Detecting earthquakes around Salton Sea following the 2010 Mw7.2 El Mayor-Cucapah earthquake using GPU parallel computing

Xiaofeng Meng, Xiao Yu, Zhigang Peng, & Bo Hong

Published June 2012, SCEC Contribution #1564

The recently developed matched filter technique is effective in detecting earthquakes during intensive aftershock or swarm sequences. However, currently our detection code can only process on single-CPU computers, which takes a long time to perform the cross-correlation between continuous seismic data and template events. In this paper, we present a GPU-based computation method to significantly accelerate the detection algorithm. By dividing the procedure into several routines and processing them in parallel, we achieve ~40 times speedup for one Nvidia GPU card compared to sequential CPU code. We apply the paralleled code to search around the Salton Sea geothermal field for missing earthquakes in a 90-day time window around the occurrence time of the 2010 Mw7.2 El Mayor-Cucapah earthquake. We obtain ~70 times more earthquakes than listed in the official Southern California Seismic Network catalog. These newly detected events could be used to help to better understand how earthquakes are triggered in the immediate vicinity of a mainshock rupture.

Citation
Meng, X., Yu, X., Peng, Z., & Hong, B. (2012, 6). Detecting earthquakes around Salton Sea following the 2010 Mw7.2 El Mayor-Cucapah earthquake using GPU parallel computing. Oral Presentation at International Conference on Computational Science. doi: 10.1016/j.procs.2012.04.100.