基于建筑物破坏的地震人员伤亡快速评估方法是一种静态评估方法，具有较大的局限性，因此迫切需要建立地震人员伤亡动态评估方法以提高评估结果的有效性。以川滇地区为例，基于历史震例人员伤亡资料和相关科研成果，利用统计分析、案例对比法以及大数据分析等方法，从影响人员易损性因素的地区差异性角度研究地震人员伤亡动态修正方法。为此提出了人口动态分布、建筑物抗震能力及其致死率、地震地质灾害致死率等3类地区差异性因子，并建立了川滇地区的人员伤亡动态评估修正因子表。通过对川滇地区2008—2019年发生的6次强震进行动态评估试算，结果表明，通过动态评估模型评估的人员伤亡数量能够与实际伤亡人数保持数量级一致，评估结果准确性优于传统评估方法。

The traditional rapid assessment method of earthquake casualties based on building destruction is a static method with some limitations in practice. It is urgent to establish a dynamic method for earthquake casualties assessment to improve the effectiveness of the assessment results. Taking Sichuan and Yunnan Provinces as the examples,based on historical earthquake data and related scientific research results,we used statistical analysis,case comparison method and big data analysis method to establish the dynamics earthquake casualties assessment method by using the regional vulnerability for assessment results correction. In order to realize the regional dynamic assessment,three types of disparity factors were proposed,i.e. regional dynamic population distribution,regional building vulnerability and lethal rate in earthquake,and regional fatal rate of geological secondary disasters of earthquake. Correction factors for dynamic earthquake casualties evaluation were established. In order to evaluate the dynamic assessment method,trial calculation of five strong earthquakes occurred in Sichuan and Yunnan Province from 2008 to 2017 was conducted. The results showed that the number of casualties estimated by the dynamic evaluation model were consistent with the actual number of casualties. The accuracy of our method was better than that of traditional evaluation methods.

P315

国家重点研发计划（2016YFC0803109、2018YFC1504506）共同资助