S039-0008
Earthquake depths, focal mechanisms, and stress in the Lower St. Lawrence Seismic Zone
Abstract:
Focusing on 2015-2020, after seismometer coverage in the eastern LSZ improved, we apply PhaseNet [Zhu and Beroza, 2018] and REAL [Zhang et al., 2019] to detect 70 new earthquakes (vs. 175 catalog earthquakes in the same timespan/region). A new 1D, gradient velocity model is produced via a Monte Carlo search; vP/vS is assumed to be constant at 1.77, which was computed using the Wadati method. Hypocenter estimates are refined via the triple-difference method, with additional depth constraints from the timing of sP depth phases. Focal mechanisms of ∼150 earthquakes are estimated with automatic P-wave first-motions and (absolute value) P-SV-SH amplitude ratios. A grid search and Bayesian analysis allow for robust uncertainty estimates, which in turn allow for uncertainty estimates on the state of tectonic stress. The recovered WNW-ESE σ1 is consistent with a stress tensor controlled by deglaciation, and in agreement with a previous estimate. The orientation of relocated aftershocks to the 1999 M5.1 event, as well as the stress inversion result, both favor a SE-NW strike for the mainshock fault plane. This is consistent with an earlier interpretation [Lamontagne et al., 2003], but is perpendicular to the expected orientation of paleo-rift faults that are expected to host earthquakes. Many of the recovered focal mechanisms similarly suggest fault planes perpendicular to the rift, although dense east-west lineations of earthquakes may indicate that paleo-rift faults are the largest (and most hazardous) faults in the LSZ.