We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Cross Validation of Stress Drop Estimates and Interpretations for the 2011 Prague, OK, Earthquake Sequence Using Multiple Methods.
- Authors
Pennington, Colin N.; Chen, Xiaowei; Abercrombie, Rachel E.; Wu, Qimin
- Abstract
We compare source parameter estimates for earthquakes in the 2011 Prague Mw 5.7, Oklahoma, sequence to investigate random uncertainty and systematic bias, and resolve reliable relative variations in stress drop. Source parameters provide insight into the earthquake rupture processes but large variations between studies occur. The Prague earthquake sequence is a prime example of this, with different studies reaching contrasting interpretations of the effects of injection on source parameters. We examine the Prague earthquake sequence using a single coherent catalog for all the events detected by the Oklahoma Geological Survey (OGS) and McMahon et al. (2017). We use three principal approaches to estimate stress drop in order to understand the biases of each: a spectral decomposition method based on stacking, individual event spectral modeling, and a spectral ratio method based on highly correlated events. We also compare our results with previous studies for the Prague sequences aftershocks, as well as past results for the Mw 4.8 foreshock and Mw 4.8 aftershock and Mw 5.7 mainshock. The absolute values of stress drop vary significantly between methods, but the relative patterns remain consistent, except when low quality or low bandwidth data are included. The consistent relative patterns reveal that the stress drops of aftershocks are dependent on the fault orientation and the proximity of the events to the mainshocks slip. These results indicate that fault structure as well as past events play an important role in stress drop patterns. Plain Language Summary: The area of slip in an earthquake, and the stress released (the stress drop) are often estimated from the frequency content of the seismic waves using simple models. Many methods have been developed to separate the effect of the wave traveling through the Earth and the frequency content of the earthquake source itself. The stress drops calculated in different studies using different methods, can vary significantly for the same earthquakes. Such variability casts doubt on their use for predicting ground motion or understanding the rupture process. We use a range of approaches to analyze earthquakes in Oklahoma, and compare our results with past estimates. We find the chosen method or data can systematically change the absolute stress drop measurements, but the relative differences in stress drop between individual earthquakes are more stable, meaning that the patterns in space and time can be interpreted. We observe that a faults orientation impacts the stress drops of earthquakes along it. Also, that the slip due to the largest earthquake affect the stress drops of the earthquakes that follow it. We recommend using cross validation with multiple methods in studies that examine large numbers of earthquakes, to identify method issues or data problems. Key Points: Data selection and method choice can systematically bias absolute stress drop estimates and significantly affect apparent variabilitySpatial and temporal variation of relative stress drop estimates are more consistent across methods with similar datasetsConsistent spatial patterns across methods indicate that stress drops are dependent on fault orientation and proximity to mainshock slip
- Subjects
PRAGUE (Okla.); EARTHQUAKES; GEOLOGICAL surveys; EARTHQUAKE aftershocks; CHEMICAL weathering
- Publication
Journal of Geophysical Research. Solid Earth, 2021, Vol 126, Issue 3, p1
- ISSN
2169-9313
- Publication type
Article
- DOI
10.1029/2020JB020888