HR: 0800h
AN: S11A-0279 [Abstracts]
TI: Study of the Probability Law Governing Ground Motion Metrics Recorded During the 2004 Parkfield Earthquake
AU: * LAVALLEE, D
EM: daniel@crustal.ucsb.edu
AF: University of California, Santa Barbara, Institute for Crustal Studies
1140 Girvetz Hall
MS 1100, Santa Barbara, CA 93106, United States
AB:
Based on the superposition of seismic waves and the Central Limit Theorem, we developed the basis for a
unified picture of earthquake variability from its recording in the ground motions to its inference in source models.
According to this theory, the random properties of the ground motions and the source for a single earthquake
should be both (approximately) distributed according to the Levy law. Computation of the probability density
function (PDF) of the peak ground acceleration (PGA) of the 1999 Chi-Chi, the PDF of the PGA and the PDF of the
peak ground velocity (PGV) of the 2004 Parkfield earthquakes confirms this theory. As predicted by the theory, we
found that the tails of the PDF, characterizing the slip and the PGA, are attenuated according to power laws with
exponents (denoted Levy indexes) that take almost the same values close to 1. Computations of the PDF of the
PGA recorded at the surface and the PDF of the PGA recorded in borehole during the 2003 Tokachi-oki
earthquake lead to a similar conclusion. The PDF tail measures the frequency at which large events occurred and
thus quantifies the probability to observe large acceleration values and large velocity values during an
earthquake.
We extend our analysis of the random properties to other ground motion metrics. To lessen the dependency due
to the source-to-site distance, we consider the ratio of the PGV to the PGA, the ratio of the two horizontal
components of the PGA to the vertical component of the PGA, and the ratio of the horizontal components of the
PGV to the vertical component of the PGV. In this analysis, we use the ground motions recorded during the 2004
Parkfield earthquake, arguably the best-recorded earthquake in history for the density of near-source data. We
select stations located within a closest distance to the rupture surface that varies from 0 to 180 km. To test the
effect of the distance on the computed random properties, these stations are divided into several subsets or
windows. For each subset of stations, we compute the PDF and characteristic function (CF) of the ground motion
metrics and compile the parameters of the Levy law that best fit the PDF and CF. We find that the tails of PDF of
ground motions metrics decrease with power law behaviors controlled by Levy indexes with values close to 1.
DE: 3265 Stochastic processes (3235, 4468, 4475, 7857)
DE: 4468 Probability distributions, heavy and fat-tailed (3265)
DE: 7212 Earthquake ground motions and engineering seismology
DE: 7223 Earthquake interaction, forecasting, and prediction (1217, 1242)
DE: 7260 Theory
SC: Seismology [S]
MN: 2007 Fall Meeting