Premium
Stress drops of repeating earthquakes on the San Andreas Fault at Parkfield
Author(s) -
Abercrombie Rachel E.
Publication year - 2014
Publication title -
geophysical research letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.007
H-Index - 273
eISSN - 1944-8007
pISSN - 0094-8276
DOI - 10.1002/2014gl062079
Subject(s) - seismology , geology , magnitude (astronomy) , earthquake magnitude , borehole , cluster (spacecraft) , spectral line , observatory , stress (linguistics) , seismic moment , fault (geology) , geodesy , physics , geometry , astrophysics , mathematics , linguistics , philosophy , geotechnical engineering , astronomy , scaling , computer science , programming language
Abstract I calculate well‐resolved corner frequencies and stress drops for 25 earthquakes (1989–2006) in the three repeating sequences targeted by the San Andreas Fault Observatory at Depth, using borehole data and multiple, highly correlated empirical Green's functions (EGFs). The earthquakes in the largest magnitude ( M ~ 2.1) cluster exhibit source spectra well‐fit by a circular source model. The corner frequencies correlate with those from the regional study by Allmann and Shearer ([Allmann, B. P., 2007]), suggesting that the interevent variability is resolvable. The earthquakes have stress drops between 25 and 65 MPa, with a gradual increase before the 2004 M 6 earthquake, followed by an immediate decrease, then a rapid return to previous levels. The spectra of the cluster of M ~ 1.9 earthquakes include high‐frequency energy not fit by simple source models and so stress drops are unreliable, and probably underestimated (1–20 MPa). There is no correlation with previous studies, and interevent variation is not resolvable. The earthquakes in the smallest magnitude cluster ( M ~ 1.8) have the highest corner frequencies, but similar stress drops (4–120 MPa). The stress drops exhibit the same temporal variation as the first cluster, but there is poor correlation with Allmann and Shearer ([Allmann, B. P., 2007]), probably because their frequency bandwidth is too limited.