Vertical Crustal Motion Across Southern California Fault Systems from GPS and InSAR Abstract


  • We combine GPS and InSAR measurements to measure the contemporary vertical motion across the active fault systems of southern California. We use data from all of the public domain high precision GPS networks available in the region including the NSF EarthScope Plate Boundary Observatory, SCIGN, and other municipal networks. Solutions are obtained using the GIPSY OASIS II software in a mega-network analysis that includes over 12,000 stations globally distributed. The solutions are aligned to the new North America fixed reference frame (NA12) that provides strong vertical reference, providing a means to compare vertical rates across different regions within the Pacific/North America plate boundary. To supply additional constraints on vertical deformation at wavelengths near or shorter than the spacing of GPS stations, we incorporate InSAR data from the WinSAR archive. We use over 750 ERS and ENVISAT radar scenes from between 1992 and 2009, including 4 frames from 5 tracks to form over 10,000 interferograms, providing line-of-sight (LOS) velocities for overlapping domains. To separate the contributions from vertical and horizontal signals, we subtract the line-of-sight signal of horizontal deformation estimated from a GPS strain rate map, and un-project the remaining signal into the vertical rate map aligned to NA12. To minimize systematic misfits between InSAR and GPS rate fields we employ a time series model that includes and exponential decay from viscoelastic relaxation following the 1992 Landers and 1999 Hector Mine earthquakes. InSAR and GPS motions track one another well, with RMS difference in vertical rate of 1.0 mm/yr, where the signal of vertical rate varies between -5.0 and 2.6 mm/yr. The vertical rates show both basin-scale pockets of subsidence and regional wavelength variations in uplift rate. An uplift feature of near 2 mm/yr is centered on the SAF near its junction with the San Jacinto fault and eastern Transverse Ranges. We find that this feature can be explained using a viscoelastic earthquake cycle model, and suggests a slip rate on the San Bernardino section of the SAF of 15 mm/yr, in better agreement with geologic rates compared to earlier models.


publication date

  • 2014

presented at event