Comparison of different site classification methods based on statistical analysis of strong ground motion records

The one-index site classification based on the equivalent shear wave velocity of rock and soil layers within a specified depth as the control parameter, and the two-index site classification based on the equivalent shear wave velocity of soil layers within a specified depth and the thickness of overlying soil layers as the control parameters, are the two major categories of site classification methods. The applicability and rationality of the two categories of methods have consistently been focal points of attention and discussion within the engineering community. The primary purpose of engineering site classification is to judiciously consider the site effects on strong motion to determine the seismic design ground motion parameters for construction projects. The degree of consistency in the spectral characteristics of strong motion records within the same classified site stands out as the most effective criterion for elucidating the rationality of the site classification methods. Utilizing strong motion records from the Japanese strong-motion seismograph network KiK-net and related site borehole data of the observation stations, the statistical analysis was conducted to delve into the distinctions and merits of the existing site classification methods. Initially, the site classification of the observation stations was carried out using the site classification methods including the specified in seismic design codes from countries such as China, the United States, and Japan, and several recently proposed ones, and then the collected strong motion records are grouped accordingly. Subsequently, statistical analyses of acceleration response spectra, including the peak acceleration, were performed for ground motion acceleration records from the observation stations with the same site class specially for different site classification methods. The study also scrutinized the consistency and clustering property of acceleration response spectra for strong motion records from the observation stations with the same site class across various site classification methods through the statistical residual analysis. The research findings reveal that, among the site classification methods studied, the site classification method of the Chinese seismic design code, a two-index site classification method based on the equivalent shear wave velocity of soil layers and the thickness of the overlying soil layers as the control parameters, is deemed as the optimal method.