Resolving multidisciplinary challenges of seamless integration of data on people, business, and the environment through applying a Discrete Global Grid System Framework

Crossman Shane1, Bell, Joseph1, Bastrakova, Irina1

1Geoscience Australia, Canberra, Australia

Increasingly, crisis situations require the ability to rapidly integrate diverse data from many sources: this is essential for timely and effective delivery of complex solutions to enable effective decision and policy making. Solutions commonly have to simultaneously cover multiple diverse use cases (e.g. social, environmental and economic) and be transparent, verifiable and trusted. The 2020 Australian bushfire crisis and the global COVID-19 pandemic are examples of these complex crisis events.

The unifying factor for these events is location: everything is happening somewhere at some time. Inconsistent representation of location (e.g. coordinates, statistical aggregations and descriptions) and the use of multiple techniques to represent the same data creates difficulty in spatially integrating multiple data streams often from independent sources and providers. A Discrete Global Grid System (DGGS) is a developed through OGC emerging cutting-edge technology. It provides a common framework capable of integrating very large, multi-resolution and multi-domain datasets together, and is a very efficient way of handling multiple data streams. The DGGS is changing the way how spatial data are enabled leading to an endless range of diverse and powerful data integration possibilities.

The DGGS aims to greatly increase the amount of location intelligence by flexibly linking big and small data in multiple formats, types and structures, and provides a framework for quick, reliable, repeatable, reusable infrastructure and codes. It fosters cross community collaboration and facilitates quick responses to stakeholder needs and for multiple use cases.

This paper will outline how Geoscience Australia and its partners implement the DGGS to address cross-portfolio needs around location-based data to provide a consistent way for seamless integration of data on people, business, and the environment.  Two use cases will be highlighted:

  1. Providing a good analytical basis to understand the environmental health issues such as human vulnerability during extreme natural events such as heatwaves; and
  2. Allowing for rapid and repeatable analysis of cross-portfolio information and decision making in response to devastating events of the 2020 Australian Bushfires.


Shane Crossman has worked in the Geospatial industry for the past 25 years working collaborative between Commonweath and State governemnts to build foundation spatial data.

Assimilation-Calibration Technique for Integrating Satellite Data with Hydrological Models

Khaki, Dr Mehdi1

1University Of Newcastle, Newcastle, Australia

Hydrological models are also crucial for simulating water storage and water fluxes. They reflect our knowledge of hydrological processes and are often used to study the impact of climate variability and changes on the Earth system. The ability of the models to accurately simulate and more importantly (reliably) predict water storage and fluxes highly depends on their parameters. This study applies an innovative sequential data assimilation technique to enhance a hydrological model’s simulations using multivariate satellite remote sensing while calibrating its parameters. The assimilation-calibration approach is based on the recently proposed data assimilation method, the Unsupervised Weak Constrained Ensemble Kalman Filter (UWCEnKF), extended to calibrate model parameters simultaneously with the state. The filtering method applies the joint state-parameter filtering problem, which results in a dual-type filtering scheme separately updating the state and parameters using two interactive Ensemble Kalman Filters (EnKFs). Numerical experiments are designed over globally distributed basins to test the proposed approach over two periods: calibration from 2003 to 2013 and prediction from 2013 to 2016. Results indicate that simultaneous assimilation of multiple satellite products combined with parameter estimation improves model predictions compared with single satellite products and/or state estimation alone. This improvement is achieved during both experiment periods, e.g., ~32% groundwater RMSE reduction over the calibration period and soil moisture correlation increase from ~0.69 to ~0.78 during the prediction period.


After my PhD graduation in 2018 from Curtin University, I started working as a lecturer at the University of Newcastle (Australia). My research is centred about the application of geodetic and remote sensing techniques and their integration with available models to improve their simulations at various scales.

Predicting Seafloor Lithology with a Neural Network Based Classification Model

Dietmar Müller1, Adriana Dutkiewicz1, Joyce Yu2

1EarthByte Group, School of Geosciences, University of Sydney, Sydney, Australia, 2CSIRO, Data61, Sydney, Australia

