Elastic Imaging and Full Waveform Inversion for Carbon Sequestration

Abstract

Geological sequestration of carbon dioxide (CO2) is essential to reduce the effects of global warming. However, this requires careful monitoring. Reflection seismics collects data from a large area and is therefore the most useful method for detecting CO2 migration. The existing imaging and Full Waveform Inversion (FWI) methods that produce subsurface images from seismic data are mathematically and computationally demanding. Improving these methods is, therefore, an important topic of research. I demonstrate that elastic ray-Born modelling can be used in multiparameter imaging and FWI for CO2 monitoring purposes. I propose different imaging and FWI methods and investigate which of these is the most effective in mitigating parameter crosstalk, a known problem in multiparameter inversion. FWI using the local optimisation method L-BFGS give the most accurate result. This method successfully mitigates crosstalk artefacts between the P- and S-velocity images and recovers the strong perturbations associated with the CO2 plume. The work of this thesis can serve as a starting point for a more sophisticated analysis of the imaging and FWI methods developed. Further research could focus on accelerating the computational speed, for example, using graphics processing units (GPUs). Moreover, the methods should be validated before being applied to real data, for example, by testing the methods' performance on synthetic data generated with elastic finite difference