Coseismic offsets recorded by borehole strainmeters from the 2014, Mw 6.0 South Napa, California earthquake: Reconciling tidal calibrations with earthquake source models.

Abstract:

The 24 August 2014 Mw 6.0 South Napa, California earthquake produced significant offsets on 12 borehole strainmeters in the San Francisco Bay area. These strainmeters are located between 24 and 80 km from the source and the observed offsets ranged up to 400 parts-per-billion (ppb), which exceeds their nominal precision by a factor of 100. However, the observed offsets in tidally-calibrated strains have RMS deviation of 130 ppb from strains predicted by previously published moment tensor derived from seismic data. Here, I show that the large misfit can be reduced by a combination of better tidal calibration and better modeling of the strain field from the earthquake. Borehole strainmeters require in-situ calibration, which historically has been accomplished by comparing their measurements of Earth tides with the strain-tides predicted by a model. Although borehole strainmeters accurately measure the deformation within the borehole, the long-wavelength strain signals from tides or other tectonic processes recorded in the borehole are modified by the presence of the borehole and the elastic properties of the grout and the instrument. Previous analyses of surface-mounted, strainmeter data and their relationship with the predicted tides suggest that tidal models could be in error by 30%. The poor fit of the borehole strainmeter data from this earthquake can be improved by simultaneously varying the components of the model tides up to 30% and making small adjustments to the point-source model of the earthquake, which reduces the RMS misfit from 130 to 18 ppb. This suggests that calibrations derived solely from tidal models limits the accuracy of borehole strainmeters. On the other hand, the revised calibration derived here becomes testable on strain measurements from future, large Bay area events.