Deep-sea sediments record changes in climate, paleogeography, surface environments and biogeochemical cycles through geological time.  Yet, their composition and distribution through time is not well known due to the sparse distribution of ocean drilling sites. We have developed a neural network-based methodology that allows us to produce deep sea lithology maps from sparse samples in combination with other features that can be reconstructed through geological time.  These include bathymetry and features derived from paleo-climate models including sea surface temperature, salinity and upwelling intensity as a proxy for productivity. We test this model based on a present-day lithology, bathymetry and oceanography data set to map seafloor lithology with a probabilistic multi-class neural network model trained with about 12,500 sample points. We reduce seafloor lithologies to  five end-member classes:  (1) gravel/sand/silt, (2) clay, (3) calcareous ooze and other fine-grained calcareous sediments, (4) radiolarian ooze and (5) diatom ooze.  K-fold cross validation techniques were adopted to avoid overfitting. A final accuracy score of 86% was achieved in performance for our classification model. We compare this with a previously applied support vector machine (SVM) approach.  When applied to our five end-member classes its classification accuracy is 79%. Unlike the SVM, our neural network model also provides probability estimates, which helps quantify the level of uncertainty in predicting each lithology. Our method represents a low-cost way of obtaining lithology and paleo-lithology maps, and has the potential to be extended to produce sediment accumulation rate maps, allowing an analysis of the deep ocean sediment flux through time in a context of changing paleogeography and climate.


Dietmar Müller is Professor of Geophysics at the School of Geosciences, University of Sydney. He received his PhD in Earth Science from the Scripps Institution of Oceanography in 1993. With his EarthByte research group he combines tectonics, geodynamics, and sedimentology to build virtual Earth models from global to basin scales.

The AuScope Geochemistry Network and AusGeochem

Dalton, Hayden1, Prent, Dr Alexander2, Boone,Dr Samuel1, Florin, Dr Guillaume3, Greau, Dr Yoann3, McInnes, Professor Brent2, Gleadow, Professor Andrew1, O’Reilly, Professor Suzanne3, Kohn, Professor Barry1, Matchan, Dr Erin1, Alard, Dr Olivier3, Rawling, Dr Tim4, Kohlmann, Dr Fabian5, Theile, Moritz5, Noble,Dr Wayne5

1School of Earth Sciences, University of Melbourne, Melbourne, Australia, 2John de Laeter Centre, Curtin University, Perth, Australia, 3Department of Earth and Planetary Sciences, Macquarie University, Sydney, Australia, 4AuScope, Melbourne, Australia, 5Lithodat Pty Ltd, Melbourne, Australia

In 2019, AuScope, in response to a national expression of need for better organisation and coordination of geochemistry laboratories and data, established the AuScope Geochemistry Network (AGN). The AGN aims to foster and coordinate a national geochemistry laboratory infrastructure that incorporates earth science institutions across Australia in order to solve the national challenges of today and tomorrow. The AGN’s goals include (but are not limited to): i) promotion of capital and operational investments in new, advanced geochemical infrastructure; ii) endorsing existing geochemical capability and supporting increased end user access to laboratory facilities across Australia; iii) fostering collaboration and professional development via online tools, training courses and workshops; and iv) developing and maintaining a FAIR Australian geochemistry data ecosystem capable of hosting a diverse suite of geochemistry and geochronology data (AusGeochem). The AGN is led out of Curtin University with partner ‘nodes’ currently comprising The University of Melbourne and Macquarie University. The AGN actively encourages all Earth Science institutions from government, academia and industry, to register their interest in becoming a data contributing partner of the network and collaborate towards a national geochemistry infrastructure.

The AGN and collaborator Lithodat are making significant progress towards the goal of developing the AGN’s data repository and platform, AusGeochem, to become the interface between the institutional, collaboration and public domains, facilitating laboratory data upload and dissemination. Using AusGeochem, institutes and geoscientists will be able to upload, disseminate and publish their datasets while maintaining data privacy control and plot and synthesise their data within the context of a wealth of publicly funded geochemical data aggregated by all data contributing partners.

The AGN is working with a number of Expert Advisory Groups (EAGs) to build common technique specific interlaboratory metadata templates and data models, currently for SHRIMP U-Pb, LA-ICP-MS U-Pb and Lu-Hf, Ar/Ar, fission track and (U-Th-Sm)/He, with expansion to more data types on the horizon. Comprising geochemical specialists from across Australia, the EAGs are providing invaluable advice regarding data reporting best practices, data quality assessment and visualisation tools to be incorporated into AusGeochem. The AGN has also been teaming up with the Australian Research Data Commons (ARDC) to integrate International Geo Sample Number (ISGN) minting capabilities into AusGeochem, allowing users to simultaneously mint their samples when uploading their data into the platform.

The AGN plans to grow its network and to continue engagement with the geoscience community through its monthly webinar series and hosting national workshops to share best practice and foster new collaborations across Australia. 


Hayden Dalton is presenting on behalf of the greater AuScope Geochemistry Laboratory Network. Hayden is a PhD researcher in the School of Earth Sciences at the University of Melbourne, his research focuses on the geochronology and geochemistry of kimberlites.

