ABSTRACT

Recent discoveries of slow earthquakes, tectonic fault tremor, low frequency earthquakes and other modes of fault slip provide new insight into plate boundary fault mechanics, but demand the development of a new paradigm in terms of earthquake physics. These phenomena, known collectively as slow earthquakes, represent modes of failure that were recently thought to be theoretically impossible. Slow earthquakes are now well documented to occur on a global range of tectonic faults, and they occur along the subduction plate interface at depths varying from shallow near-trench regions to the down-dip end of the seismogenic zone. Slow slip redistributes stored elastic strain without catastrophic deformation, but the changes in plate boundary loading that accompany these events can trigger coseismic events and Great Earthquakes such as the 2011 Tohoku earthquake. The processes that control the mode of slip on tectonic plate boundaries are here informed by observations from ancient subduction fault zones in the Kodiak and Shimanto ancient accretionary complexes, with well-studied examples of exhumed faults that record deformation related to plate motions in the Cretaceous and Paleogene. These rocks exhibit evidence for two modes of slip behavior: 1) slow slip and quasi-dynamic fault motion across wide (10’s of m’s) zones, and 2) coseismic slip in a narrow fault gouge zone (~1 m) at the top of the plate boundary shear zone. Moreover, the observations of ancient faults indicate that the behavior of the subduction interface is buffered at high fluid pressure through a relationship between physical processes such as fracturing and fluid flow, and chemical processes such as pressure solution and mineral redistribution. Based on observations of microstructures in the ancient fault zones, a kinetic model is constructed to estimate the time required to seal a single fracture, which could be a proxy for the rates of healing or the increases in contact area that influence rupture propagation. Crack sealing is driven by diffusive redistribution of silica from solid-solid surfaces to undersaturated cracks where they precipitate as quartz. Our calculations predict that cracks heal on maximum time scales of hundreds of years, and that healing rates related to silica redistribution likely differ significantly for subduction zones with different thermal structures—a potential explanation for differences in observed slip behavior. For slow earthquakes, we propose that ruptures propagate at rates dictated by shear processes within a zone of finite thickness. Stress rises at the front of a propagating slow slip instability, leading to plastic failure along scaly slip surfaces in the footwall. Development of these microfaults causes initial weakening, but each slip surface hardens due to progressive increase in contact area related to dissolution and crack sealing. Unlike regular earthquakes, slow slip events are analogous to Volterra dislocations in crystals or self-healing slip pulses that have an inherent slip weakening mechanism followed by hardening that “puts on the brakes”.

About the Speaker:

Donald Fisher received his AB from Franklin Marshall College in 1983. In 1988, he completed a Ph. D at Brown University with a dissertation related to the tectonic and structural history of the Kodiak Islands in Alaska. He moved directly to Penn State University as an Assistant Professor and was promoted to Professor in 2001. Over his time at Penn State, he has conducted NSF-funded research in Alaska, Tohoku and southern Japan, Taiwan, New Zealand, Costa Rica, offshore Sumatra, and Panama. His research has focused on the processes that govern the rates of deformation, uplift history, structural evolution, and fault slip behaviors of active tectonic convergent plate boundaries.