SCEC Award Number 14162 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title Prospective Earthquake Forecasts based on Coulomb Stress
Investigator(s)
Name Organization
David Jackson University of California, Los Angeles
Other Participants Anne Strader
SCEC Priorities 2b, 4e, 2d SCEC Groups EFP, Seismology, SDOT
Report Due Date 03/15/2015 Date Report Submitted N/A
Project Abstract
 We showed that conclusions of Deng and Sykes, 1997, still hold. With 95% confidence, M≥ 2.8 earthquakes preferentially nucleate where shear or Coulomb stress increased. On average, 59% of earthquakes occurred within stress-enhanced zones, regardless of the choice of rupture plane or type of stress change. The 1999 Hector Mine, California, earthquake increased the seismicity in regions of positive and negative stress change but more so in the positive regions. Earthquakes frequently occur after their calculated peak stress has occurred, indicating that elastic calculations do not tell the whole story. We found no significant evidence that stress change affects the magnitude of receiver quakes. Thus, instantaneous Coulomb and shear stress change apparently influence the locations but not the magnitudes of future earthquakes.

We compared Coulomb forecasts with smoothed seismicity forecasts. Smoothed seismicity works better near recently ruptured faults, while Coulomb works better at several km distance. Triggering effects of smoothed seismicity and Coulomb stress both decay with the age of the source. Likelihood tests show that Coulomb stress variation significantly alters aftershock rates from what smoothed seismicity models would forecast. An optimized combination of smoothed seismicity and Coulomb stress may show improved success in prospective tests.
Intellectual Merit The project makes several contributions not previously published. First, we examined the effect of many choices made in modeling the relationship between Coulomb stress change and earthquake occurrence. Those choices include which components of stress to include, what to assume about receiver fault orientation, what magnitude, time, and space limits to sample, when to start the clock for change calculations, etc. We found the results pretty tolerant of those choices. Second, previous studies generally asked whether earthquakes in the catalog were in areas of enhanced stress. We asked the reverse question, whether areas of enhanced stress had increased seismicity. The old way didn’t count areas where stress increased but seismicity didn’t. The dog that didn’t bark, so to speak. Third, we asked whether earthquakes in stress enhanced zones occurred at the time of maximum stress. Surprisingly, many such earthquakes occurred long after stress had peaked and began to decline.
Broader Impacts This project resulted from the PhD dissertation work of Dr. Anne Strader, who finished her PhD in 2014 and is now working with Dr. Danijel Schorlemmer in Potsdam, Germany.
Exemplary Figure Figure 1. Top: ΔCFF bounds for each stress spatial quantile in southern California, sorted by increasing ΔCFF. Each quantile contains 100 cells, 0.1 by 0.1 degrees per cell, not necessarily contiguous. Bottom: Fraction of cells within each quantile containing at least one earthquake. The orange line marks the seventeenth ΔCFF quantile, within which ΔCFF = 0. To the right of that line, the earthquake probability increases with increasing stress change, as we expected. To the left, the probability also increases with decreasing (more negative) stress, which we did not expect nor can explain.
Linked Publications

Add missing publication or edit citation shown. Enter the SCEC project ID to link publication.