(left) Map view of relocated (Trugman 2020) Ridgecrest earthquakes included in the two-week (7/4/2019 - 7/17/2019) dataset of over 12,000 earthquakes M1+ provided in study. (top right) Comparison of stress drop estimated by several methods for the Prague, OK sequence show variability even when performed by the same researcher or using the similar methods, from Pennington et al. (2020). Number on violin plots give number of common earthquakes, with standard deviation shown below the method name. (bottom right) Initial comparison of stress drop for several M4 - M5 Ridgecrest aftershocks between Trugman (2019) spectral decomposition method and Parker et al. (2020) Arias ground-motion based method show a lack of obvious correlation. Figure from Rosas et al., 2020.

Introduction

Stress drop is a fundamental earthquake source parameter that in theory relates the average slip on a fault to rupture area, and in practice inherently indicates how energetic the earthquake is. Thus, earthquake stress drop is a key parameter in many ground motion, rupture simulation, and source physics problems in earthquake science. 

Many SCEC research goals require a closer understanding of stress, including quantifying stress heterogeneity across faults, considering how stress influences rupture propagation, and constraining how stress varies with depth on faults. One method of querying stress heterogeneity and relationship to controlling physical factors is through the estimation of earthquake stress drop. However, stress drops are notoriously variable and difficult to measure. Estimates by different researchers, using different methods or datasets, have yielded highly inconsistent values. This wide scatter may mask physical trends (such as depth, mechanism, regional variation, or dependence on fault heterogeneity) and implies that significant underestimated uncertainties. This SCEC Technical Activity Group (TAG) has been formed to understand the physical controls and methodological reasons for variance in stress drops, so that they can be used reliably by the earthquake science community.

This TAG began in 2021 with a study to compare and validate stress drop estimates for the 2019 Ridgecrest sequence of earthquakes. Using a common, provided dataset, researchers are asked to independently calculate and submit stress drop estimates on a shared platform for comparison and analysis. Anyone interested is invited to join this community validation study and the new Stress Drop TAG research efforts. See below for more information on how to participate.

Research Priorities

To address SCEC’s research priority of understanding stress, we need a coordinated, community effort to develop better methods for estimating stress drop and determining its dependency on earthquake and material heterogeneities. In this context, the goals of this Community Stress Drop Validation Study TAG are to understand the nature and causes of discrepancies in earthquake stress drop, as well as where random and physical variability arises. We want to answer the questions: 

1. How do differing methods and model assumptions affect stress drop estimates? How do
2. different researchers approach similar methods?
3. How does data quantity, quality, selection and processing affect stress drop estimates? 
4. How do physical source (mechanism, depth, radiation pattern, directivity), path (geometrical 
5. spreading, attenuation), and site (soil conditions, site attenuation) features affect the estimates?

The ultimate goal is to determine the most appropriate method for estimating stress drop for a given situation.

At the conclusion of the TAG, we hope to provide guidance on best practices for estimating stress drop and its real uncertainties, and assessing reliability of published measurements. This will allow future studies of source physics and ground motion to be guided by reliable stress drop measurements.

Current Validation Study: 2019 Ridgecrest Earthquake Sequence

The current community stress drop validation study is focused on the 2019 Ridgecrest earthquake sequence using a set of common waveforms. We invite any and all researchers to estimate stress drop or other earthquake source parameters by any method or means desired. 

The study is divided into two main activities: 1) Independent analysis of stress drop for the Ridgecrest sequence by researchers, and submission to the group validation; and 2) Meta-analysis to compare the submitted results. These activities will be iterative. That is, once the first round of estimates are submitted, initial analysis may indicate that researchers need to return to their methods and try something new. We will work together as a community to get the best understanding of stress drop. 

How to Participate in the Current Study

Everyone is encouraged to join the community stress drop validation study. You can download the data and perform analysis for stress drop, corner frequency or other source parameters. You can become involved in the meta-analysis comparing different results. Or participate in the upcoming workshop this fall. Sign up for the group mailing list to be informed of upcoming activities, events and results or email Annemarie Baltay and Rachel Abercrombie to let them know of your interest.

COMING SOON! Specific details on how to format and submit your stress drop estimates will be posted here and announced on the mailing list. One month prior to the first workshop, study participants will submit stress drop estimates for  basic meta-analysis, along with a detailed description of the parameters assumed and estimated. For example, what value of shear wave velocity was used? Did you estimate corner frequency and moment? For each component at each station, or for the earthquake as a whole? We ask that all study participants provide results in the asked-for formats in order to facilitate group comparisons. 

