The depth of pseudotachylyte formation from detailed thermochronology and constraints on coseismic stress drop variability

Abstract

Pseudotachylytes are accepted as recording paleo-seismicity in the rock record. However, the interpretation of the mechanics of faulting based on pseudotachylyte generation is often hindered because the depth at which they form is poorly constrained. Here, we use thermochronology to determine the depth at which pseudotachylytes in the Sierra Nevada, California, formed. The pseudotachylytes formed in ≤ 10 m long patches over a rupture surface, the rest of which comprised cataclasites that did not melt. The age of the pseudotachylytes is found to be 76.6 0.3 Ma (2σ) from 40Ar/ 39Ar dating of pristine vein matrix. A suite of thermochronometers define the temperature-time path of the host rock granodiorite from ∼550 to 60°C. When the pseudotachylytes formed, the ambient temperature was 110 to 160°C, implying a depth of ∼2.4 to 6.0 km under typical geothermal gradients. At these depths, the failure stress on optimally oriented faults with Byerlee friction and hydrostatic pore pressure was ≤51 MPa. Following melting, the dynamic stress acting on the fault is the melt shear resistance, which we calculate to be <0.2 MPa, suggesting that the stress drop associated with melting was complete. To conform with seismologically observed dynamic stress drops averaged over an entire rupture (1 to 10 MPa), dynamic stress drop must have varied by at least an order of magnitude between the parts of the fault that melted and those that did not. Constraining the depth of pseudotachylyte formation using thermochronology therefore provides a quantitative estimate of the degree and scale of coseismic stress heterogeneity. © 2012 American Geophysical Union. All Rights Reserved.