A Real-Time Seismogeodetic Network Using MEMS Accelerometers and Its Performance in Kinematic Slip Inversions

Dara E Goldberg, Scripps Institution of Oceanography, La Jolla, CA, United States, Jennifer S Haase, Institute of Geophysics and Planetary Physics, La Jolla, United States, Diego Melgar, University of Oregon, Department of Earth Sciences, Eugene, United States, Yehuda Bock, Scripps Institution of Oceanography, University of California San Diego, Institute of Geophysics and Planetary Physics, La Jolla, United States, Jianghui Geng, Wuhan University, GNSS Research Center, Wuhan, China and Jessie K Saunders, Caltech, Pasadena, United States
Abstract:
The seismogeodetic combination of high-rate GPS observables and seismic acceleration captures the broadband on-scale recording of earthquake ground motions. The use of these data for determining rapid centroid moment tensor solutions (“fastCMT”) has been demonstrated in the post-analysis of the 2010 Mw 7.2 El Mayor-Cucapah earthquake. This seismogeodetic combination will improve source inversions for future earthquakes, but large-scale accelerometer deployment at the many available permanent GPS stations is limited by the cost of traditional observatory-grade accelerometers. Instead, we improve feasibility by installing SIO Geodetic Modules and low-cost MEMS accelerometers at 17 GPS stations in southern California near the San Andreas, San Jacinto, and Elsinore faults, transmitting data in real time for analysis of seismic velocity and displacement waveforms. We examine the performance of our seismogeodetic subnetwork using the El Mayor-Cucapah earthquake as our focus. We calculate a kinematic slip inversion, using the small set of seismogeodetic waveforms available at the time of the event, and assess the reliability of the result in comparison to the fastCMT solution. We evaluate reliability by using our model to predict ground motion at independent stations, and using recorded data as verification at a range of frequencies. Next we supplement the dataset by including realistic simulated waveforms for the additional 17 seismogeodetic stations, adding realistic seismogeodetic noise, and demonstrate the improved reliability of our result in terms of reducing the space of possible solutions due to better geometric constraints. The MEMS accelerometer has higher noise than the observatory-grade accelerometer, which we quantify using strong motion recordings from a series of UCSD NEES outdoor shaketable experiments conducted in December 2013 and January 2014. Results will provide confidence in the use of the MEMS accelerometer for large-scale deployment as an alternative to an observatory-grade accelerometer, as well as the prospects for the increased station density to improve the source parameters of future events, in particular a large earthquake forecast for the southern San Andreas fault.
Back to: Strong Motion Data, Networks, and Seismological Studies II