基于K-NET数据库的地震动持时预测模型  

Time-duration Prediction Model of Ground Motion Based on K-NET Database

作　　者：崔如玉 张齐 黄昊 谢礼立[3] CUI Ruyu;ZHANG Qi;HUANG Hao

机构地区：[1]三峡大学土木与建筑学院,湖北宜昌443002 [2]防灾减灾湖北省重点实验室,湖北宜昌443002 [3]哈尔滨工业大学土木工程学院,哈尔滨150090

出　　处：《科技创新与应用》2025年第10期6-9,共4页Technology Innovation and Application

基　　金：国家自然科学基金面上项目(52079072)。

摘　　要：该文收集来自日本K-NET数据库的104次地震事件,以5%~75%显著持时作为研究对象,研究震级、震源距和地表以下30 m范围内的等效剪切波速(Vs30)对显著持时的影响。通过随机效应回归方法和残差分析,得到一个对数据适应性较好的地震动显著持时预测模型,并且持时会随着Vs30的增大而减小,近场情况下,超过300 m/s持时对Vs30的依赖性会减弱。In this paper,104 seismic events from the K-NET database in Japan are collected,5%~75%of the significant duration are used as the study object.The study of the relationship between the magnitude,hypocentral distance and time-averaged shear-wave velocity of below 30 m(Vs30),and the effect of three parameters on duration,so developed a predictive model for earthquake ground motion duration by random effects regression and residual analysis.The model reveals that ground motion duration decreases as Vs30 increases.Furthermore,in near-field fields,the dependency of duration on Vs30 will be weakened if it exceeds 300 m/s.

关 键 词：K-NET数据库 强震观测记录 显著持时 残差分析 预测模型 

分 类 号：P315.9[天文地球—地震学]