A STUDY ON THE GROUND MOTION CHARACTERISTICS OF TAIPEI BASIN, TAIWAN, BASED ON OBSERVED STRONG MOTIONS AND MEASURED MICROTREMORS

Transcription

1 A STUDY ON THE GROUND MOTION CHARACTERISTICS OF TAIPEI BASIN, TAIWAN, BASED ON OBSERVED STRONG MOTIONS AND MEASURED MICROTREMORS Ying Liu 1, Kentaro Motoki 2 and Kazuo Seo 2 1 Eartquake Engineer Group, OYO Corporation, Tsukuba, Japan 2 Department of Built Environment,Tokyo Institute of Tecnology,Yokoama, Japan 2 Professor, Department of Built Environment,Tokyo Institute of Tecnology,Yokoama, Japan ABSTRACT : ryu-ei@oyonet.oyo.co.jp, kmoto@enveng.titec.ac.jp,seo@enveng.titec.ac.jp Taipei basin suffered serious damage on building structures during te 1999 Ci-Ci eartquake, even toug it was located 0 km away from te epicenter. Te purpose of tis paper is to discuss te relationsip between te damage occurred and te ground motion caracteristics observed in te Taipei basin and to draw a seismic microzoning map for te region by taking into account te analyzed results about te eartquake motions, microtremors and te damage distribution. For more discussion about site effect evaluation, observed strong motions of two eartquakes, wic appened in 1994, 1995 will be used in addition to te 1999 eartquake. Suc records of strong motions of tose eartquakes were offered by te Central Weater Bureau (CWB, Taiwan. Measurements of microtremors inside Taipei basin were made in 2000 after te Ci-Ci eartquake. KEYWORDS: Ground Motion Caracteristics, Taipei Basin, Seismic Microzoning, Te 1999 Ci-Ci eartquake, Building damage 1. BACKGROUD AND PURPOSE Strong motion sites Strong motion & microtremor sites A TAP Holocene Pleistocene Pleistocene extrusives 35 Miocene 3 Oligo A' 5 71 Figure1 Geological condition and Location of CWB strong motion stations around Taipei basin

2 Taiwan, located in te western part of Circum-Pacific seismic zone, often suffers eartquake disaster. About 1/3 rd of te population of Taiwan concentrates in Taipei basin, and te urban development is expending from Capital Taipei to te wole of Taipei basin. So it is very important to make sure te relationsip between eartquake damage distribution and te ground motion caracteristics of Taipei basin. Taipei basin (Figure 1 as a triangle saped alluvium structure, and te area (about 240 square kilometers is almost flat wit an altitude lower tan 20 meters. Te geological structure inside te basin (Figure2 consists of te Quaternary layers, wic were classified into four stratigrapic formations (Deng, 199 above te Tertiary base rock. Figure 2 A-A section Regarding te previous studies on Taipei basin, Wen et al. (1995, 2000 and Sokolov et al. (2000 ave discussed on te seismic amplification of te basin. Tsai et al. (1999 identified Taipei Basin into 4 groups using te sapes of response spectral ratios according to te predominant frequencies. In tis study, te measured microtremors (Figure 1 and observed strong motions of tree eartquakes, tose appened in 1994, 1995 and 1999 (Table 1, Figure 3 ave been used, considering site 3, wic is on te ard rock in Taipei region as reference point, to classify te site caracteristics of Taipei basin Kilometers 0 Figure 3 Location of epicenters and te objective region Table 1 Focal information of te eartquakes appened in 1994, 1995 and 1999 EQ Origine Time Lon Lat Dept ( ( (km M L /0/0 09: /0/25 14: /09/21 01: GROUND MOTION CHARACTERISTICS USING STRONG MOTIONS AND MICROTREMORS 2.1. Ground Motion Caracteristics of Strong Motions For te evaluation of ground motion caracteristics in te basin, it is necessary to remove te source and propagation pat effects. Ten we considered Site 3 wit te ardest site condition as te reference point in

3 tis study. Figure 4 sows te response spectra of strong motions at te reference point Site 3. Figure 1 sows te sceme of te Taipei Basin and sites of Taipei strong motion observation network tose were used in tis study. Figure 3 sows te epicenters of eartquakes and te objective region. Table 1 sows te focal information of te eartquakes tose appened in 1994, 1995 and Figure 4 Te response spectral of strong motion at Site 3 As Taipei basin suffered serious damage during te 1999 Ci-Ci eartquake, it is important to investigate te ground motion caracteristics due to te 1999 Ci-Ci eartquake data. Considering Site 3 as te reference point, te response spectral ratios of te 1999 Ci-Ci eartquake were obtained as Figure 5. Following to different surface geologic conditions, suc as aard rock, bbasin edge, csoft rock wit tin sediments, dtin sediments and etick sediments regions, we identified observed strong motion caracteristics into 5 regions (Figure 5. Figure 5 Identification of te response spectral ratios using te 1999 eartquake data (=5% Similar identification was also done using te 1994 and 1995 eartquakes. Figures and 7 sow te results of identified response spectral ratios using te 1994 and 1995 eartquakes, respectively. Figure Identification of te response spectral ratios using te 1994 eartquake data (=5%

