Geophysics

Our aims

Our global geophysics efforts are oriented towards integrating observations with models of plate tectonics, mantle convection and landscape evolution, which in turn are tied to basin-scale problems and resource exploration.

We unlock the potential of existing large-scale datasets through innovative machine-learning techniques, lead marine expeditions that collect new data in the sparsely sampled seas around the Australian continent, and develop key community infrastructure for modelling various aspects of Earth evolution and its uncertainty.

Our integrated effort places us in a unique position to analyse the history of continents and ocean basins through deep-time and to explore the feedback between processes occurring across scales, from the Earth’s interior to the surface. Our research feeds into our goal of improving resource exploration (Mineral and Petroleum Resources Research Group).

Our research

Global geophysics

Despite over 50 years of research in geodynamics, there are still many fundamental unresolved questions with regard to how the Earth has evolved since its accretion about 4.5 billion years ago.  When exactly did plate tectonics evolve, and what triggered its inception?  What exactly drives the plates and what determines the geometry of the plate tectonic system?  What controls the time-dependence of mantle convection?  How do solid Earth processes contribute to sea-level variations through time?

Our development of the open-source and cross-platform GPlates plate reconstruction software, funded through AuScope’s Simulation and Modelling Group, is the enabling engine of our efforts to build and improve global plate tectonic reconstructions that we use to constrain mantle convection models, e.g. CitcomS. This stable community software has been used for many years, by Sabin Zahirovic, Maria Seton, and Dietmar Müller and their collaborators, to investigate the connection between plate motions, subduction, and the evolving mantle structure, particularly aimed at understanding dynamic topography and abrupt changes in the plate-mantle system.  However, the increasing uncertainties in the geological record back in time make it difficult to constrain the evolution of the Earth before Pangea breakup.

This means we need to improve our understanding of the physics behind plate motions. To make progress in this direction, Claire Mallard and collaborators use recently-developed fully-dynamic global models using the code StagYY, which self-consistently generate Earth-like mantle currents together with plate-like surface tectonics. The virtual planets produced this way provide access to a range of different evolving parameters representing plate-mantle evolution. Our evolving 4D solid Earth models help us understand not only the evolution of the mantle and plate geometries through time, but also how convection influences vertical motions of the Earth’s surface.

Modelling sedimentary basin structure and evolution

Sedimentary basins capture Earth’s sea level, climate history, and the variation of the surface topography due to geodynamic, tectonic, and surface processes. We rely on sedimentary basins for oil, gas, geothermal energy, and water, and managing these competing uses requires a thorough understanding of the processes that form, deform, and fill the basins. As part of the ARC Basin Genesis Hub,  we simultaneously model deep Earth and surface processes, spanning continental to basin scales – the key to an integrated understanding.

Our major advance in software and workflow engineering opens emerging petascale computing resources to basin modelling and model-data assimilation, leading to a new generation of predictive 5D basin models of transformative significance for the oil and gas industry. Our surface process models using the Badlands software (https://github.com/badlands-model/pyBadlands), developed by Tristan Salles, depend on a range of parameters, including dynamic and isostatic topographic change, rock erodibility and precipitation through time. We explore parameter combinations to find the best-fit models constraining erosion and sediment accumulation, using a range of observations, and formally evaluate model uncertainty through a Bayesian statistical framework (Bayeslands) developed by Rohitash Chandra, Tristan Salles and their collaborators at the Centre for Translational Data Science at the University of Sydney.

Examples of how these models are being applied to the Australian continent include: an understanding of the interplay between the formation and disappearance of the Cretaceous Eromanga Sea and the subsequent uplift of the eastern highlands of Australia; the development of delta systems on the Northwest Shelf of Australia; and the dependence of sedimentation in the Gulf of Papua on upland erodibility.

Coupling lithospheric deformation and surface process models

Feedbacks between mantle flow, crustal deformation, erosion, sediment transport and deposition drive the structure and evolution sedimentary systems. Patrice Rey, Tristan Salles and Claire Mallard are applying a range of different 3D thermo-mechanical models for different tectonic contexts using the Underworld (http://www.underworldcode.org/) numerical modelling framework. 

It is now possible to link these models with surface process models in order to model basin stratigraphy via our Badlands software, coupled to Underworld. This 4D simulation of surface processes, enabled by a high-performance parallel computing approach, allows us investigate the effect of lithospheric rheology and extension speed/obliquity on the removal of up to several kilometres of material during rifting, as well as associated sedimentary deposits. These models are being applied to basins in a variety of tectonic settings around the world.

Our people

• Dr Adriana Dutkiewicz
• Dr Claire Mallard
• Professor Dietmar Müller
• Associate Professor Patrice Rey
• Dr Tristan Salles
• Dr Maria Seton
• Dr Sabin Zahirovic
• Dr Vasileios Chatzaras

Opportunities

For information about opportunities to study with us, contact Professor Dietmar Muller via Research Supervisor Connect.