Characterizing the Physical and Statistical Properties of Microseismicity and Earthquake Swarms in Western Nevada and Eastern California
AuthorRuhl, Christine Joy
AdvisorSmith, Kenneth D
Geological Sciences and Engineering
StatisticsView Usage Statistics
Abundant microseismicity is a poorly understood component of seismic hazard and diffuse plate boundary deformation in western Nevada and eastern California. Characterizing the details of such seismicity is an important observational step in improving our understanding of earthquake clustering, and earthquake physics in general. We develop a high-precision earthquake relocation routine for the Reno-Tahoe area of western Nevada and eastern California to investigate seismicity-defined fault structure in a diffuse deformational zone. Combining precise relocations with large sets of focal mechanisms enables description of incoherent seismicity clouds into distinct spatiotemporal clusters with defined structural patterns. Relocations from 2000 to 2015 reveal that microseismicity in the Reno-Tahoe region occurs primarily off the mapped faults, at the ends of faults, and on perpendicular structures to major faults; and in three main seismic source zones. Inversion of fault plane data reveals T-axis rotation into the Sierra Nevada block and a change in variability of the P-axes, which marks the abrupt transition from en echelon normal faulting and transtension of the central Walker Lane to northern Walker Lane style strike-slip dominated deformation.The strike-slip Mogul earthquake swarm dominates the easternmost seismic source zone in our study area and is exceptionally well recorded by near-source temporary seismic stations. Relocations and focal mechanisms indicate an internally clustered sequence in which foreshocks evolved on multiple well-resolved structures that generally surround the eventual mainshock fault rupture area. We objectively group events into clusters using space-time-magnitude distances between events and a nearest-neighbor threshold developed through randomization of the relocated catalog. Resulting clusters closely match subjective clusters grouped manually by careful inspection of timing, location, and mechanism type. Clusters highlight a narrow fault damage zone that widens upwards from ~0.25 km wide at 5 km depth to ~1.25 km at 2 km depth as well as a well-organized fault-fracture mesh offset to the west from the main fault zone and en echelon fault-perpendicular structures at the northwest termination of the main fault structure. The latter are similar to seismicity-defined structures observed around the Mohawk Valley fault zone, for example. These geometries have been associated with magmatic and fluid-driven swarms and suggest that fluid is a major player at Mogul. Swarm events migrate away from initial activity at a rate consistent with fluid flow and arrivals on near-source stations suggest an elevated velocity ratio for fault zone seismicity that is possibly related to fluid. To investigate the physics of the unusually shallow Mogul swarm, we estimate source parameters for 81 earthquakes in the complex 2008 Mogul earthquake sequence using both P- and S-wave EGF-derived spectral ratios. Temporary broadband seismometers deployed in the source area before the MW4.9 mainshock provide high quality records of many foreshocks and aftershocks, and an ideal opportunity to investigate variation of source parameters related to space, time, depth, style of faulting, and magnitude. High-quality EGFs are chosen by testing earthquakes that are (1) between 2.5 and 0.5 magnitude units smaller than each of the 95 target events and (2) within one estimated focal distance to each target event and accepting only those which meet our strict objective criteria. We compare corner frequencies estimated using (1) the Brune and Boatwright spectral models, (2) inverse-variance weighted-mean and stacked-spectral approaches, (3) various cross-correlation limits, and (4) various bandwidth limits. The Boatwright source model fits the data with lower variances and tighter minimums than the original Brune corner and results in more mainshock-EGF pairs that meet our criteria. Stacked results are more stable than individual weighted-mean results, and can use more ratios. The maximum frequency band used affects both the Boatwright and Brune corner frequency, with a stronger fall off observed at ~1/2 the maximum frequency band for the Boatwright model. We find variation within the sequence greater than the error range of each individual estimate, with no clear dependence on location, timing, depth, seismic moment, or mechanism type. Stress drops for this shallow swarm are similar to previous studies of stress drops, implying that other shallow, fluid-driven earthquakes (e.g., induced seismicity) do not have lower stress drops and, therefore, have similar (or higher, due to proximity to the surface) expected ground motions compared to typical earthquakes.