Measuring Long-Term Motion of the Sierra Nevada/Great Valley Microplate with GPS: Effect of Earthquake Cycle Transients Abstract

abstract

  • A goal of the EarthScope program is to chart the patterns of permanent deformation of the North American continent that reveal its dynamics and earthquake hazards. Doing so requires inferring the patterns of tectonic deformation that occur on time scales longer than the period of observation. One step in this process is to account for transient earthquake cycle effects that affect contemporary measurements. Failing to do so can bias hazard assessments where recent earthquakes have occurred. The EarthScope Plate Boundary Observatory geodetic data show postseismic earthquake cycle effects in their time series, e.g. non-linear motion. While some transients are easy to detect and distinguish from the background, signals from earthquakes that occurred years to decades earlier may linger and are more difficult to separate from long-term motion. The Sierra Nevada/Great Valley is a microplate (SNGV) that has few earthquakes, faults, or other obvious signs of internal deformation. However, GPS now show ~2 mm/yr of contraction along the 600 km length of the SNGV, apparently at odds with its being a microplate. It also exhibits a counterclockwise vertical axis spin superimposed on its northwest translation parallel to the San Andreas fault system. In this presentation we ask the question: Can the anomalous deformation and spin of the SNGV be explained by postseismic viscoelastic relaxation? To investigate we have developed a model of the time-variable deformation field associated with viscoelastic relaxation following the largest historic earthquakes California and Nevada (all events with M>7.0 and some smaller). The model is estimated from published earthquake parameters and mapped fault traces, and assumes a 1-dimensional, 3-layer Maxwell viscoelastic rheology. We use the VISCO1D software to generate model displacement time series at every GPS station and subtract the transient from the GPS time series before estimating the long-term trend. We systematically vary the model viscosities of the lower crust and upper mantle and select the model that best explains the observed strain and vertical axis spin. Preliminary results suggest that the estimated pole of rotation and strain rate of the SNGV are very sensitive to the assumed viscosity structure, and thus accurate models of mantle and crust rheology are needed to estimate long-term deformation patterns from PBO data.

publication date

  • 2012

presented at event