|
|
 |
|||||||||||||||||
|
|||||||||||||||||
 |
|||||||||||||||||
| JOURNAL HOME | HELP | CONTACT PUBLISHER | SUBSCRIBE | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Article |
The 1999 Chi-Chi, Taiwan, earthquake (Mw 7.6) struck central western Taiwan at 1:47 a.m. local time (on 21 September or at UTC 17:47 20 September). The causal fault was immediately identified to be the mapped active Chelungpu thrust fault as it produced a remarkable 100-km-long surface break, with fault scarps as high as 8 m in places. The nucleation point of this long rupture was defined by an epicenter at 120.82 °E and 23.85 °N, with a focal depth of 8 km. A first-motion solution was reported to be striking N20 °E, dipping 30 °SE with an average rake of 85° (Chang et al., 2000). The event was officially named after the nearest town Chi-Chi by the Central Weather Bureau (CWB) (Shin, 2000). This was the most devastating earthquake to hit Taiwan in modern times. The Chi-Chi earthquake and its energetic aftershock sequence inflicted a casualty toll of 2435 and an estimated US$ 4 billion property loss (Tsai et al., this issue). This catastrophe was nonetheless fortunate in several aspects. The mainshock struck when schools were not in session; the loss of student lives would have been unthinkable in view of the total collapse of hundreds of school buildings. In emergency response, it was fortunate that the CWB completed in 1995 the Taiwan Rapid Earthquake Information Release System that was capable of electronically releasing pertinent information within a couple minutes of an earthquake occurrence. Rapid earthquake information greatly facilitated the disaster relief missions, and it mitigated further secondary losses. Scientifically, because of the completion of the Taiwan Strong-Motion Instrumentation Program (TSMIP), the occurrence of the Chi-Chi earthquake sequence has resulted in the richest seismological data recovery in the world to date, in both quality and quantity. These data were open to the world with a data CD of the Chi-Chi mainshock distributed by the CWB three months after the event (Lee et al., 1999). This data release has significantly helped foster international research collaborations. Many of the results are collected in this issue and are grouped by dominant subject: overview, fault geology and seismotectonics, crustal studies, source studies, strong ground motions and impact on society, earthquake engineering analysis, preprocessing of the Chi-Chi data, and CD-ROM attachments.
To begin with, Shin and Teng give an overview of the Chi-Chi earthquake from the observational standpoint. The seismic instrumentation program in Taiwan and its recent accomplishments on earthquake rapid reporting are reported. A movie is constructed, based on the TSMIP data from more than 400 strong-motion stations. This movie documents the nature and evolution of the rupture initiation, wave generation, propagation, and attenuation, and shows that the rupture velocity is clearly not uniform over a rupture surface that is nonplanar. A jumping dislocations model is proposed to describe the observed rupture process.
There are seven articles on fault geology and seismotectonics. Lee et al. report a trenching study across the causal Chelungpu fault, giving updated paleoseismic information. Kelson et al. report on the geomorphic deformation style along the Chelungpu fault and show how surface ruptures respond to man-made structures. Huang et al. investigate the surface-normal acceleration and discuss a possible mechanism for the initiation of large landslide blocks kilometers in size. As the NS-trending Chelungpu rupture grows northward, it terminates into a NE-trending jog of strike-slip motion; Lin et al. discuss the mechanism that might control the initiation and termination of a rupture. Ouchi et al. correlate damage from the Chelungpu rupture and the strong ground motions generated. They point out that damaged areas appear to cluster in patches fractions of kilometers in size, and, similar to the observation of Kelson et al., that large building sites sometimes modify the surface break. Y.-G. Chen et al. report the surface measurements on the fault scarp and find that the Chelungpu fault is quite segmented and has varying slip vectors over its length. W.-S. Chen et al. give detailed fault slip measurements and discuss the variation of stress field in different rupture segments.
Six articles are devoted to crustal studies. Yu et al. have carried out a long-term crustal deformation program in Taiwan, using both classic geodetic and GPS networks. They report important measurements on preseismic deformation, coseismic displacements, and postseismic slips. This article gives perhaps one of the most complete sets of crustal deformation observations of a major dynamic period in Taiwan. C.-H. Chen et al. give 3D Vp and Vp/Vs models from a tomographic inversion for the source area using data before and after the Chi-Chi earthquake. They report that the crust is highly heterogeneous, but good correlations can be found between damage fault zones and zones of low Vp and high Vp/Vs in the upper crust. Their results also show significant evolution of the upper crustal properties in response to the Chi-Chi event. Kao and Angelier performed an extensive stress tensor inversion analysis and discuss the seismotectonics in Taiwan in terms of the regional plate collision. Wang and Chen performed a static deformation analysis and discuss the effects on the stability of the surrounding fault systems due to the stress transfer induced by the Chi-Chi event. Based on the distribution of hypocenters before and after the Chi-Chi event, Lin proposes a crustal wedge model and speculates that stress focusing may have defined the eventual Chi-Chi rupture surface. For ground water conservation and management, Taiwan has constructed extensive networks of observational water wells. Chia et al. recovered and reported excellent water-level recordings with sharp impulsive responses over the Choshui River alluvial fan on the footwall of the Chelungpu fault. Careful examination of the data shows that the responses are coseismic. The varied amplitude responses reveal the rapid change of permeability in shallow aquifers.
