本研究主要是利用有限斷層模型探討台北盆地之場址效應。利用有限斷層模型計算所有屬於基盤站的地動加速度，與觀測值得到良好的擬合後，便可利用此校準後之模型計算沖積層測站在不考慮場址效應下之地動加速度，比較觀測值與模擬值之頻譜得到兩者水平向傅氏頻譜比值。如此便可探討台北盆地內所有強震站位置的場址放大效應。從模擬出來的結果可以看出，五分山雷達測站(TAP103)屬於岩盤站(Site class B)，但具有明顯主頻且放大倍率偏高，顯示此測站有必要深入研究造成震波放大的原因。而盆地內主頻低區分佈在五股、蘆洲、士林、松山及大安一帶，主頻為1.1Hz以下，而盆地周圍主頻則為1.7~2Hz或更高。經由TSMIP H/H Ratio 及H/V Ratio資料佐證，台北主要共振主頻在0.5~1.5Hz，只有信義區及樹林附近對2 Hz以上有明顯放大。而從模擬集集主震PGA的過程顯示，集集地震震波在台北盆地的放大應該有二維或三維效應存在，且單站頻譜比法比值所得之倍率在部分地區較不準確，但對淺層土層的共振頻率反應敏銳。 In this study, the method of the stochastic finite-fault modeling is used to analyze site effect in the Taipei basin. Local site effects are considered by soil/rock spectral ratios calculated as ratios between spectra of chi-chi earthquake record and finite-fault modeled for rock site. The dominant frequency and amplification are the indicator clearly show that local site effect. The data records at 14 Site class B TAP stations are aimed the model calibration. Simulation and spectra ratio analysis indicate that TAP103 is an unusual rock site which is recorded large ground motions and may be caused by geography effect. In Taipei basin the Dominant frequency range from 0.5Hz to 1.5Hz, but in Xinyi and Shulin are about 2.0Hz, higher than other sites. The calibrated model is then used to simulate chi-chi earthquake ground motion on soil site with average transfer function and main shock transfer function that estimated by the spectra-ratio technique. The result shows that the H/V technique no absolute response the chi-chi earthquake amplification effect, revel there are 2D-3D effects influence the site effects.
