HR: 1330h
AN: S42B-0162 [PDF]
TI: ElarmS: Earthquake early warning for southern California
AU: * Allen, R M
EM: rallen@geology.wisc.edu
AF: University of Wisconsin, Dept. Geology and Geophysics
1215 W Dayton St, Madison, WI 53706 United States
AU: Kanamori, H
EM: hiroo@gps.caltech.edu
AF: California Institute of Technology, Seismological Laboratory
MC 252-21, Pasadena, CA 91125 United States
AB:
The location of faults directly beneath metropolitan areas in southern California requires any earthquake early warning
system to make use of the first arriving energy, the P-wave, to estimate the hazard potential of the earthquake. Using the
P-wave arrival maximizes the warning time available and may also make it possible to issue a warning before peak ground
motion at the epicenter where the warning is most needed. We present a methodology that could potentially provide an
earthquake early warning system for southern California. The necessary components include rapid detection and location of
earthquakes, magnitude estimation using the P-wave, and prediction of peak ground motion variations across the affected
region. As a result of the dense distribution of seismic stations in the region (TriNet), typically two or three stations
trigger within one second of the first station to detect an earthquake. This allows for accurate location of the event
within a second of the first trigger. Magnitude is estimated from the predominant period of the P-arrival at each triggered
station. At any given time after the P-trigger, the maximum period observed thus far is converted to a magnitude. This
magnitude represents a minimum estimate; a larger predominant period observation at some later time would indicate a larger
magnitude earthquake and the magnitude estimate would be updated. For smaller earthquake (less than magnitude 4.5) the
maximum period is usually observed within 1 sec, for larger events up to 4 sec of data is needed to capture the best
magnitude estimate. Given the location and magnitude estimates, the distribution of peak ground motion is predicted using
simple 1D attenuation relations. As time progresses observations of the peak ground motion at stations close to the
epicenter are used to adjust the attenuation relations and improve predictions of ground motion for more distant regions.
UR: http://www.geology.wisc.edu/~rallen/ELARMS/
DE: 7215 Earthquake parameters
DE: 7223 Seismic hazard assessment and prediction
SC: Seismology [S]
MN: 2003 Fall Meeting