Presentation by Dr. Hillary Chang.
Challenges in microearthquake source parameter modeling: Are the observations source variabilities or site effects?
Source properties of small earthquakes, such as source dimension and stress drop, help us to constrain source physics and assess seismic hazards. Small events carry information about the stress state in the subsurface. They also help us predict the behavior of larger earthquakes. However, the source properties of small earthquakes (magnitude less than 3) are poorly constrained because the high-frequency signals of small earthquakes are subject to wave propagation effects such as attenuation, which can result in an apparent breakdown of earthquake self-similarity. To date, researchers have been trying to improve the accuracy of source parameter measurements. In this talk, I share my quest to understand the cause of variability in source parameter measurements for induced seismicity in Oklahoma during my PhD. I use a dense array to analyze the relationship between ground motion, subsurface structure, and source parameter measurements. The dense array provides an unprecedented opportunity to understand the influence of site effects. My analysis shows the importance of considering near-surface attenuation in source parameter modeling and how results with large biases can still appear to make a good fit to the data. Many microearthquake studies using surface sensors are likely subject to trade-offs with near-surface attenuation. Alternative approaches to reduce the potential site-effect bias include using a small event as an Empirical Green’s Function or using borehole sensors. In this context, borehole distributed acoustic sensing (DAS) that uses fiber-optic cables as seismic sensors is a promising tool. I developed a workflow using borehole DAS to characterize microearthquakes at the Cape Modern geothermal field. The downhole cable provides densely distributed sensors in the subsurface, allowing us to resolve attenuation structure and minimize site effects in source parameter modeling.