Mapping subglacial sedimentary basin distribution in Antarctica using Random Forest method

Li, Lu1, Aitken, Dr Alan1, Lindsay, Dr Mark1, Jessell, Dr Mark1

1The School of Earth Sciences, The University Of Western Australia, Perth WA, Australia

Antarctica preserves the largest ice sheet in the world, which has a potential contribution to future sea-level rise up to 60m. Understanding subglacial sedimentary basin distribution is essential for studying ice sheet behaviour, as it forms an important basal boundary condition for ice sheet dynamics. It also records the geological history for tectonic evolution and past ice sheet behaviour. However, the subglacial sedimentary basin distribution is poorly known in Antarctica. A map of sedimentary basin distribution is a prerequisite to improving understanding of current and past ice sheet behaviour, aiding to project future ice sheet change and sea-level rise.

In this study, we present a sedimentary basin distribution likelihood map for Antarctica using the supervised machine learning method Random Forest. We apply this to generate a model based on the current understanding of Antarctica bedrock type distribution. We label the sedimentary basin and crystalline basement distribution driven from sparse rock outcrops, seismic imaging, and potential field data interpretation. Evidence layers are chosen from the available continental-scale geophysical datasets. By applying variable importance selection, we remove the unimportant and highly correlated evidence layers. After that, a strata sample process ensured a balanced class of each bedrock type during the training process. A sedimentary basin distribution map is then spatially predicted. The model accuracy is evaluated based on block cross-validation to overcome the underestimate of prediction error in the spatial correlated geophysical data.

Our results confirm the existence of previously documented subglacial sedimentary basins in Antarctica, and in general define the margins and extents of sedimentary basins in more detail. Specifically, in West Antarctica Rift System, the model delimits boundaries between sedimentary basins and volcanic rocks. We find a potential sedimentary basin preserved in Byrd Subglacial Basin, which is highly likely contributing to the currently fast flow in Thwaites Glacier. Further, our model shows more widely distributed subglacial sedimentary basins in East Antarctica than been previously recognized. Properties of geophysical and remote sensing data in Recovery Glaciers suggest a high probability of sedimentary basin preservation.


Lu Li is a PhD student at The University Of Western Australia to study geophysics. His PhD project is using geophysics to understanding lithosphere structure of Antarctica, and its influence on cryosphere/solid-earth interaction.

Bayesian inversion of 3D groundwater flow within the Sydney-Gunnedah-Bowen Basin

Ben Mather1, Dietmar Müller1, Craig O’Neill2, Louis Moresi3

1EarthByte Group, School of Geoscience, The University of Sydney, Camperdown, NSW, Australia, 2Department of Earth and Environmental Sciences, Macquarie University, North Ryde, NSW, Australia, 3Research School of Earth Sciences, Australian National University, Canberra, ACT, Australia

In the driest inhabited continent on Earth, aquifers of the Sydney-Gunnedah-Bowen Basin are essential for Australian agriculture production, yet they experience progressively declining water level trends. In addition, groundwater discharge from the basin into the coastal ocean, a process now widely recognised as being important for providing significant inputs of nutrients and solutes to the oceans, has never been modelled. We have constructed a 3D Bayesian numerical groundwater flow model spanning the entire width and depth of this continent-scale basin. Our model assimilates groundwater recharge rates from water chloride concentrations, and borehole temperature measurements to constrain hydrothermal flow within the basin. We show that inland aquifers exhibit slow flow rates of 0.5 cm/day, resulting in a groundwater residence time of approximately 383 thousand years. In contrast, coastal aquifers have flow rates of approximately 30 cm/day, and a groundwater residence time of just 182 years. Our open-source modelling approach can be extended to any basin and help inform policies on the sustainable management of groundwater. In the future, our approach will enable time-dependent modelling of groundwater flow in response to uplift, erosion and climate change.


Ben is a postdoc with the EarthByte Group at the University of Sydney. He is interested in coupling geophysical and geochemical observations with numerical solid Earth models to constrain crustal evolution. Ben is aiming to apply for a DECRA in 2021.

About the GSA

The Geological Society of Australia was established as a non-profit organisation in 1952 to promote, advance and support Earth sciences in Australia.

As a broadly based professional society that aims to represent all Earth Science disciplines, the GSA attracts a wide diversity of members working in a similarly broad range of industries.

Conference Managers

Please contact the team at Conference Design with any questions regarding the conference.
© 2020 Conference Design Pty Ltd