Poster #026, San Andreas Fault System (SAFS)

Creep rate gradient through the locked-to-creeping transition zone measured by differential lidar on the San Andreas fault at Parkfield, CA

Michael Vadman, & Sean P. Bemis
Poster Image: 

Poster Presentation

2020 SCEC Annual Meeting, Poster #026, SCEC Contribution #10558 VIEW PDF
The San Andreas fault (SAf) near Parkfield, CA, transitions from fully locked to creeping over 50 km of fault length, approaching a surface creep rate of ~25 mm/yr. The gradient in surface creep rate is illustrated through measurements at several locations through the transition zone, but not at sufficient resolution to fully characterize how the change is distributed through this zone. To quantify how the surface creep rate changes through the transition zone, we use differential lidar to measure changes between the 2005 B4 Project and the 2018 Parkfield NCALM dataset. We calculate displacements (both magnitude and direction) over a 13-year period on a 28 km x 1.75 km swath, roughly centere...d on Parkfield, using a windowed Iterative Closest Point (ICP) algorithm. With 25 m spacing between each 50 m window, this process outputs ~58,000 displacement vectors. The resulting displacement field appears to exhibit both tectonic and non-tectonic signals, including mass movements, land-use changes, and an apparent global coordinate shift. We isolate the tectonic displacement signal by filtering the data using a series of logical and site-specific criteria, such as removing outlier displacement values and directions (e.g., those highly oblique to the SAf trend). To extract the surface creep rate from this displacement field, we subdivide the dataset along the fault, averaging each section to assess along- and cross-fault patterns. These measurements illustrate a trend of small (but variable) displacements southeast portion of the swath, increasing to ~10 mm/yr near the center of the swath at Parkfield, and increasing further to 18-20 mm/yr in the northwestern end of the swath. These creep rates are consistent with prior studies and the expected transition from mostly-locked to mostly-creeping, but the variable rates and directions through the southeastern end of the swath indicate additional non-tectonic signals present in the data. These non-random, erroneous displacements are common on flat, agricultural lands, suggesting that the primary texture of the fields changing over time leads to a non-unique ICP fit. Examining multi-year time windows complicates differential lidar analysis because the non-tectonic noise potentially increases relative to the tectonic signal. Site-specific assessment of possible non-tectonic signals, such as land-use changes, mass movements, and regions of poor suitability of ICP analysis, should isolate the tectonic surface displacements.