Plain Language Summary
Earthquake stress drop (the change in shear stress before and after the earthquake) reflects the properties of the fault where earthquakes occur. Determining the scaling relationships between earthquake stress drop and magnitude enables us to predict the behavior of infrequent large earthquakes from the measurements of the more abundant small earthquakes. Measurement of stress drop is challenging, especially for small earthquakes. Different studies can obtain different stress drop values for the same earthquakes, leading to different interpretations. We use recordings from a high-resolution borehole network in Parkfield, California to measure source parameters for small earthquakes. We examine the influence of data limitation (e.g., frequency bandwidth) and corrections for material properties on stress drop resolution. We find that if the bandwidth is too narrow, the resulting source parameters can be systematically underestimated with large scatter. Insufficient corrections of material properties can lead to biased interpretation of stress drop patterns and scaling relationships. Our final results show that the stress drops of earthquakes in Parkfield do not depend on magnitude or depth, but exhibit strong spatial variations that are stable with time. The 2004 M6 earthquake caused temporal variations of stress drop, and the temporal changes are different at different fault patches.