If you’re not quite ready to jump into the validation study, you are welcome to  join the first  workshop planned for fall 2021!

Common Dataset: 2019 Ridgecrest Earthquake Sequence

Study information and dataset is available through SCEDC: https://scedc.caltech.edu/data/stressdrop-ridgecrest.html.

The provided common dataset consists of ~13,000 earthquakes of magnitude 1+ over two weeks from July 4 until July 17. This contains the M7.1 and M6.4 mainshocks, 3 M5 earthquakes and 86 M4 events. This somewhat arbitrary two-week window was chosen to avoid introducing selection biases, yet retain a set of earthquakes sufficient for the wide variety of expected stress drop analyses. It is unlikely that any individual contributor will analyze all the earthquakes, but the approaches of different groups will be suitable for different subsets, allowing significant overlap.

Waveform Data: The provided data is recorded on 107 local and regional stations within 1-degree (~110km), and consists of broadband, accelerometer and geophone instruments (nodals excluded) with all horizontal and vertical components. The length of each record is proportional to the magnitude, with the record starting 15s before the origin time and ending 60s after for M1; for the M6+ the records start 90 before OT and end 310s after. The waveforms are in mseed format. 

Meta Data: 

• Full earthquake catalog with SCSN magnitudes and relocations from Trugman (2020)
• P- and S-wave phase picks for each waveform
• Vs30 estimates at all 98 stations
• Simple 1D velocity model for those wanting depth-dependent rupture velocity correction (provided by Malcom White, MIT)
• obspy Python script used to download the data

Benchmark Exercise: The first benchmark exercise is strict. Participants are asked to use only the earthquakes and waveforms provided in the common dataset, so that we can more easily compare methods. You do not need to use all of the data, just do not to add any other data. (If your method cannot be performed using the provided data only, please contact the Co-Leaders.)

Optional Exercise: The second component is optional and open ended. Participants may add events or waveforms to ensure your method works best. Please provide a description on how these results may differ from the benchmark exercise. 
Download the waveforms as a tar file, or use the provided Python script to pull straight to your machine. For the benchmark exercise, do not modify the Python script and use only the selected data (with the exception being if you require longer waveforms than provided, all else remaining the same). In the optional exercise, you may use the Python script to request additional stations or earthquakes. Download the dataset and script at:  https://scedc.caltech.edu/data/stressdrop-ridgecrest.html.

References

• Baltay, A., R. E. Abercrombie, and T. Taira (2021), A Community Stress Drop Validation Study Using the 2019 Ridgecrest Earthquake Dataset. SSA Annual Meeting 2021.
• Lin, G., Shearer, P. M., Hauksson, E., and Thurber, C. H. (2007), A three-dimensional crustal seismic velocity model for southern California from a composite event method, J. Geophys. Res., 112, B11306, doi:10.1029/2007JB004977.
• Parker, G., A. Baltay, J. Rekoske, E. M. Thompson (2020), Repeatable Source-, Path-, and Site-Effects from the 2019 Ridgecrest M7.1 Earthquake Sequence, Bull. Seis. Soc. Am. doi: 10.1785/0120200008
• Pennington, C., X. Chen, R. E. Abercrombie, and Q. Wu (2020). Cross Validation of Stress Drop Estimates and Interpretations for the 2011 Prague, OK, Earthquake Sequence Using Multiple Methods, J. Geophys. Res., https://doi.org/10.1029/2020JB020888
• Rosas, V. G. and A. S. Baltay (2020), Analyzing Stress drops and other earthquake parameters from the 2019 Ridgecrest Earthquake Sequence. Poster Presentation at 2020 SCEC Annual Meeting. SCEC Contribution 10738
• Trugman, D. T. (2020), Stress‐Drop and Source Scaling of the 2019 Ridgecrest, California, Earthquake Sequence. Bulletin of the Seismological Society of America 2020; 110 (4): 1859–1871. doi: 10.1785/0120200009
• White, M. C. A., H. Fang, R. D. Catchings, M. R. Goldman, J. H. Steidl and Y. Ben-Zion (2021), Detailed traveltime tomography and seismicity around the 2019 Mw 7.1 Ridgecrest, California, earthquake using dense rapid-response seismic data, Journ. Geophys. Int’l, in press. 
• Zhang, Q. and Lin, G. (2014). Three-dimensional Vp and Vp/Vs models in the Coso geothermal area, California: Seismic characterization of the magmatic system. J. Geophys. Res., 119(6):4907–4922.  https://doi.org/10.1002/2014JB010992