Revisiting Parkfield Mainshocks: A Characteristic Snowflake Model

Susan E. Hough, William L. Ellsworth, Aarnav Agrawal, Luke Blair, & Morgan T. Page

Submitted August 30, 2026, SCEC Contribution #15303, 2026 SCEC Annual Meeting Poster #TBD

We use early instrumental and macroseismic data to improve the characterization of nine moderately large earthquakes in central California, including seven events on the Parkfield segment of the San Andreas Fault and three earthquakes that may have been on the adjacent creeping section. We develop updated shaking intensity maps using available archival sources for earthquakes between 1857 and 1922, and data from postcard surveys of felt shaking for the 1934 and 1966 Parkfield mainshocks. The surveys included a standard macroseismic questionnaire, a forerunner of the questionnaire used by the US Geological Survey “Did You Feel It?” system. Our results confirm preferred moment magnitude estimates, M, of 6.0 ± 0.1 for the 1881, 1922, 1934, and 1966 mainshocks, and largest foreshock of the 1857 Fort Tejon mainshock. We estimate M 6.3 for the 1901 Parkfield mainshock. The shaking distributions reveal compelling evidence for rupture directivity, with northwest-propagating ruptures felt more strongly in the southern San Francisco Bay Area than in northern Los Angeles County, and the reverse for southeast propagating ruptures. We conclude the 1881, 1901, and 2004 mainshocks ruptured primarily to the northwest, while the 1922, 1934, and 1966 mainshocks ruptured primarily to the southeast. Our results support the characteristic earthquake model for the Parkfield segment, with the stipulation that M ~ 6 Parkfield mainshocks tend to be predominantly unilateral within the Parkfield segment, but do not always nucleate under Middle Mountain. Mainshocks have unique slip patterns and moment magnitudes close to but not always identical to 6.0. We also review available data for moderately large earthquakes in 1860, 1877, and 1885. We conclude that available evidence does not support the earlier suggestion that these events occurred on the creeping section of the San Andreas Fault. Lastly, we consider the report that dead pine trees were broken on Middle Mountain by the 1934 Parkfield mainshock, in the same location that severe shaking was recorded during the 2004 mainshock. This observation suggests that severe shaking in this location is robust across a variety of rupture scenarios.

Key Words
earthquake cycles, ground motions

Citation
Hough, S. E., Ellsworth, W. L., Agrawal, A., Blair, L., & Page, M. T. (2026, 08). Revisiting Parkfield Mainshocks: A Characteristic Snowflake Model. Poster Presentation at 2026 SCEC Annual Meeting.

Related Projects & Working Groups
Seismology