Normal, Drought, Heavy Precipitation Year: the Contribution of Variable Seasonal Loading on the Horizontal Strain Field in California

We quantify time-dependent horizontal transient strains using cGPS data between 2007 and 2019 within the plate boundary zone in California and investigate the link between observed seasonal horizontal strain anomalies and hydrologic loading patterns. Our long-wavelength transient strain model shows dilatational strain variations of ± 10 - 20 x 10 -9 and the associated Coulomb stress changes of ± 1 kPa on faults. During the summer, a prominent zone of dilatation and shear develops along the San Andreas Fault (SAF) zone. The Great Valley and the Sierra Nevada also experience dilatational strains and displace a total of about 2 mm east with respect to Pacific Plate reference frame (Fig. 1b). During winter, on the other hand, the dilatational strain patterns along SAF reverse and the Great Valley and Sierra Nevada experience compression, moving westward in a total of 2 mm (Fig. 1a). During the drought years between 2012 and 2015, the winter patterns are slightly diminished, while the amplitudes of summer deformation patterns are similar to those in normal summers. During heavy precipitation winters, we detect a significant dilatational compression along the Great Valley and Sierra Nevada. To investigate the link between the observed seasonal strain changes and the physics behind it, we analyze the UNAVCO hydrologic model [ Puskas et al., 2017] and surface water estimates [Argus et al., 2017]. Both of the horizontal elastic Green's function responses predict, to first order, similar spatial and temporal patterns of deformation, supporting our hypothesis that the source of our long-wavelength horizontal strain field is mostly from surface hydrologic loads. We also investigate links between seasonal stress changes and seismicity, applying a rate-and-state friction model. In the Bay Area, for instance, the predicted seismicity rate variations are in accord with the observed rates of seismicity. Furthermore, in the regions of 2014 M 6.0 South Napa and 2019 M 7.1 Ridgecrest Earthquake, our seasonal deformation model predicts extensional dilations a month prior to these events, indicating that the large events may have been triggered by seasonal stress changes (Fig 1c, d). These results suggest that seasonal stress changes may provide important insights into mechanisms that influence seismicity rate changes in California.