Skip to main content
The steaming mouth of a volcano

Earth Systems research

Understanding Earth’s evolution
We are developing an experimental virtual planet to assimilate sparse geological observations into deep time Earth models to understand the intricate pathways in Earth’s evolution.

Our aims

Earth is unique in the solar system, forming a resource-rich, oxygenated habitable planet. Our civilization is predicated on stable climate and coastlines, yet the geological record reveals numerous episodes of major change, biological and chemical innovation, radiation and extinction, in the co-evolution of the solid Earth and habitable surface environments.  We need to understand the underlying mechanisms and feedbacks to predict possible future paths of the Earth System. Human civilisation equally depends on a range of metal resources to build renewable energy infrastructure. To identify portions of the Earth’s crust rich in these resources, we need to understand how bursts in mineralisation are related to the interplay between deep Earth processes, the crust and surface environments.

Our research

Rohitash Chandra, Adriana Dutkiewicz, Tristan Salles, Dietmar Müller (EarthByte Group), Jody Webster (Geocoastal Group), and their collaborators are using machine learning approaches to understand global processes from the sedimentation history of the ocean basins, and the connection of deep-sea sedimentation to ocean circulation, to local processes in basins and coral reefs.

The Sydney Informatics Hub and the Centre for Translational Data Science is also part of this research, the global research aspects of this initiative are focused on the synthesis of large sedimentological and geochemical datasets in order to provide new insights into fundamental processes underpinning the composition of marine sediments and ocean chemistry. Bayesian inference methods  are being used  to estimate model parameters and constrain the uncertainty in large-scale surface process models over geological time. This  requires major advances in the parallelisation for speeding-up of models that include sediment transport and carbonate growth. extension of Markov Chain Monte Carlo (MCMC) methods that implement Bayesian inference to cater for higher-dimensional parameter space and computationally expensive models using a high performance computing environment.

A major goal is to model Australia’s landscapes, basin sedimentation and carbonate growth through deep time with comprehensive uncertainty quantification that takes into account the  initial model boundary conditions, time-dependent dynamic and isostatic topographic driving forces, spatially-varying erodibility, spatially and temporally varying precipitation, and sea level fluctuation through time. In terms of ocean basin applications, our research is focussed on understanding both clastic and carbonate sedimentation, and their temporal and spatial variations and driving forces.

The paradigm of Earth System Science embodies the connections and feedbacks between the solid Earth and the hydrosphere, atmosphere, biosphere, and continental surface processes including chemical and physical weathering, erosion, sediment transport and sedimentation, particularly in terms of greenhouse gases. For example, sea level change affects continental silicate rock weathering, and the rate at which this process draws C02 out of the atmosphere, while sea-floor spreading and subduction processes drive the long-term degassing of the Earth, modulating the carbon cycle both via emission and sequestration of carbon. In turn these processes modulate the ability of corals and other marine calcifiers to store carbon in reefs and the deep sea.

Over long time spans, plate tectonics is the key component controlling the exchanges of fluids and volatiles between deep and shallow reservoirs. These exchanges include carbon (CO2, CaCO3, or a range of other molecular forms), H2O, and other volatiles to form physical Earth systems, that in addition to the biosphere, collectively contribute to the life-support mechanisms that make our planet unique in the Solar System. Sabin Zahirovic, Adriana Dutkiewicz, Dietmar Müller and colleagues in the Deep Carbon Observatory, are using the EarthByte Group’s global tectonic models to investigate fluxes of inorganic and organic carbon along divergent and convergent plate boundaries, and those related to weathering and erosion on the continents in relation to carbon sequestration in the marine environment, through deep time.

This project leverages existing and emerging open source tools such as GPlates and other community Earth-modelling platforms, including geodynamic models. This international community effort has the potential to revolutionise our understanding of the deep-time planetary carbon cycle by combining cutting-edge ‘big data’ and modelling approaches.

The topography of the continents and bathymetry of the ocean basins is a first-order earth observation, largely shaped by plate tectonics. Although we know the Earth’s present day elevations quite well from mapping and GPS measurements, reconstructing the past elevations of the continents and bathymetry of the oceans is more problematic as we need to account for episodes of continental collision and separation, large-scale volcanism, dynamic topography, erosion, sedimentation and the advance and retreat of ice-sheets. Maria Seton, Sabin Zahirovic, Dietmar Müller and colleagues have been developing state-of-the-art global topography and bathymetry models of the Earth from the Proterozoic to the present day using the power of GPlates and pyGplates.

Working with palaeoclimate modelling experts, they link their models to a range of ocean circulation models (both high- and low-resolution) to explore the connection between Earth’s changing surface and abrupt oceanographic and climate shifts. A key area of research is to investigate the role of oceanic gateways (narrow passageways facilitating the exchange of water, nutrients, salinity and species between ocean basins) in modulating Earth’s climate at key periods during the planet’s transition from a “Greenhouse” to “Icehouse” World. For example,  whether the opening and closure of oceanic gateways is the primary control on abrupt climate shifts or whether changes in atmospheric CO2 concentrations dominate.

Punctuations in Earth’s steady states are expressed  in surface environmental crises, burst and biological evolution as well as of mineralization. In collaboration with our national and international partners, we are developing a new approach to integrative Earth System Modelling aimed at quantifying Earth’s long-term evolutionary cycles and their driving forces to understand crustal and mineral evolution through time in a global palaeogeographic context.

Our objective is to determine first-order relationships between mineralization style and tectonic, magmatic, geodynamic settings and surface environments. This will be made possible by linking mineral systems to specific tectonic settings and geodynamic boundary conditions using a combination of geochemical, isotopic, structural and geophysical signatures by close examination of case studies, where the tectonic environment is well-known, but also through numerical modelling approaches. Igneous rocks provide records of past events that are crucial for understanding the magmatic and hydrothermal systems associated with many ore types. As part of the first research team to enunciate an Ore System approach, Wyman developed what would come to be known as the “orogenic model” for vein gold deposits in Archean terranes and pioneered the use of computer-based plate reconstruction software in studies of porphyry deposit distribution.

Our GPlates and pyGPlates software can be used together to integrate time-dynamic isotopic maps within a reconstructed plate tectonic framework over a billion years and beyond to develop time-dependent digital maps of isotopes and major/minor element concentrations that will track deep crustal structure and mantle input over space and time. Similarly, structures controlling mineral deposits can be reconstructed to their orientation and kinematics during periods of known mineralisation. Our integrative approach also allows us to combine tectonic and geodynamic models with erosion, surface drainage and material dispersion numerical models to understand deep weathering, regolith remobilisation and relief inversion in “stable” terrains. Long-wavelength lithospheric deformation driven by tectonic and mantle convection forces results in regolith (surface weathering) and landform evolutionary paths that have resulted in the formation of some of the world’s largest regolith-related iron and aluminium deposits.

Our people

  • Dr Rohitash Chandra
  • Dr Adriana Dutkiewicz
  • Professor Dietmar Müller
  • Associate Professor Patrice Rey
  • Dr Tristan Salles
  • Dr Maria Seton
  • Associate Professor Jody Webster
  • Associate Professor Derek Wyman
  • Dr Sabin Zahirovic


For information about opportunities to study with us, contact the researchers above.