Back to the deformation and Stress Change Modeling home page

Back to the team's online recent papers pages

Summary. This paper discusses the M6 earthquake which struck the San Andreas fault at Parkfield, CA on Sept. 28, 2004.  This earthquake was the sixth in a series of M6 earthquakes occurring in 1881, 1901, 1922, 1934, 1966, and now 2004.  Based on observations of the M6 earthquakes at this locale prior to the 2004 event, a prediction had been made that a M6 quake would occur in 1988 +/- 5 years.  As a result, the Parkfield prediction experiment was initiated, and this locale became one of the most heavily instrumented for seismic monitoring in the world. 

Unlike what is known about previous earthquakes at Parkfield, the 2004 earthquake initiated about 15 km to the southeast of previous epicenters, and slip was concentrated farther northwest than that of the 1966 event.  Despite M5 foreshocks preceding the 1934 and 1966 Parkfield earthquakes, and anecdotal reports of surface offset in the days prior to the 1996 event, no foreshocks or other precursory signals were observed associated with the 2004 earthquake.  A strong postseismic signal was observed in the Global Positioning System (GPS) and creepmeter measurements of ground surface displacement in the weeks to months following the earthquake.  Measurements of strong ground shaking in the Parkfield area showed large variation in the ground acceleration experienced by neighboring sites, and the highest recorded ground acceleration for California was observed during this earthquake.