crustal domain is overlain by thick onlapping Lower-Middle Triassic sediments which form a triangular shape depocentre in the inboard northern Carnarvon Basin, wrapping around the edge of the Pilbara Craton. The location of this initial thick sediment package suggests that it was controlled by the inherited thermal structure of the Late Permian-early Triassic rift architecture that is associated with some volcanics related to a large igneous province extending across the central North West Shelf. As described in the Rowley Sub-basin, the petroleum prospectivity of the northern Carnarvon Basin remains in the overlying sediments showing similar characteristics and indicating good to excellent hydrocarbon generative potential.

The Houtman Sub-basin is an under-explored region of Australia’s continental margin. It is located at the transition between the non-volcanic margin of the northern Perth Basin and the volcanic province of the Wallaby Plateau and lies adjacent to the Wallaby-Zenith Transform Margin (WZTM). In 2014, Geoscience Australia acquired new 2D seismic data (3300km) across the northern Houtman Sub-basin to better image deep crustal structures in this frontier province. Interpretation reveals that this depocentre contains up to 19 km of sediments and regional correlation of the seismic stratigraphy across the northern Perth Basin suggests this includes up to 16 km of Permian—Cretaceous succession. However, the depth and nature of the crystalline basement, the total crustal thickness as well as the extent and distribution of Seaward Dipping Reflector Sequences (SDR) and intra-basinal volcanics associated with development of the Wallaby Plateau volcanic province and the WZTM remain poorly constrained. An integrated geological and geophysical study, based on available seismic and potential field data was undertaken to aid the structural interpretation of the deep crust and Moho in order to better define the basin’s crustal architecture. In addition, the transition between non-volcanic and volcanic margin segments was delineated and, in conjunction with the regional seismic interpretations, better understanding of the timing, distribution and magnitude of multiple basin forming events was gained. The Ocean-Continent Transition (OCT) shows along strike and dip variations from extended and hyperextended (<5 km thick) continental crust beneath the main Permian depocentre to a zone of volcanic SDRs located outboard. Continental thinning and stretching phases occurred during both the Permian and Late Jurassic extensional phases. Volcanic margin development began in the Early Cretaceous, immediately prior to the separation of Greater India and Australia, suggesting that the volcanic margin experienced a phase of hyperextension before the magmatic break-up. Structural inheritance played an important role in basin development. It is likely that Early Permian graben formation was influenced by rheological contrasts in the underlying Proterozoic basement. The distribution of Permian rifts in turn further localised strain during Jurassic—Early Cretaceous rifting, strongly influencing the location and style of rifted margin development during Valanginian continental break-up.

Crustal magnetism is predominantly caused by the abundantly distributed ferrimagnetic mineral magnetite which posses the property of spontaneous magnetisation. Such magnetisation is dependent on temperature, which if high enough, will cause magnetite minerals to lose their magnetic property of spontaneous magnetisation and become paramagnetic. This temperature, known as the Curie point isotherm, occurs at ~580oC for magnetite. As temperature increases with depth in the crust, the Curie point can be taken as the depth at which the crustal magnetism ceases to be recorded. Using power spectral analysis of aeromagnetic data, we have generated a Curie point depth map for the Olympic Dam region in South Australia, host to the world's largest iron oxide-copper-gold-uranium deposit. The map shows an approximately 55 km long by 35 km wide and 40 km deep hemispherical depression in the Curie point depth beneath Olympic Dam, from a background average of around 20 km. Olympic Dam is notable for its large iron and uranium content, and it is located in a region of unusually high heat flow (av. 73 mWm-2). With such high heat flow one would expect the Curie point depth to be shallow. The paradox at Olympic Dam is that the Curie point depth is deep, raising questions about the geothermal gradient, depth-integrated abundance of heat-producing elements, and the source of the iron. A possible solution to the paradox is to interpret the deep Curie point depth as a giant hydrothermal alteration zone, where the heat-producing elements have been scavenged and concentrated into the upper crust, along with the gold and copper. The iron must have a significant mantle source as it is measured throughout the full crustal column. As iron is electrically conductive, such an interpretation is supported by the high conductivity measured deep beneath Olympic Dam.

Various aspects of isostasy concept are intimately linked to estimation of the elastic thickness of lithosphere, amplitude of mantle-driven vertical surface motions, basin uplift and subsidence. Common assumptions about isostasy are not always justified by existing data. For example, refraction seismic data provide essential constraints to estimation of isostasy, but are rarely analysed in that respect. Average seismic velocity, which is an integral characteristic of the crust to any given depth, can be calculated from initial refraction velocity models of the crust. Geoscience Australia has 566 full crust models derived from the interpretation of such data in its database as of January 2012. Average velocity through velocity/density regression translates into average density of the crust, and then into crustal column weight to any given depth. If average velocity isolines become horizontal at some depth, this may be an indication of balanced mass distribution (i.e., isostasy) in the crust to that depth. For example, average velocity distribution calculated for a very deep Petrel sedimentary basin on the Australian NW Margin shows no sign of velocity isolines flattening with depth all the way down to at least 15 km below the deepest Moho. Similar estimates for the Mount Isa region lead to opposite conclusions with balancing of average seismic velocities achieved above the Moho. Here, we investigate average seismic velocity distribution for the whole Australian continent and its margins, uncertainties of its translation into estimates of isostasy, and the possible explanations for misbalances in isostatic equilibrium of the Australian crust.