4 Figure 7 Identification of te response spectral ratios using te 1995 eartquake data (=5% Te group of respective sites are categorized as following. a Hard rock region (34, 7, 71, 8,98 b Basin edge region (2, 7, 9, 29, 30, 35, 43, 48, 53, 91, 93, 94, 95 c Soft rock wit tin sediments region (15, 27, 32, 33, 42, 51, 52, 87, 88, 89,90, 92 d Tin sediments region (4, 5,, 8, 12, 13, 14, 19, 22, 23, 24, 2, 28, 31 e Tick sediments region (3,, 11, 1, 17, 20, 21, 25, 37, 38, 99, 0, TAP Te predominant peak period around 2.0 seconds can be seen in all regions during te 1999 Ci-Ci eartquake. It is probably because of te surface wave caused by te deep sediments in te basin. In te 1994 eartquake, 1.0 second peak period can be seen at te basin edge and soft rock wit tin sediments regions. On te oter and, 0.35 seconds and 0.8 seconds componentsare obvious in te response spectral ratios of te 1995 eartquake Ground Motion Caracteristics of Microtremors We measured te microtremores along te section A-A of Taipei basin (Figure 1. Because we want to try te similar identification of site conditions wit microtremors as well as strong motions. Were te caracteristics of microtremors will be evaluated using H/V spectral ratios. Tey were classified into te following tree groups. 1 Soft rock wit tin sediments region (15, 52, 87, 88 2 Tin sediments region (12, 13, 19, 22, 31, 97 3 Tick sediments region (3,, 11, 1, 17, 20, 21, TAP Here te groups 1, 2 and 3, classified wit microtremors are in correspondence wit te group c, d and e identified wit strong motions, respectively. Figure 8 Classification of microtremors H/V spectral As sown Figure 8, we can see quite different ground motion caracteristics on microtremors. For example, tey look very flat in te soft rock wit tin sediments region. Tere is no sarp peak in tick sediments region, and 1.0 second peak period can be seen very clearly in tin sediments region.

5 Figure 9 Spectral comparison of microtremors wit eartquakes at Site 3 Figure 9 sows te comparison of response spectral ratios among te 1994, 1995, 1999 tree eartquakes response spectral ratios and te H/V spectral ratio of microtremors at Site 3. It was observed tat te H/V spectral ratio of te microtremors were almost similar to te response spectral ratios of strong motions, especially similar to tat of te 1994 eartquake. 3. COMPARISON OF THE DAMAGE DISTRIBUTION AND THE GROUND MOTION CHARACTERISTICS IN TAIPEI BASIN DURING THE 1999 CHI-CHI EARTHQUAKE According to Te first report of te building damage during te 921 Ci-Ci eartquake (1999, te most serious damage was found along te seismic fault regions of course, but te second serious damage appeared in Taipei basin, wic was more tan 0km away from te epicenter of te 1999 Ci-Ci eartquake. It is reported tat 500 buildings were damaged inside te basin. On te oter and, tere was almost no damage outside te basin (cm/s Floor Complete Damage Floor Partial Damage Floor Complete Damage Figure a Comparison of te distribution of damaged 1-3,4- floor buildings and te average response velocities in seconds period during te 1999 Ci-Ci eartquake TAP

6 TAP (cm/s Floor Partial Damage Figure b Comparison of te distribution of damaged 12-13,17 floor buildings and te average response velocities in seconds period during te 1999 Ci-Ci eartquake From te response spectral ratios during te 1999 Ci-Ci eartquake, peak period of 0.4 seconds can be seen in every region. So we tink tat te serious damage appened because of suc component. Figure a sows te comparison of damage distribution on te 1-3, 4- floor buildings and period average response velocities of te strong motions in seconds period. Similarly, peak period of 2.0 seconds can be seen too. Ten, Figure b sows te comparison of damage distribution on te 12-13, 17 floor building and period average response velocities of te strong motions in seconds period. As we can see in Figures a and b, te relationsip between te building damage distribution and te distribution of related response velocities are almost consistent eac oter. 4. DISCUSSION AND CONCLUDING REMARKS As tere are a lot of ig buildings in te basin, we can assume tat te strong motions of around 2 seconds period will affect tose buildings. We draw te amplification contour of te average response spectral ratio in te period ranges of 1.0 ~3.0 seconds during te 1999 Ci-Ci eartquake, considering Site 3 as reference point (Figure 11. We can see te maximum amplification reaced to times around Site 3. So it is te most vulnerable area for te ig buildings in Taipei basin, in te case tat a very large and sallow eartquake occurs around te area. Sokolov et al. (2000 tougt te seismic network in Taipei basin, despite of large number of te stations, couldnot provide a reference site wic is suitable to be accepted, undoubtedly, as ard-rock reference site. So tey modeled a ypotetical very ard rock site of te 1999 Ci-Ci eartquake in Taipei basin by te single-corner frequency Brune ω -2 source model. Tey Considered te very ard rock site as te reference point, te contour for distribution of te maximum amplitudes of te amplification in te period ranges 1.0 ~3.0 seconds during te 1999 Ci-Ci eartquake, was drawn. Compared wit teir study, te amplification distribution obtained in our researc is almost consistent Floor Partial Damage Floor Complete Damage

