Abstract:

Site shear wave velocity is a pivotal parameter for site classification and for quantitatively assessing the site's impact on ground motion. It has extensive applications in engineering seismic design and rapid post-earthquake damage assessment. China's seismic design standard, GB50011-2010, primarily uses two indicators for site classification: the thickness of the soil layer and the equivalent shear wave velocity of the top 20m of soil. In contrast, the United States and Europe utilize the average shear wave velocity, V_S30, at a 30m depth for site classification. Studies have indicated that considering only the top 20m of soil in classification overlooks the influence of deeper low-velocity layers on long-period structures. Additionally, reliance on the top 20m's shear wave velocity can be problematic due to its sensitivity to the properties of the fill layer and the potential unreliability of measurements in this shallow depth. To address these issues, scholars in China advocate increasing the depth considered in site classification from 20m to 30m. Current standards focus on soil layers not exceeding 20m, resulting in engineering boreholes and shear wave velocity measurements that rarely exceed this depth, especially in harder sites where boreholes often extend less than 10m. The development of site shear wave velocity extrapolation models is crucial for accurate site classification and ground motion parameter determination, particularly in the absence of deep borehole data.
Various extrapolation methods have been proposed, including the constant velocity method, velocity gradient method, and conditional independence method. The constant velocity method assumes a uniform velocity below the measured depth, while the velocity gradient method fits empirical relationships in a linear or logarithmic form. The conditional independence method leverages correlations between instantaneous and average shear wave speeds at various depths. Domestic research has led to the establishment of regional shear wave velocity extrapolation models, though their applicability is often limited to specific regions. The selection of the most suitable extrapolation method for a given region requires further investigation.
This study focuses on Shandong province, a region within China's North China Seismic Zone with a significant risk of strong earthquakes. With nearly 80% of the province requiring seismic fortification of at least Ⅶ-degree intensity, research into shear wave velocity extrapolation models is of practical importance for site categorization and seismic defense. Utilizing extensive shear wave velocity profiles and borehole lithology data, this study applies constant velocity, velocity gradient, and conditional independence methods to establish regional extrapolation models. It evaluates the applicability and accuracy of these methods in Shandong and recommends an empirical model for shear wave velocity extrapolation.
Key findings include: 1)For borehole depths less than 10m, the empirical extrapolation models for V_S20 and V_S30, utilizing the three discussed methods, exhibit considerable inaccuracies. Caution is advised when applying the wave velocity predictions from this study to depths under 10m. Notably, the BCV method tends to significantly underestimate when extrapolating from shallow data. The BCV method's predictions become more reliable and exhibit reduced error only when borehole depths exceed 10m for V_S20 and 15m for V_S30; 2)The empirical extrapolation models for V_S20 and V_S30 in Shandong province, developed using the velocity gradient method, align well with actual measurements. These models' regional applicability is supported by comparative regional analyses. The V_S30 predictions for Shandong are found to be generally lower than those in Japan but closer to those in California and the Beijing plain, with a slight increase in the higher wave speed range; 3)Considering the models' accuracy and regional applicability, the study advocates for the empirical extrapolation models of V_S20 and V_S30 for Shandong Province based on the conditional independence method. These models minimize total prediction errors across various depths. While the BCV model's performance improves at greater depths, the velocity gradient extrapolation model's efficacy diminishes.
Overall, this study contributes to the advancement of seismic design practices in Shandong province by offering empirical extrapolation models for V_S20 and V_S30, enhancing the understanding of ground motion characteristics and supporting more robust seismic resilience strategies.

Key words: site classification, extrapolation models, average shear-wave velocity, applicability, V_S30

摘要：

场地剪切波速是进行场地分类和定量估计场地对地震动影响的重要参数, 在工程抗震设防和震后震害快速评估等方面具有广泛应用。文中利用山东地区1 336个工程场地的剪切波速剖面数据, 分别基于常速度外推方法、 速度梯度外推方法和条件独立方法建立了山东地区场地剪切波速V_S20 和V_S30 的经验外推模型。研究结果表明, 常速度外推方法对较浅的钻孔进行波速外推时会产生明显的低估, 且预测误差较大。速度梯度外推方法的拟合结果表现出明显的区域性特征, 山东地区的V_S30 预测结果与美国加州和北京平原地区的结果相比, 总体上较为接近, 但明显低于日本地区。综合考虑建立的3种区域外推模型的精度和预测误差, 文中建议优先采用基于条件独立方法建立的山东地区的V_S20 和V_S30 经验外推模型, 获得的波速外推结果可以较好地为山东地区场地分类提供依据。

关键词: 场地分类, 外推模型, 平均剪切波速, 适用性, V_S30