| Description |
Due to its location in the interior of the Colorado Plateau of east-central Utah, the San Rafael Swell experiences very little seismicity. As such, any earthquake sequence in the region is important as a tool to understand the geological structures and geophysical processes in the area. The 2019 Clawson sequence-189 earthquakes that occurred on the northwest edge of the San Rafael Swell from March 13 to May 27, 2019-provided an excellent opportunity to uncover fault structures within the region. In this thesis, I perform high-precision relocations of the 2019 Clawson sequence and estimate the orientation of the presumptive fault plane. Using time windows of -0.6 to 1.4 s around P-wave arrivals and -1.2 to 4.8 s around S-wave arrivals, P- and S-waves for each earthquake were cross-correlated to produce 332,631 differential P-travel times and 99,513 differential S-travel times. These values were then used as inputs for GrowClust, a cluster-based relative relocation program. Bandpass filters of 1 to 10 Hz, 1 to 5 Hz, and 2 to 15 Hz were tested to determine which produced the best differential travel times and subsequent relocations. Additionally, each event was artificially set to be located at 39.04˚N, -110.95˚E with depths of 5 km, 8 km, and 18 km before being used as inputs to GrowClust in order to determine the stability of the apparent fault orientation. Each relocation produced a single cluster that contained a majority of the events. This cluster aligned with a northeast dipping fault and a potential secondary fault near the south end of the sequence. The 2-to-15 Hz filter was chosen to measure the strike and dip of the relocated sequence using principal component analysis. The resulting fault plane had a strike of 335.18˚ ± 8.78˚, a dip of 89.05˚ ± 1.03˚, and planarity of 0.973 ± 0.006. Limiting calculations to events iii relocated north of 39.044˚, and thus removing the influence of the secondary southern fault, the fault plane had a strike of 319.61˚ ± 0.38˚, a dip of 84.44˚ ± 0.41˚, and planarity of 0.992 ± 0.002. While poor depth constraints on the initial earthquake locations cast doubt upon the low formal uncertainty of the dip measurement, it is likely the fault is near vertical. The high planarity especially supports this possibility. Regardless, the fault strike is likely to be highly accurate given the consistency of the relocations. Earthquake detection techniques based on cross-correlation of template waveforms could prove invaluable for further clarifying the structures found in this study. If newly detected events are relocated alongside the initial catalog events, then the fault plane may be refined and even better constrained. Initial work indicates that over 500 new earthquakes could be added to the Clawson sequence using template-based cross-correlation methods. |
