Predicting earthquakes and mitigating earthquake hazards requires understanding how faults work; for example what controls or limits the size of earthquakes and the recurrence time of earthquakes on a particular fault. Most of the slip on continental strike-slip faults such as the San Andreas occurs during earthquakes, while most of the slip on mid-ocean ridge transform faults (RTFs) is accomplished by aseismic creep not related to earthquakes, and the largest events on RTFs are comparatively small, magnitude 6-7. This study will be the first to model the behavior of oceanic transform faults using laboratory-derived rate and state friction laws that have been successfully applied to continental faults. Models will be constrained by seismicity on the faults recorded by ocean bottom seismographs. Among the broader impacts of this study are improved understanding of earthquakes and earthquake mechanisms, and support for a graduate student.
This project developed numerical models of earthquake rupture on oceanic transform faults. This type of faults have the unique advantage that the size distribution of their largest earthquakes can be predicted directly from the velocities of the tectonic plates. The goal of the numerical models was to recreate these scaling relationships between fundamental fault parameters, such as plate motion velocity and the size-distribution of earthquakes on different faults. The outcome of the modelling study was that first-order descriptions of fault failure derived from lab experiments on rocks can explain the observed saling relations but only in a narrow parameter space and possibly with some behavior that contradicts observations. The primary discrepancy was in wether the large ruptures in the model occurred on the same fault patch over and over again as is seen in at least some oceanic transform faults. The modelled ruptures moved around from one cycle to the next on the fault while still only accounting for about 1/5 of plate-motion. To match the spatial complexities in earthquake behavior seen on real faults requires more complex rock friction laws with greater heterogenetiy in material properties. In terms of broader impacts.this project provided the funding for training two post-doctoral investigators in numerical modeling of earthquake rupture.