The section on source studies includes 10 articles. This large number reflects both a broad interest in this subject and the availability of many data types, including acceleration data from the TSMIP program, velocity data from local short-period and broadband networks, displacement data from both GPS and ground-truth fault scarp measurements, and data from the Global Seismic Network (GSN/IRIS). These articles employ a range of computational methods, make different rupture model assumptions, use different crustal velocities (therefore Green's functions), group different combinations and sizes of the data sets, and apply varied bandwidths of the data that are processed. There are broad similarities in their research approaches, and their basic findings show a similar overall picture of the Chi-Chi rupture source, but there are significant differences in details of the outcome models. As data of better quality and more thorough coverage are used, the resolution of the source inversion correspondingly improves. Improved resolution not only reveals more detailed features of the resulting source models, it often raises new questions about some fundamental assumptions on the source models and on the adequacy of the crustal structures used. Among these articles, Ma et al. use near-source (TSMIP) and telesismic (GSN/IRIS) data as well as GPS data. Three groups of researchers (Zeng and Chen, Oglesby and Day, and Wu et al.) use TSMIP and GPS data sets of varying sizes. Chi et al. emphasize the need for a composite rupture plane; authors of the previous four papers share this viewpoint. Dalguer et al. attempt to use a 2D discrete element model to study the differences of the rupture process between the northern and southern sections of the Chelungpu fault. Huang et al. study the stress drop along the fault rupture. Wang et al. derive the source slip distribution from surface displacement measurements, while Yoshioka derives the source slip distribution from the GPS data. Huang et al. study the rise time and slip velocity based on the observed near-source spectrums.
Nine articles discuss strong-motion data analysis. Boore clarifies an old problem with new data on the instrumental drift effect due to strong shaking and gives examples of integrated velocity and displacement waveforms. The baseline correction method Boore suggests is widely used in articles in this issue for integrated velocity and displacement calculations. In a second article, Boore reports that the Chi-Chi ground motions can be adequately predicted by the empirical relationship largely based on the California data. Wu et al. use large data sets from Taiwan and devise a method of obtaining near-real-time PGA and PGV maps for future large earthquakes in Taiwan. Chang et al. discuss the attenuation of PGA in Taiwan. Wen et al. examine the site conditions of station TCU129, which has recorded anomalously high PGA. Lee et al. use the surface geology to classify the site conditions of the entire 650 TSMIP stations. Their primary results are included in a file on the CD-ROM attached to this issue. Chen et al. report the near-field observations of rupture pulses from the Chi-Chi mainshock seismograms. Using a short-period seismic array and a single-station microtremor method, Satoh et al. report a 3D shear-wave velocity for the Taichung basin, where heavy damage and soil liquefaction were extensive. From thorough governmental statistics, Tsai et al. discuss the spatial distribution and age dependency of the fatality rate during the disastrous Chi-Chi event. This study is unique because Taiwan keeps excellent household registration data and the distribution of the postdisaster governmental relief fund requires accurate casualty statistics. The result casts interesting light on the demographic probability of injury or death during an earthquake catastrophe.
Two articles deal with a new data processing technique that is potentially useful in nonlinear analysis. Huang et al. outline the basic methodology. Loh et al. apply the new technique to identify the dynamic properties of structures under strong shaking. Wang et al. give some stochastic properties of the Chi-Chi strong-motion response spectra.
Finally, W.H.K. Lee coordinated a CD-ROM with data files related to this issue, principally the Chi-Chi free-field strong-motion data. Also included are the Taiwan broadband data, short-period network data, other miscellaneous free-field strong-motion data, Taiwan crustal structures used by different authors and their derived models, the Chi-Chi movie, and long tables. A short note section includes one-page readme-type articles for each file on the CD-ROM. The large volume of materials on the CD-ROM should be useful for future research. Clearly, this issue marks only the beginning of a research program investigating the details of the Chi-Chi earthquake, which so far is the most thoroughly observed earthquake ever, and may remain so for a long time.
 |
References |
|---|
 Top References |
|---|
Lee, W. H. K., T. C. Shin, K. W. Kuo, K. C. Chen (1999). CWB free-field strong-motion data from 921 Chi-Chi earthquake: Volume 1. Digital acceleration files on CD-ROM, Prepublication version (6 December 1999), Seismology Center, Central Weather Bureau, Taipei, Taiwan.
Shin, T. C. (2000). Some seismological aspects of the 1999 Chi-Chi earthquake in Taiwan, TAO 11, no. 3, 555-566.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| JOURNAL HOME | HELP | CONTACT PUBLISHER | SUBSCRIBE | ARCHIVE | SEARCH | TABLE OF CONTENTS |
