Where the Fault Meets the Road: Structure, Deformation and Rheology of the Urban Hayward Fault

We are measuring and modeling the near-fault interseismic deformation of the Hayward Fault to better understand the fault zone properties and their variations in space and time, using SAR interferometry, Lidar scanning, and GNSS analysis. We will use model inversion methods to estimate the accumulation and release of strain on the Hayward Fault to estimate the interseismic coupling at depth and assess off-fault deformation in the zone near the fault. The Hayward Fault has been assessed in the Uniform California Earthquake Rupture Forecast version 3 (UCERF3) as the second most likely location in California for an earthquake larger than magnitude 6.5 in the next 30 years and runs through the extensively urbanized eastern San Francisco Bay area. Interseismic deformation on the Hayward Fault is therefore important for understanding earthquake risk in the area as well as other basic processes such as the evolution of faults over longer periods of time. In light of the fact that the Hayward Fault last ruptured in a large M ~6.5 earthquake in 1868 and paleoseismic studies show it has a roughly 150-year recurrence interval, it is imperative that we understand in detail how stress is accumulating in this fault zone. The main sources of interferometric SAR (InSAR) data are the NASA UAVSAR high-resolution L-band SAR acquired over the area since 2009, along with Copernicus Sentinel-1A and -1B for finer temporal resolution. We do time-series analysis of the UAVSAR data with the InSAR Scientific Computing Environment (ISCE) and MintPy at about 4 meter spatial resolution. We are starting the development of models to use InSAR and Lidar deformation constraints on interseismic deformation to infer the spatio-temporal, along-fault distribution of slip on the Hayward Fault and the distribution of strain across the uppermost fault zone. We will use Bayesian inference methods to estimate the uncertainties of the results including both data covariance and prior model assumptions.