S017-05
Direct-waves passive monitoring of fault zones: processing strategy and source/structure sensitivity trade-offs
Wednesday, 9 December 2020: 04:18
Virtual
Aurélien Mordret1, Korbinian Sager2, Laura Pinzon Rincon3, Chloé Gradon3, François Lavoué4, Florent Brenguier5, Pierre Boué5, Yehuda Ben-Zion6 and Frank Vernon7, (1)ISTerre, Université Grenoble Alpes, Saint Martin d'Hères, France, (2)Brown University, Department of Earth, Environmental and Planetary Sciences, Providence, RI, United States, (3)ISTerre Institute of Earth Sciences, Saint Martin d'Hères, France, (4)Univ. Grenoble Alpes, ISTerre - Sisprobe, Grenoble, France, (5)Univ. Grenoble Alpes, ISTerre, Grenoble, France, (6)University of Southern California, Department of Earth Sciences and Southern California Earthquake Center, Los Angeles, CA, United States, (7)University of California San Diego, La Jolla, CA, United States
Abstract:
Monitoring faults before and during the nucleation of earthquakes might be the key to unlock the possibility of forecasting them. To do so, we need to probe the fault at seismogenic depths. This can be done in various ways, but the most logistically achievable is using surface sensors. Continuous monitoring, the only way to possibly detect any subtle signal from the fault, necessitates waves shot at regular time intervals, diving inside the fault from the surface. Here we use seismic waves generated by freight trains and recorded by two dense seismic arrays on both sides of the San Jacinto fault to reconstruct, by interferometry, direct P-waves propagating at depth between the two arrays.
To obtain stable waveforms, we detect and construct a catalog of trains that allows us to correlate only the right segments of seismic records. In doing so, we obtain extremely stable daily waveforms with a signal to noise ratio sufficient to reach a travel-time difference accuracy of few milliseconds between each day. We improve by more than 25% the accuracy of the travel-time measurements compared to a direct stack of all the available noise.
However, using moving sources to monitor structural changes via interferometry, adding the constraint of the dense arrays, poses several challenges: the reconstructed direct waves might not be part of the exact Green's function. Therefore, it can have sensitivity outside of the classical direct P-wave sensitivity between the stations. To better understand the source effects and the structural effects on this retrieved direct wave's sensitivity, we conduct full-waveform synthetic simulations of its sensitivity kernels. We try different structural changes and source geometry scenarios and highlight the different source/structure sensitivity trade-offs.
