A Compact Probabilistic Representation of the Chi-Chi Earthquake Ground Motion - Reinforced Soil Engineering: Advances in Research and Practice - page 469

improved design standards for civil structure and to perform reliability studies.
A comparison is made of the preliminary results of this study and those obtained
from a similar analysis performed on an ensemble of ground motions from the
1994 Northridge earthquake.
1 INTRODUCTION
Given the proliferation of dense seismic arrays around the world, it is possible to
glean statistical information about the characteristics of major seismic events and
their potential effects on structural designs. An analysis procedure developed by
Masri et al. (1990), and later generalized in Smyth (1998) and Masri et al. (1998)
for the representation and transmission of random excitation processes, provides
a new tool to characterize strong ground motions from large data sets. For details
of this analytical compaction, representation, and transmission procedure, the
reader is referred to Masri et al. (1998). In summary, the method involves two
main stages of compaction of the random excitation data. The first stage is based
on the spectral decomposition of the covariance matrix by the orthogonal
Karhunen-Loeve expansion. The dominant eigenvectors are subsequently least-
squares fitted with orthogonal polynomials to yield an analytical approximation.
This compact analytical representation of the random process is then used to
derive an exact closed-form solution for the nonstationary response of general
linear multidegree-of-freedom dynamic systems.
2 THE ENSEMBLE OF CHI-CHI EARTHQUAKE GROUND
MOTION DATA
An ensemble of Chi-Chi earthquake ground motion data was gathered from the
extensive seismic network in Taiwan. Specifically, these are records from
stations denoted by “TCU” distributed around the Taichung region (in the west
coast of the central part of Taiwan) and also records from stations denoted by
“HWA,” which are from the Hwa-Liang area (east coast).This Chapter
presents the results from the 51 TCU stations. Because the records come from
a specific geographic region relative to the Chelungpu fault, they are treated as
statistically representing the ground motion in that region. It is probably
judicious not to mix records from areas that are too varied. A map of the TCU
seismic sensor locations in Taiwan is shown in Fig. 1 .
A representative sample of some of the time histories used to create the
ensemble are shown in Fig. 2 . The data was downsampled from the original