The integration of regional reflection seismic profiles and gravity datasets with different spatial coverage associated with geological models

To enable the integration of reflection seismic and gravity datasets with different spatial coverage we explore methodologies drawn from implicit modelling and machine learning approaches. This thesis examines implicit modelling approaches in the form of Level set gravity inversion in different scenarios ranging from constrained inversion to the cooperative inversion of seismic and gravity datasets. These methodologies are tested on synthetic datasets and on two case studies in Australia. Finally, a deep learning approach is employed for predicting the main structural feature of the 3D geological models. These integration approaches produce 3D models compatible with geological modelling parameters and structures.