The Role of Clays in Faults: From Earthquake Heating to San Andreas Fault creep
Julia Krogh
UC Santa Cruz
- Date & Time
- Location
- Hybrid - in person and online via MS Teams
- Summary
Both long-term and transient fault strength are affected by the production and localization of frictionally weak phases, such as clays. Here I will discuss two sets of experiments that shed new light on clay behavior in faults.
First, I discuss whether clay alteration from frictional heating is a successful indicator of earthquake slip in the rock record, as has been previously suggested. I conducted ramped and isothermal heating experiments up to 900°C on smectite clays using in situ X-ray diffraction (XRD) to monitor reactions as they progressed. I observed that the extent of the high-temperature decomposition reaction was systematic with heating duration and peak temperature, enabling the quantification of reaction kinetics. Coupling our empirical decomposition kinetics to a model of fault heating enables determination of how much clay decomposition should occur in a single earthquake. I find that a large shallow earthquake similar to the 2011 Tohoku earthquake would cause only 5%–10% smectite decomposition. Since amorphous material is unlikely to persist and accumulate over earthquake timescales, a large temperature rise during earthquakes could occur without evidence of decomposed smectite in the rock record.
Second, I examine the role of frictionally weak phases in the creeping section of the San Andreas Fault. Recent work has identified an area in the northern transition region (north of San Juan Bautista) where three creeping strands identified by road offsets are located near the mapped main strand of the SAF, which is not actively creeping. Gouge samples from road outcrops as well as a drilling project in the summer of 2022 were analyzed via semi-quantitative XRD and friction experiments. I find no lithologic difference between the non-creeping and creeping strand samples, and all samples are frictionally strong and velocity strengthening. Healing rates are negligible at low normal stress and increase significantly with greater effective stress. These results indicate that lithology is not likely responsible for the disparate slip behavior observed between the fault strands. Instead, local variations in effective stress could play a significant role in reducing healing rate and therefore increasing stability on certain strands. I posit that local fault geometry plays a crucial role at this location, with the creeping strands dipping more shallowly than the main strand, resulting in lower normal stresses on the creeping strands.