09:00 〜 09:15
[SSS04-01] The SCEC/USGS Community Stress Drop Validation Study using the 2019 Ridgecrest Earthquake Sequence Data
★Invited Papers
We present preliminary comparisons from the SCEC/USGS community stress drop validation study using the 2019 Ridgecrest earthquake sequence, in which researchers are invited to use a common dataset to estimate earthquake stress drop. We seek to understand the physical controls and methodological reasons for similarity or differences in stress drop estimates, so that they can be used more reliably by the earthquake science community. The common dataset consists of 2 weeks of earthquakes following the 2019 Ridgecrest M6.4 earthquake, including nearly 13,000 events of M1 and greater, recorded on stations within 100 km. As a community study, all are invited to join; more information can be found at https://www.scec.org/research/stress-drop-validation.
At the November 2021 SCEC workshop, attended by over 100 participants from 14 countries, 11 research groups submitted initial results using the common data set. The analytical approaches used include spectral decomposition/generalized inversion; spectral ratios or eGf in the frequency domain; source time functions or eGf in the time domain; and other methods such as ground-motion and single-station approaches. Direct comparison reveals considerable scatter, yet stronger correlations between results using similar methods. We identify a set of 22 events of M3 to M4 that were used in at least 8 analyses; some of these events show significantly reduced inter-method scatter than others. Since the workshop, researchers are refining results to focus on a subset of the data: 50 select events of M2+, including the 22 previously identified, representing a range of depths, mechanisms and azimuths. We also consider uncertainty in other parameters, such as moment, record quality control, earthquake depth, and path considerations. Lastly, we consider the use of a dataset of synthetic earthquake records, for which we would prescribe the stress drop, as a way to test the methods and assumptions alone, while controlling for path and recording station variability.
At the November 2021 SCEC workshop, attended by over 100 participants from 14 countries, 11 research groups submitted initial results using the common data set. The analytical approaches used include spectral decomposition/generalized inversion; spectral ratios or eGf in the frequency domain; source time functions or eGf in the time domain; and other methods such as ground-motion and single-station approaches. Direct comparison reveals considerable scatter, yet stronger correlations between results using similar methods. We identify a set of 22 events of M3 to M4 that were used in at least 8 analyses; some of these events show significantly reduced inter-method scatter than others. Since the workshop, researchers are refining results to focus on a subset of the data: 50 select events of M2+, including the 22 previously identified, representing a range of depths, mechanisms and azimuths. We also consider uncertainty in other parameters, such as moment, record quality control, earthquake depth, and path considerations. Lastly, we consider the use of a dataset of synthetic earthquake records, for which we would prescribe the stress drop, as a way to test the methods and assumptions alone, while controlling for path and recording station variability.