How fault geometric complexity and frictional properties affect seismic fault behavior and accumulation of slip along strike-slip faults

How does slip at a point vary in and between individual earthquakes? What controls timing and magnitude of future earthquakes? Is single fault behavior inherently characteristic or does magnitude vary randomly from event to event? Addressing these questions has motivated studies in seismology and earthquake geology for the last hundred years. I reevaluated the surface slip distribution along the south-central SAF (a cornerstone in formulation of earthquake recurrence models) capitalizing on the high spatial resolution of the "B4" LiDAR topographic data set that covers it. My results show that slip during the Mw7.9 earthquake of 1857 along the Carrizo segment was 5.3+/-1.4m---significantly lower than the previously reported 8-10m. The updated 1857 surface slip distribution eliminates a core assumption for a strong Carrizo segment that dominates the fault behavior of the south-central SAF. A second portion of my dissertation regards numerical earthquake simulations. It is an outstanding question whether fault behavior of single faults is better described by characteristic earthquake (CE) or Gutenberg-Richter (GR) behavior. My results show that the maximum coseismic stress drop at depth zp within the seismogenic zone (related to temperature dependence of friction behavior) forms a strength barrier that limits down-dip rupture width of partial rupture earthquakes (events that rupture less than the full seismogenic width). The transition from partial-rupture (PR) to full-rupture (FR, events that rupture the full seismogenic width) is associated with an abrupt increase in rupture width, creating a bimodal magnitude-frequency distribution. I further created seismic catalogs for faults in different fault evolutionary stages using fault roughness a proxy for fault maturity. I find that bimodality and overall fault behavior change as a function of fault roughness. Bimodality and CM (characteristic magnitude of FR events) increase as faults become more mature. Slip-per-earthquake at a point is essentially variable at all evolutionary stages however variability is systematic as it reflects the inherent differences between PR and FR earthquakes. Model results suggest that the recurrence time of FR earthquakes at a point is related to the stress-renewal time at the depth of maximum coseismic stress drop z p.