7 TAP Figure 11 Contour of average response spectral ratios for te period range 1.0~3.0 seconds in Taipei basin, using te 1999 Ci-Ci eartquake data Group a Group b Group c Group d Group e 25.0 Figure 12 Holocene Pleistocene Pleistocene extrusives Miocene Oligo Seismic microzoning map of Taipei basin We tried to draw te seismic microzoning map of Taipei basin (Figure 12, considering te observed strong motions and measured microtremors, and also te damage distribution in Taipei basin during te 1999 Ci-Ci eartquake. We identified different site conditions into 5 groups. Te caracters of respective groups can be summarized as follows. Group a on te ard rock, in te east of te basin: Te response spectral ratios are almost flat and te amplification is very low.

8 Group b on te edge of basin: Te sort peak period of 0.4 seconds is obvious. Group c on te soft rock wit tin sediments: Wen a deep eartquake occurs, te peak period of 0.8 seconds will be observed. Wen a sallow eartquake occurs, te peak period will become muc longer. If te scale of eartquake is larger, te peak period will be longer. Group d on te tin sediments: Wen a deep eartquake occurs, te amplification will be small. Wen a sallow eartquake occurs, te amplification will be big. Group e on te tick sediments: In case of a sallow eartquake, long peak period of 2.0 seconds will be resonant. Wen a deep eartquake occurs, sort peak period of 0.35 seconds will be obvious. According to Tsai et al. (1999, te objects are 52 sites in te Taipei basin and te strong motion data were obtained from more tan 20 eartquakes. Taipei basin was classified into 4 groups by response spectral ratios. Te peak periods of Group1, Group2, Group3, Group4 are 2 seconds, 1.5 seconds, 0.15 seconds and 0.9 seconds, respectively. Compared wit tis study, te amplification distribution obtained in our researc is almost consistent. In conclusion, we found tat te sediments regions inside te basin will resonate around 2.0 seconds period, wen a large and sallow eartquake occurs even far from Taipei basin. So tose buildings, wic ave a predominant period around 2.0 seconds, will be in danger. Te application of microtremors is an effective metod for compensating for te microzoning as microtremors are in correspondence wit strong motions. ACKNOWLEDGMENTS We igly appreciate Dr. Yi-Min Wu of Central Weater Bureau for offering strong motion data set and cooperating wit us in te measurement of microtremors. Prof. Yi-Ben Tsai and Prof. Kuo-Liang Wen of National Central University provided us wit a lot of te materials and information. We gratefully acknowledge teir elp and guidance. REFERENCES WEN, K. L., PENG, H. Y., LIU, L. F., and SHIN, T. C. (1995. Basin effect analysis from a dense strong motion observation network. Eartq. Eng. Struct. Dyn.24, WEN, K. L., PENG, H. Y., CHANG, C. L., and LIU, L. F. (2000. Site response in te Taipei basin from te 1999 Ci-Ci, Taiwan eartquake sequence. Proc. Of International Worksop on Annual Commemoration of Ci-Ci Eartquake,September 18-20, 2000, NCREE, Taipei, vol. 1, SOKOLOV, V., OVCHARENKO, A., LOH, C. H., and WEN, K. L. (2000. Seismic Hazard Assessment for te Taiwan Region on te Basis of Recent Strong-Motion Data and Prognostic Zonation of Future Eartquakes, National Center for Researc on Eartquake Engineering, Taiwan. DENG, S. Y. (199 Investigation Researc of Underground Geological Features and Engineering Environment around Taipei basin, Te report of Central Geological Survey, MOEA, TSAI, Y. B (1999 Te Crisis of basin, wen eartquakes appen in Taipei, THE EARTH. 141, 38. ARCHITECTURE AND BUILDING RESEARCH INSTITUTE (1999. Te first report of te building damage 921 Ci-Ci eartquake, Ministry of te Interior, Taiwan.