Multiscale Dynamics of Aseismic Slip on Central San Andreas Fault

Understanding the evolution of aseismic slip enables constraining the fault's seismic budget and provides insight into dynamics of creep. Inverting the time series of surface deformation measured along the Central San Andreas Fault obtained from interferometric synthetic aperture radar in combination with measurements of repeating earthquakes, we constrain the spatiotemporal distribution of creep during 1992-2010. We identify a new class of intermediate-term creep rate variations that evolve over decadal scale, releasing stress on the accelerating zone and loading adjacent decelerating patches. We further show that in short-term (<2year period), creep avalanches, that is, isolated clusters of accelerated aseismic slip with velocities exceeding the long-term rate, govern the dynamics of creep. The statistical properties of these avalanches suggest existence of elevated pore pressure in the fault zone, consistent with laboratory experiments. Plain Language Summary The movement between tectonic plates can be accommodated through either earthquake or continuous aseismic sliding, known as fault creep. Here we studied the creeping behavior on the central similar to 130 km segment of the San Andreas Fault. Enabled by high-resolution satellite radar measurements, we tracked the surface deformation caused by fault movement between 1992 and 2010 and developed kinematic models to constrain the time evolution of creep at depth. We found that at decadal time scale, creep acceleration characterized some parts of the fault, imparting stress on the neighboring decelerating and locked zones. We showed that the 2004 M-w 6 Parkfield earthquake was preceded by a similar decadal-scale acceleration on the adjacent creeping segment. We further showed that the rate of creep additionally varies on shorter (