Assessing creep event rheologies using their temporal evolution

Aseismic creep events have been observed along the San Andreas Fault since the 1960s; yet, the fault properties that create these events still need to be identified. Here, we examine creepmeter data from events at Harris Ranch, CA, at the northern end of the San Andreas Fault's creeping section. We assess how well different rheological models describe the displacement-time evolution of surface-rupturing creep events. Using a basin-hopping approach, we minimise the misfit between the observed creepmeter data and predictions from five plausible rheological models for near-surface fault slip: 1) rate and state friction near steady state, (2) rate and state friction above steady state, (3) rate and state friction below steady state, (4) power-law flow, and (5) linear flow. We find most events are best fit by either a power-law flow or by rate and state friction below steady state. "Standard" velocity-strengthening friction models do not match the data well. This initial test of the rheologies could have one of several implications: (1) power-law creep somehow dominates even at near-surface conditions, (2) the near-surface is weak and mostly tracks the rheology of evolving slip at depth, or (3) a more complex rheology is required, perhaps including dilatancy or pore pressure evolution.