Project Abstract
|
The objective of this project is to address the post-seismic deformation following two major ~M7 earthquakes along the southern San Andreas fault system. We focused our initial investigation into the deformation following the 2010 M7.2 El Mayor-Cucapah earthquake in northern Baja California, Mexico. PBO installed 9 continuously operating GPS (CGPS) sites in the months following the earthquake, providing for the only continuous sites in close proximity to the earthquake, as well as to the south and west of the earthquake. Pre-existing CGPS sites are focused north of the border, providing resolution for only a portion of the region. However, to fully understand the post-seismic deformation observed at the newly constructed sites, we needed to estimate the pre-EMC velocities at each of these newly constructed sites. We accomplished this through analysis of continuous and campaign GPS data for the study region. We estimated interseismic velocities for the two decades leading up to the EMC earthquake, through analysis of the coordinate time series for each site, accounting for annual and semi-annual perturbations to the secular motions. We included the velocity estimates from the SCEC CMM4 velocity solution by rotating the estimates into our reference frame through a minimization of motions between common sites. We constructed an elastic block model for the study region, using mapped faults to guide the location of our model faults. We estimated slip rates for each of the model faults as well as block motions for each of the model crustal blocks. From these block motions, we were able to estimate the pre-EMC velocity at each of the newly constructed CGPS sites in northern Baja California. The first of these new PBO sites was installed in 2010.8, so we investigated the time series for all 9 newly constructed sites, as well as 12 additional CGPS stations with similar spatial extent in southern California for the time period 2010.8-2012.5. Seeing no obvious exponential or logarithmic decay signal in the time-series for these sites, we estimated a velocity following the same procedure as for the pre-EMC period. Upon comparison of these post-EMC velocity estimates to the pre-EMC estimates, we find that the two estimates don’t agree, and the differences decrease systematically with distance from the earthquake with orientations exhibiting a butterfly pattern, as expected from a strike-slip earthquake. We attribute these differences in secular velocity to unmodeled post-seismic deformation. We modeled the observed post-seismic deformation through relaxation of a viscoelastic medium. We used a simple 3-layer viscoelastic model, with an elastic upper crust overlying a viscoelastic lower crustal layer overlying a viscoelastic halfspace representing the uppermost mantle. Models with a slightly stronger lower crust compared to the uppermost mantle minimized the misfit to the observed values, with the preferred model of 4e19 Pa-s for the lower crust and 9e18 Pa-s for the uppermost mantle. We anticipate that more detailed rheologic models, particularly in the Salton Trough region where high heat flow and thinning of the crust is present, will provide a better fit to the observed data. |
