Improving Velocity Models for Microseismic Imaging

Microseismic imaging is a useful tool for monitoring migration of the CO2 plume and subsurface pressure changes during geologic carbon storage. An accurate velocity model is crucial for microseismic event location and focal mechanism inversion. For cost-effective long-term monitoring, the number of seismic stations are often limited, and their spatial distribution is usually sparse. To obtain an accurate velocity model for microseismic imaging, we develop a double-difference tomography method with a compressive sensing technique. The compressive sensing technique was developed to extract information from sparsely measured signals. We adapt this technique into double-difference tomography to alleviate inversion artifacts caused by the sparse distribution of seismic stations. We validate our new method using synthetic microseismic data and show that our new method significantly improves the accuracy of microseismic velocity inversion for a sparse seismic network.