Seismogeodesy for Rapid Response: The 2014 Mw 6 South Napa Earthquake

Rapid assessment of medium to large events with finite source models at regional scales has traditionally been delayed by limitations in seismic-only monitoring systems. Broadband and short period sensors saturate and strong motion sensors are affected by baseline offsets. In practice, this means that strong motion data must be high pass filtered. This approximation breaks down as the earthquakes get larger since the long period band of the spectrum best characterizes these events. In contrast geodetic data record long period motions down to the coseismic offset. They are however less sensitive, with single-epoch noise levels of the order of 1cm. With data from 34 collocated GPS/strong motion stations around the South Napa earthquake, we demonstrate a seismogeodetic algorithm that produces broadband strong motion velocity and displacements. We further show that these data can be ingested into source modeling algorithms with minimal operator interaction. Through retrospective analysis off the South Napa data, we show that one can produce a full suite of models from rapid moment tensors to kinematic slip inversions. While displacement amplitudes are small, often below GPS noise levels, the seismogeodetic solution provides an objective constraint on the accelerometer integration. The rapid inversion to find shallow slip can provide an indication of surface rupture and indicate that enhanced response is likely for road repairs and such. We quantify the improvements and advantages to using 5Hz GPS over 1Hz data.There are over 600 real-time monitoring GPS stations in the western US maintained by several operators; the NASA-sponsored READI project aims to unify these into a network of networks for prototyping earthquake and tsunami warning systems. We argue that the results shown here demonstrate the need for full integration of geodetic networks into seismic monitoring systems and that with current deployed stations this is within reach. However, only a small number of seismic and GPS stations are collocated in the western US In abstracts by Haase et al. and Goldberg et al. we show that low-cost MEMS accelerometers can be used to upgrade existing real-time GPS stations to seismogeodetic capability; 17 stations in southern California have already been upgraded and similar upgrades in the Bay area are forthcoming.