Source Imaging With a Multi‐Array Local Back‐Projection and Its Application to the 2019 Mw 6.4 and Mw 7.1 Ridgecrest Earthquakes. Issue 10 (5th October 2021)

• Record Type: Journal Article
• Authors: Xie, Y.; Bao, H.; Meng, L.
• Is Part Of: Journal of geophysical research. Volume 126:Issue 10(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-10-05
• Subjects: back‐projection -- local array -- Ridgecrest earthquake; Geomagnetism -- Periodicals; Geochemistry -- Periodicals; Geophysics -- Periodicals; Earth sciences -- Periodicals; 551.1
• DOI: 10.1029/2020JB021396
• Languages: English
• ISSNs: 2169-9313

Abstract

Abstract: This study aims to improve the multi‐array local back‐projection (MLBP) approach and to apply it to the 2019 M w 6.4 and M w 7.1 Ridgecrest, California earthquakes to resolve more details about their kinematic process, with the dense seismic network in California. Compared with teleseismic BPs, seismic array processing with stations located at local to regional distances images earthquake rupture process more quickly with higher resolution. To increase the objectivity of stations selection criteria and to be prepared for future real‐time BP implementations, we improve MLBP with an automatic procedure to group stations based on waveform coherence. We also apply empirical aftershock calibrations to account for the 3D path effect. Our MLBP highlights the rupture complexity in a multi‐fault system. The M w 6.4 quake initiates on a 5‐km‐long NW‐trending segment, then ruptures the primary SW‐trending fault at the speed of ∼1.3 km/s. The M w 7.1 quake ruptures bilaterally for 10 and 22 km, on the NW and SE portion of the fault, respectively, at the speed of 1–1.6 km/s. The rupture paths agree with aftershock distributions and surface rupture estimated from satellite imagery. The slow rupture propagation may be driven by the low structural maturity of the fault. This case study demonstrates the effectiveness of MLBP for earthquake source imaging and rapid hazard assessment. Plain Language Summary: Arrays of closely spaced sensors receiving earthquake signals have advantages over single stations for studying the features of earthquakes, by separating the similar, useful signal and noise. Theoretically, seismic arrays that are close to the earthquake source (within 300 km) have a higher resolution and shorter data collection time than more distant arrays (3, 300–10, 000 km away). But the data of the closer seismic arrays are more challenging to process due to low similarity between signals. We aim to better utilize local seismic arrays so that we can use the existing seismic stations in California to study the details of the 2019 M w 6.4 and M w 7.1 Ridgecrest, California earthquakes. We group the stations that have the highest similarity into multiple subarrays to improve the stability and resolution of the results. Our results reveal complex ruptures of multiple faults during the Ridgecrest earthquake sequence, which agree with aftershock locations and surface ruptures estimated from satellite imagery. We make this grouping process automatic to increase its objectivity and prepare it for future real‐time implementations, which is important for the estimates of earthquake source characterization and rapid hazard assessment. Key Points: We develop a multi‐array local back‐projection approach for earthquake source imaging using seismograms in 0.1–2.0 Hz The multi‐array local back‐projection method increases accuracy and stability by automatically grouping stations based on waveform coherence The back‐projections of the 2019 Ridgecrest earthquakes show bilateral rupture, which is consistent with aftershocks and surface ruptures