Site response analyses are crucial for estimating local effects on ground shaking during earthquakes. However, recent investigations utilizing KiK-net borehole array data have revealed a consistent underprediction of high-frequency ground motion by bo...
Site response analyses are crucial for estimating local effects on ground shaking during earthquakes. However, recent investigations utilizing KiK-net borehole array data have revealed a consistent underprediction of high-frequency ground motion by both equivalent linear and fully nonlinear methods, contrary to expectations. This study reassess the accuracy of 1D site response analysis methods, including equivalent linear, frequency-dependent equivalent linear model, and nonlinear analysis, by integrating depth-dependent stiffness and adjusted shear wave velocity. Nine instrumented vertical arrays, featuring a total of 132 recorded ground motions, are subjected to analysis. For the sites and ground motions considered, the results indicate that the nonlinear method performs without significant bias, whereas the equivalent linear and frequency-dependent equivalent linear methods tend to respectively underestimate and overestimate high-frequency results. To enhance accuracy, it is recommendedto incorporate depth-dependent stiffness and adjusted shear wave velocity when predicting high-frequency ground motion in site response analyses.