經由過去許多地震學的研究，已經認知到地震動會受到土層狀況不同而產生差異，此種現象稱之為場址效應。場址效應往往造成某些區域受災特別嚴重，因此在工程場址的選擇時，必須要對場址的特性有進一步的瞭解。在探討場址效應時，最主要的工作之一就是要確認場址的土層性質，包含質量密度、土層厚度和剪力波速剖面等。本研究提出利用重力異常資料與微地動資料的聯合反演方法來獲得地層性質。首先將地層假設為兩層模型一為土層一為岩盤，給定假設密度，利用重力異常資料來推估大略的土層厚度。接著利用微震動資料來推估地層可能層數並利用之前給定的密度及獲得的厚度進一步進行反演，運算結束後可得到新的各層密度及厚度。由於此時的密度與厚度比先前假設的更為精確，故最後再利用重力異常資料以及獲得的密度再一步進行反演得到各層土層厚度，我們將此土層模型視為當地土層狀況。利用此種方法進行許多的數值案例，我們將這種方法的限制做個小結: 1)淺層地層處的精確度高於深層地層 2) 在地盤有突然性的形狀變化中，當厚度增加、角度變小時，誤差隨之增大 3) 在地盤有突然性的形狀變化處，利用此種方法無法在該處得到精確的反演結果 4) 在進行實地量測時，可辨識岩盤變化能力為三倍量測點間距，若岩盤變化寬度小於此數值，反演出來的準確度不高。In most previous studies relating to the estimation of underground structure, the underground structure is assumed to be horizontally layered, which confines the estimation accuracy.?Hence, we propose a method combining the merits of gravity anomaly and microtremor survey, which are usually used to identify the ground structure.?In this method, at first we assume the gravity anomaly for 2-layer model to get the approximate thickness of the depth of the entire ground structure.?Using the estimated depth, the results of microtremor survey is applied to specify the ratio of thickness of each layer.?Finally, gravity anomaly is used again to obtain more realistic density and thickness of the site.? The validity of the proposed method is examined using numerical examples.?The conclusions drawn from this study are: (1) The accuracy of estimation for shallow region is higher than that for the deep region. (2) For the case of irregular bedrock, the error in the estimation increases with increasing depth and decreasing angle of bedrock. (3) It is difficult for the proposed method to provide enough accuracy for the region near abrupt change in the shape of layer interface. (4) The capability for identifying the shape of bedrock is three times of the distance between two observation sites. It means this method cannot identify the shape of bedrock less than this range.?
