對於土壤液化潛能的影響主要取決於兩大因素，分別為地質因子與大地工程因子，地質因子主要為地震規模與最大水平地表加速度(Peak Ground Acceleration, PGA)；大地工程因子主要為土壤的組成與其強度。 然而，在這兩大因素皆受到不確定性所影響，例如：地質因子中的PGA，會因為對地面運動衰減的認知不確定，而造成預估值與實際觀測值產生落差；大地工程因子中的標準貫入試驗(Standard Penetration Test)之N值，會因為機械的缺失、人員的操作失誤等，造成其與真實值不同。並且，土壤液化評估法中的參數受到不確定性所影響，將對土壤液化評估結果造成不同程度的影響。 本研究將對液化評估法進行基於變異數的敏感度分析 [1]，採用台北以及高雄地區之真實土層與斷層資料，並搭配蒙地卡羅模擬(Monte Carlo Simulation, MCS)來進行計算，計算出各個參數其相應的敏感度，加以找出液化評估法中最為重要之參數。 研究結果顯示，在土壤液化評估法中，相較於大地工程因子的不確定性，地質因子的不確定性對於結果之影響較大。並且，將地質因子與大地工程因子再細分為各個參數後，得到之結果顯示，相較於其他參數，地質因子之PGA對於土壤液化評估法之影響最大。此結果表明，不論投入多少努力，以求獲得更加準確的大地工程參數，若無法降低地質不確定性，液化評估結果將依然存在很大的不確定性。 ;The impact on soil liquefaction potential primarily depends on two major factors, in this study calls geological factors and geotechnical factors. Geological factors mainly include seismic magnitude and peak ground acceleration (PGA); geotechnical factors mainly involve soil composition and strength. However, both of these factors are subject to uncertainty. For example, in geological factors, uncertainties in PGA arise from uncertainties in the attenuation of ground motion, leading to disparities between estimated and observed values. In geotechnical factors, the N-value of the standard penetration test can differ from its true value due to mechanical failures, operational errors, and other factors. Additionally, variations of parameters in soil liquefaction assessment methods result in varying degrees of impact on the assessment results. This study conducts a variance-based sensitivity analysis [1] to quantify geological or geotechnical uncertainties on the result of liquefaction analysis. It utilizes real soil and fault data from Taipei and Kaohsiung regions, using Monte Carlo simulation (MCS) to compute the variance of liquefaction safety factors. The sensitivity of each parameter is calculated to identify the most important parameters in the liquefaction assessment method. This study found that uncertainties in geological factors have a greater impact on the results of soil liquefaction assessment. Upon further subdivision of the two factors into individual parameters, results show that the uncertainty in the estimated PGA has the greatest impact on the soil liquefaction assessment. This suggests that if geological uncertainties cannot be reduced, significant uncertainties will be present in liquefaction assessment results.
