Summary
A large portion of the stress release on seismic faults remains silent and undetected, requiring the development of novel observation techniques. Measuring travel time perturbations from the correlation of ambient seismic noise at different stations is a well-known approach to assess temporal changes in seismic velocities, which can provide insights into hydrologic, tectonics, and volcanic dynamic processes. In this work, we study the specific case of a P-wave phase retrieved from the correlation of freight train noise in Southern California and evaluate its potential to detect localized velocity changes along the San Jacinto Fault. We use a full waveform modeling approach to simulate this P-wave interference and further assess its sensitivity to the position of the train source, near-surface velocity changes, and localized velocity changes in the fault zone. Our results show that the uncertainty in trains location can induce large travel-time biases which can be mitigated by averaging over many trains. Our results also highlight the weak sensitivity of these correlation P-waves to near-surface velocity changes, while they show significant sensitivity to localized changes at depth. This modeling highlights the potential of monitoring travel time perturbations of this ballistic P-wave interference to detect hidden slow-slip events on the San Jacinto Fault, particularly in identifying subtle velocity anomalies associated with fault zone changes that may otherwise go unnoticed by conventional seismic monitoring techniques.