在西元1999年9月21日,台灣的中部地區發生了一個芮氏規模7.3的大地震,稱為921集集大地震。在過去的研究中,我們發現了在大地震發生時地表會產生強地動,當地表土層較鬆軟時,強地動在地表土層就可能會產生場址的非線性現象。我們可以利用強震的強地動記錄來和參考值做比較來做為辨別場址非線性現象的一個方法,而我們的參考值通常是使用弱震產生的微地動記錄的平均值。我們用的方法是單站頻譜比法,它是利用單一個地表站的水平向和垂直向的地震記錄分別去做傅立葉轉換後得到頻譜,再用水平向和垂直向的頻譜相除,這個頻譜比就代表了這個測站的場址反應。葉與溫(2001)利用了這個方法去研究台灣集集大地震所造成的非線性反應,而本篇研究的目的是進一步對這樣的非線性反應做定量分析。Noguchi和Sasatani(2008)將強震和弱震的頻譜比之間的差異加總起來去量化2003年off-Miyagi地震所造成之非線性反應,我們將利用同樣的方法去量化集集地震造成的非線性反應,並嘗試不同的方法來做討論,所使用的資料為中央氣象局TSMIP所提供,計算出來的非線性度值(DNL, Degree of nonlinearity)將和觀測結果與場址來做比較以顯示場址環境對非線性的影響。On September 21, 1999, the Chi-Chi earthquake (ML 7.3) occurred in the central Taiwan. There are many studies have indicated that nonlinear site response will occur when strong ground motion wave shook surface soft soil layers. We can identify the nonlinear site response by comparing the strong ground motion record to the reference weak motion record. The method we used is the horizontal to vertical (H/V) spectral ratio of a single surface station. Yeh and Wen (2001) used this method to study the nonlinear site response during the Chi-Chi earthquake. The main objective of this study is to quantify the degree of nonlinearity (DNL). Noguchi and Sasatani (2008) used the summation of differences between the H/V spectral ratio for strong motion and that for weak motion to calculate the DNL value. The same and modified methods are tried to calculate the DNL value in this study. The DNL value will compare with the site classification and Vs30 to show the influence of site condition to the nonlinearity.