In this presentation we describe analytical techniques that can be used to predict the magnitude and areal extent of surface deformation for the purpose of designing baseline geodetic monitoring networks. Furthermore, we discuss how this forward modelling capability can be combined with geodetic observations from radar satellites (InSAR) and Global Navigation Satellite Systems (GNSS) in an inversion to enhance our understanding of the geophysical properties of reservoirs or aquifers and improve future surface deformation predictions. This technique can be applied to a variety of geophysical studies including CO2 sequestration and groundwater withdrawal. In this presentation we demonstrate our capability with a case study of the extraction of Coal Seam Gas (CSG). We apply the analytical Finite Element Modelling (FEM) technique of Geertsma and Van Opstal (1973) to predict surface deformation in Australia. Simulation of subsidence associated with CSG extraction reveals the possibility of surface deformation due to the cumulative impact of multiple operators extracting large volumes of groundwater within a limited area (and volume). Our hypothesis is that the combined influence of groundwater drawdown associated with CSG extraction from adjoining leases could result in decimetres of subsidence depending on the properties of the reservoir, depth of the coal seams and extraction rates. Surface deformation of this magnitude has the potential to impact on infrastructure (e.g. buildings, houses, well structures, gas transport pipelines) and the environment (e.g. re-routing river systems, alteration of drainage on flood plains, inducing earthquakes, etc.) within and proximal to the areas of CSG operation. Our analysis suggests that surface deformation could occur at distances up to three times the radius of the drawdown zone, potentially resulting in subsidence in regions outside the leases of CSG operators. This highlights the need for predictive modelling to ensure that geodetic networks are designed to extend beyond the deformation zone. GNSS observations and InSAR time series analysis can be used: 1) to directly validate the surface deformation predictions, and 2) in an inversion to constrain geological properties of the reservoir. We apply the Neighbourhood algorithm of Sambridge (1999a, 1999b) to identify the optimal solution based on an ensemble of forward models. This technique allows us to constrain and improve our estimates of reservoir properties and allow for more accurate surface deformation predictions in the future.

Extension in granulite-facies orthogenisses of the Biranup Complex (Albany-Fraser Orogen), Bremer Bay, Western Australia

Marine Infrastructure links northern coastal areas with Asian markets and thus is fundamental for the development of Northern Australia. Marine science can play key roles in the planning and development of coastal infrastructure and optimisation of shipping operations that underpin growth and productivity. The Australian Institute of Marine Science (AIMS) operates two coastal observing systems in Darwin Harbour and the Beagle Gulf which supply information on wind, waves, tides and suspended sediments to Darwin Port in near real time. The systems also drive numeric models that tell us about how waves, tides and sediment circulate within the Harbour throughout the year. Adoption of the model outputs by industry has led to around a 50% improvement in ship accessibility to some infrastructure in Darwin Port. The modelling could be expanded to include other northern regions slated for infrastructure development, and is multi-purpose in the respect that it provides information that is crucial to understanding the impacts of dredging on water quality. Multibeam sonar acoustic seabed mapping creates spectacular three-dimensional images of the seafloor which increase the accuracy of models, and provide high quality information for navigation. Obstacles to shipping and pipelines such as reefs and sandbanks are clearly evident in these images, as are culturally significant features such as shipwrecks. A program to map Darwin and Bynoe Harbours is currently underway, and was made possible through offset funds provided by INPEX to Northern Territory Government Department of Land Resource Management, and co-investment from Geoscience Australia and AIMS. This comprehensive program has already revealed seabed features that were previously unknown. Moreover, it provides core data sets for Environmental Impact Assessments which reduce development costs to potential investors. The program highlights how industry and government can form effective partnerships to deliver products that will help with the successful sustainable development of Northern Australia.

This study reports the findings of salt store and salinity hazard mapping for a 20-km wide swath of the Lindsay - Wallpolla reach of the River Murray floodplain in SE Australia. The study integrated remote sensing data, an airborne electromagnetics (AEM) survey (RESOLVE frequency domain system), and lithological and hydrogeochemical data obtained from a field mapping and drilling program. Maps of surface salinity, and surface salinity hazard identified Lindsay and Wallpolla Islands, and the lower Darling Floodplain as areas of high to extreme surface salinity hazard. In the sub-surface, salt stores were found in general to increase away from drainage lines in both the unsaturated and saturated zones. Beneath the Murray River floodplain, salt stores in both unsaturated and saturated zones are high to very high (100 to 300t/ha/m) across most of the floodplain. Sub-surface salinity hazard maps (incorporating mapped salt stores and lithologies, depth to water table and the hydraulic connectivity between the aquifers), identify Lindsay and Wallpolla Islands; the northern floodplain between Lock 8 and Lock 7; and northern bank of Frenchman's Creek as areas of greatest hazard. Overall, the new data and knowledge obtained in this study has filled important knowledge gaps particularly with respect to the distribution of key elements of the hydrostratigraphy and salinity extent across the Murray River and lower Darling floodplain. These data are being used to parameterise groundwater models for salinity risk predictions, to recalculate estimates of evapotranspiration for salt load predictions, address specific salinity management questions, and refine monitoring and management strategies.

