Higher Degree by Research Application Portal

Project Overview

This PhD research extends proven machine learning approaches from gravity inversion to full tensor gravity gradiometry (FTG) and magnetic inversion with remanence characterization. Building upon successful integration of geological constraints using convolutional neural networks trained on Noddy-generated synthetic models, this project addresses critical challenges in modern geophysical exploration through advanced deep learning techniques.

Background and Motivation

Traditional geophysical inversions suffer from non-uniqueness and limited resolution, particularly with complex geological scenarios. Full tensor gravity gradiometry represents a quantum leap in gravity surveying, measuring all five independent components of the gravity gradient tensor rather than just vertical gravity. This provides superior resolution of geological structures and improved signal-to-noise ratios, but presents substantial computational challenges due to increased data dimensionality and complex inter-relationships between tensor components.

Magnetic remanence characterization is equally critical, as natural remanent magnetization often dominates magnetic anomalies and can completely mask induced magnetic signatures. Accurate separation of remanent and induced components is essential for proper geological interpretation and successful mineral exploration targeting.

Expected Outcomes and Significance

This research will deliver novel inversion algorithms demonstrating superior performance in resolving complex geological structures from FTG and magnetic data. The integration of physics-informed neural networks (PINNs) will embed gravitational field theory and magnetic potential equations directly into the inversion process, ensuring solutions remain grounded in established geophysical principles while leveraging machine learning capabilities.

This research addresses fundamental challenges in modern geophysical exploration by combining cutting-edge machine learning with rigorous geophysical principles. As the mining industry increasingly targets deeper, more subtle deposits in complex geological environments, these advanced inversion capabilities become essential for exploration success. The physics-informed machine learning approach ensures interpretable results that geologists and geophysicists can trust while providing practical tools to enhance mineral exploration effectiveness.

Dr Vitaliy Ogarko

vitaliy.ogarko@uwa.edu.au

Skills in geophysics and/or mathematics, and programming capability are essential.
Willingness to learn is a requirement. The capability to work in a team and independently and to take initiative will be appreciated.

The research will be based at the Centre for Exploration Targeting (School of Earth Sciences) at the University of Western Australia (Perth, Australia) within a dynamic team of students and researchers specialising in geophysics, mathematics, geology and statistics. The projects will be carried out in close collaboration with researchers from UWA and CSIRO. The PhD candidate will also have the opportunity to collaborate with researchers at the Loop consortium (https://loop3d.org/).

The Doctor of Philosophy (PhD) is a program of independent, supervised research that is assessed solely on the basis of a thesis, sometimes including a creative work component, that is examined externally. The work presented for a PhD must be a substantial and original contribution to scholarship, demonstrating mastery of the subject of interest as well as an advance in that field of knowledge. 

Visit the course webpage for full details of this course including admission requirements, course rules and the relevant CRICOS code/s.