Statistical Properties of Induced and Triggered Earthquakes at The Geysers, California

Abstract:

This study considers the statistics of induced and triggered seismicity at The Geysers geothermal field, California. Data is considered from the regional Northern California Seismic Network (NCSN) and local Lawrence Berkeley National Laboratory Network (LBNLN). Both data sets give good GR data fits for 2009-2014 but NCSN data have b=1.15 and LBNLN data have b=1.36. Comparing 18,000 individual earthquakes we find on average M_LBNLN = M_NCSN+0.5. Thus care must be taken when both data sets are used. We hypothesize that the strain accumulated due to the plate motions is a balance by the strain released in earthquakes with a maximum upper limit M_max. We compare the strain associated with seismicity with the tectonic GPS strain being accumulated in the region. Taking the NCSN GR data with an upper magnitude cutoff, we find this cutoff to be M_max=4.74. This is consistent with an observed upper magnitude limit to The Geysers seismicity at about M=5. We present studies of aftershock statistics of four M = 4.43, 4.16, 4.62, and 4.53 earthquakes. We find both GR and Omori Law statistics to be typical of tectonic earthquakes. We suggest that the four earthquakes release accumulated tectonic stresses but injected fluids reduce the stress required for rupture initiation. We also consider triggered seismicity caused by three remote earthquakes. We obtain excellent data for the 2010 M=7.2 El Mayor-Cucapah and the M=6.0 South Napa earthquakes. In the first case a M=3.37 event was triggered and in the second case a M=4.48 event was triggered. We conclude that the observed seismicity consists primarily of aftershocks of the large triggered earthquakes and that the directly triggered earthquakes do not satisfy GR frequency-magnitude statistics.