xx

Sonic drilling is a relatively new technology that was used successfully to obtain relatively uncontaminated and undisturbed continuous core samples with excellent (>99%) recovery rates to depths of 206m in unconsolidated fluvio-lacustrine sediments of the Darling River floodplain. However, there are limitations with the standard sonic coring method. Sands, in particular, are disturbed when they are vibrated out of the core barrel into the flexible plastic sampling tube. There can be changes to moisture content, pore fluid chemistry and sediment mineralogy on exposure to the atmosphere, even when the samples are processed and analysed soon after collection. The option exists during sonic drilling to encapsulate the core in rigid polycarbonate lexan tubes. Although this increases costs and reduces drilling rates, atmospheric exposure of the core during drilling is reduced to the ends of the lexan tubes before being capped. In addition, the tubes can be purged with an inert gas such as argon. Lexan coring is best carried out below the watertable as the heat from drilling dry clays can cause the polycarbonate to melt. In the study, 60 sonic holes (4.5 km) and 40 rotary mud holes (2 km) were obtained as part of a program to map and assess potential groundwater resources and managed aquifer recharge (MAR) targets over a large area (7,500 km2) of the Darling River floodplain. Two of the sonic bores were drilled to depths of 60 metres to obtain lexan-encapsulated core samples. These cores were used to obtain less perturbed samples for pore fluid analysis (salinity, major ions, trace metals, stable isotopes), textural analysis, and analysis of mineral phases to help assess aquifer clogging potential (using XRD, XRF, SEM). An additional advantage of the lexan coring was the recovery of encapsulated and intact sediment intervals for determining porosities, effective porosities, hydraulic conductivities, and other geophysical and petrophysical measurements. By painting some tubes black, sand samples were also successfully obtained for optically stimulated luminescence (OSL) dating. Alternatively, opaque black lexan can be made to order by the supplier. Overall, the superior sample integrity obtained from lexan coring enables a greater range of hydrogeological and hydrochemical parameters to be assessed.

The 1992 Flores earthquake (Mw=7.8) occurred on a back arc thrust near the island of Flores, Indonesia, causing widespread coastal uplift etc

xx

CO2 storage capacity and mechanisms are important considerations during exploration and characterisation of prospective CO2 storage reservoirs. Here, we use a combined numerical model and experimental batch reaction approach to determine mineral and fluid trapping capacity in the low-salinity (TDS<3,000 mg/L), deep (>800 m) Hutton and Precipice sandstones in the Surat Basin, Australia. CO2 solubility is calculated with an equation of state and water-rock-CO2 reactions are simulated using geochemical model GWB. Our initial model results indicate unusually high fluid and very low mineral trapping capacity in the Surat Basin. We find low-salinity formation waters in the basin have CO2 solubilities up to 1.3 mol/L, much greater than the fluid trapping capacity of more saline sandstone formations studied elsewhere. Mineral trapping capacity is much lower, however, due to low mineralisation of formation water and low quantities of soluble mineral phases in the prospective storage formations. Lab experiments using Surat Basin core samples and synthetic formation water are being run to validate model results and to determine reaction pathways under sequestration conditions. Initial experimental results support model estimates of CO2 trapping capacity, indicating limited water-rock-CO2 reactivity and pH buffering in the Precipice sandstone. Moderate reactivity and buffering is observed in the Hutton Sandstone, where carbonate mineral saturation is reached in several reactors following dissolution of albite, chlorite, and calcite.

A novel methodology has been developed to assess potential impacts of managed aquifer recharge (MAR) and/or groundwater extraction on groundwater dependent ecosystems (GDEs) in the Darling River floodplain. Potential negative impacts on GDEs include: (1) lowering of watertables in response to groundwater extraction, with the potential for enhanced leakage from lakes and rivers; and (2) raising of the watertables as a result of enhanced recharge. The latter has the potential to lead to saturation and/or salinisation of root zones through mobilisation of salts stored in the unsaturated zone. Depending on the magnitude of watertable fluctuations, type of vegetation, and degree of groundwater dependence, this can have significant implications for vegetation function. In this study, time-series Landsat data were used to identify groundwater dependent vegetation (GDV). Field-validated GDV maps were integrated with 3D sub-surface hydrogeological, hydrogeochemical and hydrodynamic data to characterise the hydrogeological system and understand controls on GDV distribution and condition. Particular attention was paid to the distribution and integrity of near-surface aquitards and identifying potential sites for inter-aquifer leakage and enhanced recharge. This paper reports on two sites: Jimargil and Lake Menindee. GDV at both sites (e.g. Eucalyptus camaldulensis (River Red Gum) and E.largiflorens (Black Box)), depends largely on groundwater in shallow unconfined aquifers. At the Lake Menindee site, remnant woody vegetation is concentrated along lake fringes, while the Jimargil site includes a narrow strip of riparian vegetation associated with the Darling River. At the Lake Menindee site there is a high degree of connectivity between the lake, shallow watertable, and target aquifer. Current management of lake levels impacts on GDVs at this site, and negative impacts are likely to be compounded by groundwater extraction and/or MAR. In contrast, connectivity between the shallow unconfined aquifer (and the Darling River) that sustain GDV communities at Jimargil, and the semi-confined target aquifer, is localised along faults and gaps in the intervening aquitards. Groundwater extraction and/or MAR in the deeper aquifer at the site would have minimal impacts on riparian and floodplain GDV.