Recent high-performance computing and machine learning approaches have improved our ability to detect earthquakes, but characterizing these earthquakes has proven to be more challenging. One desired characterization is determining earthquake focal mechanisms which inform us about fault structures, stress orientations, and rupture kinematics. However, these mechanisms typically can only be produced for well-recorded seismicity. I’ll talk about some machine learning techniques that we can use to increase the number of measurements, verify their results, and how we are developing software with these new workflows in mind.

Rob Skoumal is a research geophysicist at the Earthquake Science Center in Moffett Field, California. His research focuses on improving our ability to characterize both natural and induced (“anthropogenic”) earthquakes. He is particularly interested in developing new methods to characterize small magnitude earthquakes, distinguishing between natural/induced seismicity, and seismogenic fault mapping.