The lower Darling Valley contains Cenozoic shallow marine, fluvial, lacustrine and aeolian sediments including a number of previously poorly dated Quaternary fluvial units associated with the Darling River and its anabranches. New geomorphic mapping of the Darling floodplain that utilises a high resolution LiDAR dataset and SPOT imagery, has revealed that the Late Quaternary sequence consists of scroll-plain tracts of different ages incised into a higher more featureless mud-dominated floodplain. Samples for OSL (Optically-Stimulated Luminescence) and radiocarbon dating were taken in tractor-excavated pits, from sonic drill cores and from hand-auger holes from a number of scroll-plain and older floodplain sediments in the Menindee region. The youngest, now inactive, scroll-plain phase, associated with the modern Darling River, was active in the period 5-2 ka. A previous anabranch scroll-plain phase has dates around 20ka. Indistinct scroll-plain tracts older than the anabranch system, are evident both upstream and downstream of Menindee and have ages around 30ka. These three scroll-plain tracts intersect just south of Menindee but are mostly separated upstream and downstream of that point. Older dates of 50 ka, 85 ka and >150 ka have been obtained from lateral-migration sediments present beneath the higher mud-dominated floodplain. Establishing a chronology for the Quaternary fluvial landscape has been important for groundwater investigations in the Darling River floodplain area. More specifically, this has assisted in constraining the 3D mapping of floodplain units, helped constrain conceptual models of surface-groundwater interaction, and aided in the assessment of managed aquifer recharge options.

The Georgina-Arunta 3D project is a body of work that is predominantly focussed around delivery of 3D geophysical and geological data and interpretations to support the seismic data in the region. By integrating all available data in the region with a wide variety of potential-field techniques, a robust 3D map is able to be produced.

Advances in computer technology have provided the opportunity to present geoscience information in new and innovative ways. The use of web-based three-dimensional interactive models, animations and fly-throughs significantly enhances our ability to communicate complex geometries and concepts not only to the geoscientific community but also, just as importantly, to the general public. Projects within Geoscience Australia currently use a range of GIS, remote sensing, and modelling packages for visualisation of fundamental and derived data. In the main each of these packages also has the ability to produce, as an output, some form of model or animation sequence displaying the results of the visualisation. In most cases however, these outputs are generally not of sufficient quality or do not provide adequate functionality without further processing or editing. Geoscience Australia has adopted a multi-disciplinary approach to 3D visualisation encompassing cartography, GIS, remote sensing, graphic design, programming, web, and video editing to the post-processing of these visualisations. This paper examines the benefits of using models and movies for the visualisation of geoscience and briefly discusses the current workflows and presentation techniques used by the Geo-Visualisation team within Geoscience Australia.

The Radiometric 3D Atlas is a series of interactive X3D Models that can be viewed in your web browser. The Atlas consists of an overall model of Australia and eight detailed regional models from each state and territory. Each model includes; images from the Radiometric, Magnetic Anomaly and Gravity Anomaly data sets; a digital elevation model; coastline, cities/towns, state borders, mines; and 1:250 000 topographic map index. Software required Geoscience Australia's X3D and older VRML models require the free plugin BS Contact and work best with the web browser Internet Explorer version 6 or higher. More information about the plugin is available from the <a href=http://www.ga.gov.au/resources/multimedia/about-3dmodels.jsp>About 3D Modelling and Required Software</a> page. Size Approximately 163 MB for all the models. Startup download is 8.52 MB - the remaining datasets download when selected.

This third edition preliminary three dimensional model has been constructed from themes compiled from a variety of sources and assembled primarily within ESRI and GoCAD applications. The display medium for web delivery has used the Virtual Reality Modelling Language (VRML) format. Geophysical modelling was done by Geoscience Australia geophysicists using data stored by GA. Interpreted geology images of the Tanami and Arunta were provided by the Nothern Territory Geological Survey. Cross-sections were geophysically modelled using ModelVision, with geological interpretation provided by the NTGS and imported into GoCAD to build three dimensional fault surfaces. This edition of the model incorporates magnetic and gravity inversion surfaces and a depth to magnetic source layer.

Seismic line 07GA-GC1, described here, forms part of the Isa-Georgetown-Charters Towers seismic survey that was acquired in 2007. The seismic line is oriented approximately northwest-southeast and extends from east of Georgetown in the northwest to south of Charters Towers in the southeast (Figure 1). The acquisition costs for this line were provided jointly by the Geological Survey of Queensland and Geoscience Australia, and field logistics and processing were carried out by the Seismic Acquisition and Processing team from Geoscience Australia. Seven discrete geological provinces have been interpreted on this seismic section (Figure 2). Two of these, the Abingdon and Sausage Creek Provinces, only occur in the subsurface. The upper crustal part of the seismic section is dominated by the Etheridge and Cape River Provinces, but the seismic line also crossed the Broken River Province and the Drummond and Burdekin Basins.

3D visualisation of the Mount Isa Crustal Seismic Survey

The Australian Government formally releases new offshore exploration areas at the annual APPEA conference. In 2012, twenty-eight areas in nine offshore basins are being released for work program bidding. Closing dates for bid submissions are either six or twelve months after the release date, i.e. 8 November 2012 and 9 May 2013, depending on the exploration status in these areas and on data availability.

Extended abstract to accompany oral conference presentation. Full version of the short abstract (GEOCAT 70799).

Significant volumes of Big Lake Suite granodiorite intrude basement in the Cooper Basin region of central Australia. Thick sedimentary sequences in the Cooper and overlying Eromanga Basins provide a thermal blanketing effect resulting in elevated temperatures at depth. 3D geological maps over the region have been produced from geologically constrained 3D inversions of gravity data. The inverted density models delineate regions of low density within the basement that are inferred to be granitic bodies. The 3D maps include potential heat sources and thermally insulating cover, the key elements in generating an EGS play. A region was extracted from the Cooper Basin 3D map and used as a test region for modelling the temperature, heat flow and geothermal gradients. The test region was populated with thermal properties and boundary conditions were approximated. Temperatures were generated on a discretised version of the model within GeoModeller and were solved by explicit finite difference approximation using a Gauss-Seidel iterative scheme. An enhancement of the GeoModeller software is to allow the input thermal properties to be specified as distribution functions. Multiple thermal simulations using Monte-Carlo methods would be carried out from the supplied distributions. Statistical methods will be used to yield the probability estimates of the in-situ heat resource, reducing the risk of exploring for heat. A fast solver for the inhomogeneous heat equation in free space has been developed using Fourier domain techniques. Typical speed-ups for this strategy over the conventional solvers is better than 1000 to 1.