Ambient noise tomography of the southern San Andreas fault using data recorded by dense fault zone array

The Southern San Andreas Fault (SSAF) is capable of hosting Mw > 7 earthquake and poses high seismic risk in California. To better understand the internal structure of the SSAF, more than 300 seismic nodes were deployed in the spring of 2020 and recorded continuous data for about one month near the Thousand Palms Oasis Preserve. The deployment consisted of a 4-km-long linear array across the SSAF, and two 2D subarrays in places where the linear array crosses the Banning Fault (BF) and the Mission Creek Fault (MCF) branches. In this study, we attempt to develop a multi-scale shear-wave velocity (Vs) model for the internal structure of the SSAF beneath the array using ambient noise data. Surface wave signals are first reconstructed from ambient noise cross-correlation (ANC) of each station pair. Clear Rayleigh and Love wave signals are observed in ANCs for frequencies between 0.5 Hz and 50 Hz. To enhance the surface wave signals, we apply denoising techniques based on three-station interferometry and double beamforming for the linear and 2D subarrays, respectively. Depending on the array aperture, we derive phase velocity models for fundamental mode surface waves from the denoised ANCs of each subarray separately. For the two 2D subarrays, we focus on surface waves in 10-50 Hz that provide high-resolution images for the top ~100 m, whereas a lower frequency band (0.5-10 Hz) is used to derive a phase velocity model based on the 4-km-long linear array for the top ~3 km. The resulting phase velocity models will be inverted for a multi-scale Vs model for the subsurface structure of the SSAF at the site, and will be presented in the meeting.