The 2004 Parkfield Earthquake and the Parkfield Earthquake Prediction Experiment
Abstract
On September 28, 2004, a M6.0 earthquake on the San Andreas fault near Parkfield, California, highlighted the success of a two-decade scientific experiment designed to record the anticipated event. With extensive instrumentation poised to capture precursory, coseismic, and postseismic signals, a wealth of information has been collected. Studies have examined detailed geologic, seismic, electromagnetic, and geodetic observations and demonstrate a range of scientific progress. Although some of the detailed findings are specific to the behavior of the San Andreas in the Parkfield region, many of the results are applicable to understanding of earthquake physics and effects for strike-slip earthquakes and transform fault settings worldwide. Among the findings (also see Harris and Arrowsmith, BSSA, 2006): 1) Predicting the general timing and size of moderate and large earthquakes is difficult. 2) Magnitude 6 earthquakes can occur without detectable short-term precursors. 3) Ground motion variability is substantial in the near-field, and source, path, and site effects all play important roles in the variability. 4) Large ground motions, liquefaction, and slumping may result from a shallow earthquake even when there is minimal surface slip. 5) Earthquake rupture extent may be affected by fault rheology, but earthquake rupture direction may not be predictable on the basis of classic observables, such as material contrasts near faults. 6) Even with numerous types of observations, it may be difficult to reconcile a single highly-detailed deep-slip model for an earthquake, however a general picture can emerge that shows distinct slip features consistent with all of the observations. 7) The locations of subsequent smaller earthquakes (aftershocks) may not be predictable on the sole basis of stress changes due to the mainshock. For Parkfield, fault rheology also appears to play a key role in determining where small aftershocks and continuous microseismicity occur. 8) The San Andreas fault appears to be segmented in the Parkfield area such that the 2004 and prior events were contained in the expected fault region. 9) The postseismic moment release for the 2004 Parkfield earthquake was similar to that released coseismically. 10) Despite the complex details of Parkfield earthquakes and the San Andreas fault zone, the long-term (million year) evolution of the San Andreas near Parkfield can be understood by modeling the basic interaction between the creeping and locked portions of the fault.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2006
- Bibcode:
- 2006AGUFM.S23C0164H
- Keywords:
-
- 1242 Seismic cycle related deformations (6924;
- 7209;
- 7223;
- 7230);
- 7212 Earthquake ground motions and engineering seismology;
- 7215 Earthquake source observations (1240);
- 7223 Earthquake interaction;
- forecasting;
- and prediction (1217;
- 1242);
- 8118 Dynamics and mechanics of faulting (8004)