I ,A̶̲̥̅♏ a graduate of Physics.
Main choices will be around whether you are undertaking fieldwork of data, interpreting existing data, or creating a new technique that enables you to work with data in an interesting way.
Working with interesting data can be useful if you have access to decent software you can use, and/or can work with a locally based oil exploration company as part of your project.
Creating a new methodology technique is good if you are mathematically inclined and have some degree of experience in coding, software development and so on; to some extent this depends on what tools (eg MatLab) are available in your department also.
Fieldwork is more challenging if you want to create a project that has relevance on large-scale oil-industry project, unless you can find a way of making it relevant to the industry. So - for example, shooting a high resolution seismic survey over an area where you have a known reservoir formation outcropping in order to look at fractures, that kind of thing.
Techniques around localising/monitoring/imaging man-made seismicity from hydraulic fracturing are quite interesting at the moment.
A lot depends on where you are studying, what equipment/expertise is available in the department, what links their are to industry and what equipment you have available. Its a lot easier if you have some of thee already in place before you start your thesis work!