November 15, 2006
American Center for Physics
College Park, MD
No methodology has yet been found to reliably predict destructive fault ruptures on time scales of a decade or less. The field of earthquake prediction research has been relatively dormant since the mid-1990's, despite the general public's continued optimism. However, the field is now seeing renewed interest, fueled by improved observational data, computer power and understanding of earthquakes physics. Methods currently being explored include analyses of patterns of small earthquakes, anomalous changes in surface strain rate, accelerating moment release, ionospheric signals and tidal forcing.
The talk will provide a snapshot of recent trends in earthquake prediction research; viewpoints for and against the likelihood of viable short-term prediction; modern approaches to overcoming the difficulties in testing and vetting proposed methods; and a history of notable attempts made over the years. It will also summarize recent activities of the National Earthquake Prediction Evaluation Council-an expert committee that advises the USGS-and its California state counterpart.
Michael Blanpied received his PhD in Geophysics from Brown University in 1989, and immediately joined the United States Geological Survey (USGS), doing laboratory studies of fault sliding and earthquake nucleation in the Survey's Menlo Park, CA research center. Around the turn of the century he co-chaired a USGS-led working group that evaluated long-term earthquake likelihood in the San Francisco Bay region. In 2003 he moved to USGS headquarters in Reston, VA. He is currently Associate Coordinator, Earthquake Hazards Program, at the US Geological Survey, and Executive Secretary, National Earthquake Prediction Evaluation Council.