The Need for Better 3D Conceptual Models of Aquifers in the Murray-Darling Basin.

There are many challenges ahead as we learn how to better manage the allocation and movement of water throughout the Murray-Darling Basin (MDB). A significant step towards better management is acknowledging the complexity of the alluvial aquifers and working out how to capture this complexity in our water management models. The New South Wales State government has extensive borehole data records. Driller logs can be interpolated to yield a comprehensive 3D faciesmodel. Geostatisticalmethods are often used for modelling facies(Deutsch 2002). The major steps are 1) establish large scale geological structures, 2) within each zone use object-based simulation to model palaeochannels(Keogh et al. 2007, Pyrczet al. 2009) 3) populate each faciesusing an appropriate geostatisticalor nonparametric classification technique (Dubois et al. 2007, Tartakovskyet al. 2007). Research within the NCGRT will be validating the appropriateness of existing algorithms and developing new methods for predicting faciesat locations where samples have not been taken. There are many advanced 3D geological modelling packages on the market. However, the algorithms behind many of the processes are not public, and the direction of development is driven by commercial interests. Students and researchers need an accessible 3D geological modelling environment to allow for the evolution of ideas and processes. Many of the pieces exist. There are scripting languages like Python, Maple, Mathematica, and Matlabthat have good database connectivity and come with high quality 3D visualisation tools. These scripting environments allow for the rapid construction of 3D geological structural models , and are ideal nonparametric modelling environments. Stochastic modelling of the faciescan be done using GSLIB or SGEMS. These models can then be imported into MODFLOW or FEFLOW. The skills required to use all these components at the level necessary to model our catchments has traditionally not been widely integrated into our University courses. These skills will be taught through the National Centre for Groundwater Research and Training. In Australia, the application of 3D conceptual aquifer models has been limited, compared to the level of adoption throughout Europe and North America. Through the development of accessible software and procedural documents it is hoped that the use of 3D geological conceptual models will be common for: - data integration; - conveying the complexity of alluvial aquifer systems throughout the MDB and other alluvial aquifers throughout Australia; - characterising contaminated sites; - developing framework models for input into groundwater flow modelling packages, and - communicating groundwater processes to all stakeholders. For the size of the country and the number of groundwater management issues Australia is confronting there are too few hydrogeologists. Through the new ARC/NWC co-funded National Centre for Groundwater Research and Training we aim to increase the number practitioners with the skills required to analyse, communicate and manage our groundwater resources. Providing accessible software is critical for the universal adoption and practice of constructing 3D geological conceptual models to advance the management of Australia’s aquifers.