Date: April 13, 2023

Time: 11am

Location: ES&T L1125

Miguel Neves

Title: A deep look into continental tectonic processes using high-resolution earthquake catalogs

Committee members: Dr. Zhigang Peng, Dr. Andrew Newman, Dr. Felix Herrmann, Dr. Samer Naif, Dr Susana Custódio

Abstract:

According to the theory of plate tectonics, the Earth’s crust is formed by rigid blocks that move relative to each other along linear faults that bound the blocks, where most deformation occurs. But it is now recognized that plate boundaries can also be broad areas of deformation, and slowly deforming regions in plate interiors can also generate significant and destructive earthquakes. However, even at linear plate boundaries, seismologists still struggle to understand the seismic cycle, and improvements in seismic monitoring also revealed different stress sources capable of interacting with faults and earthquakes. The work presented here focuses on how we can systemically and efficiently compile high-resolution earthquake catalogs and use them to better understand earthquake sequences, the kinematics of faults, and stress interactions in different continental regions. Earthquake catalogs are often incomplete at lower magnitudes, limiting the information available for understanding the sequence and faulting processes. To improve earthquake catalogs, the past decades have seen the development of different techniques to automatically analyze seismic data, such as matched filter and deep learning earthquake detectors. I first present a study using matched filter detection to study the 2004 M6 Parkfield, California, earthquake sequence along the San Andreas Fault. I improve the earthquake catalog by about 3 times the number of events listed in the Northern California Seismic Network catalog, and use the new catalog to study the period prior to the 2004 mainshock and interactions with tidal stresses. No clear precursory signals to the 2004 mainshock are identified, but an increase in the seismic activity is observed in the creeping section of the San Andreas Fault in the weeks prior to the mainshock, accompanied by a decrease in the b-value parameter in the Gutenberg-Richter relationship. These results suggest stress is increasingly released seismically in the creeping section before the mainshock occurrence. However, seismicity rates remain stable in the Parkfield section where the 2004 mainshock ruptured. The analysis of tidal stress variations in the Parkfield segment during the 2004 sequence also reveals that microearthquakes at Parkfield are modulated by tidal stresses, but with different impacts before and after the mainshock. I then describe a project using a deep learning earthquake detector in Iberia, a mostly slowly deforming region in Southwest Europe. I analyze 7 years of seismic data from 552 stations in Iberia to improve the quality of the regional earthquake catalog and gain new insights into the seismicity behavior in the region. I fine-tune a deep-learning earthquake detector and phase picker to analyze the Iberian datasets, using a small dataset of 28,622 waveforms from the region. Using the new phase picker, I compile an earthquake catalog with 56,354 events. Additionally, I identify anthropogenic signals in the catalog revealing that most clusters in slowly deforming Iberia are connected to areas of anthropogenic signals, which suggests that induced activity is widespread in the region. Using the new catalog, I was also able to identify new lineaments in Western Iberia, likely illuminating new fault structures in that region. Lastly, combining matched filter detection and a deep learning detector, I study the August 9, 2020, M5.1 Sparta, North Carolina, earthquake sequence. This earthquake ruptured the uppermost crust near the town of Sparta, and is likely the first identified surface-rupturing event in the Eastern United States in the past 200 years. The new catalog reveals that the mainshock nucleated near an intersection point of two fault strands, a blind strike-slip fault where the rupture was possibly initiated, and a reverse fault associated with the identified surface rupture, possibly part of a flower structure like diffuse fault zone. The works presented here highlight how high-resolution earthquake catalogs can reveal fault structures and kinematics, stress interactions and seismicity patterns. These findings give insights into the seismic cycle and can have important implications on seismic hazard estimation, particularly in slowly deforming settings.