Various liquefaction vulnerability indices (LVIs) have been proposed to quantify the severity of liquefaction, and used to create the liquefaction hazard map. In producing the liquefaction hazard map, the spatial variability of LVI is important because the number of available in-situ data is typically limited, and the values at the unmeasured locations are predicted using kriging operations that consider the spatial correlation of observed values. In addition, the liquefaction hazard map may vary significantly depending on how LVI was spatially interpolated. This study aims to evaluate the spatial variability of soil properties based on cone penetration test (CPT) data, and to investigate the changes in geostatistical properties of LVIs depending on shaking intensity. A simplified CPT-based procedure was used to calculate the factor of safety against liquefaction, and the geostatistical properties of LVI according to shaking intensity were evaluated using the parameters of semivariogram. In addition, the LVIs at unmeasured locations were predicted using two geostatistical modeling approaches, and their results were compared to investigate the accuracy of geostatistical approaches in predicting LVIs according to the peak ground acceleration (PGA).

已经提出了各种液化脆弱性指数（LVI）来量化液化的严重性，并用于创建液化危害图。在生成液化危险图时，LVI的空间可变性很重要，因为可用的原位数据通常受到限制，并且未测量位置的值是使用考虑观测值的空间相关性的克里格操作预测的。此外，液化危险图可能会根据LVI的空间插值方式而有很大差异。这项研究的目的是根据圆锥渗透试验（CPT）数据评估土壤特性的空间变异性，并研究LVI的地统计特性随震动强度的变化。使用简化的基于CPT的程序来计算抗液化的安全系数，并使用半变异函数的参数评估根据振动强度的LVI的地统计特性。此外，使用两种地统计学建模方法对未测量位置的LVI进行了预测，并对其结果进行了比较，以根据峰值地面加速度（PGA）研究地统计学方法在预测LVI方面的准确性。