Abstract:

Based on the felt information from the Internet after the earthquake, this paper studied the methods of earthquake damage information acquisition and analysis. We carried out the research based on 4515 pieces of felt information after the Wenchuan earthquake acquired from the Internet. After the analysis on the publishing time, location and intensity of the feeling of the earthquake in the felt information, we found that the release of felt information was concentrated within the 3 hours after the Wenchuan earthquake happened. And we initially identified 2 289 felt data released in the subsequent three hours after Wenchuan earthquake, which distributed in 199 areas, as the data sources for spatial interpolation. In order to do spatial interpolation for these datasets, we classified the datasets into three categories, which are "unawareness", "felt" and "strongly felt", and assigned the value of 0, 1 and 2 to "unawareness", "felt" and "strongly felt" respectively. Before doing the spatial interpolation for the datasets, we preprocessed the datasets distributed in 193 different areas to remove the redundant and abnormal value data and to finally extract 193 felt data distributed in different areas. At the same time, this paper did exploratory analysis for the datasets, by which we knew the law of the datasets and determined the parameters and model for the spatial interpolation. After a series of analysis, this paper adopted IDW, RBF and Kriging methods to interpolate the felt datasets, and get the results of the felt range of Wenchuan earthquake. Then, the paper compared the results from different spatial interpolation ways by interpolation accuracy and the goodness of fit with the actual earthquake disaster. After the comparison, we found that the result of Kriging method, with a certain interpolation model and parameters, was consistent with actual disaster distribution of the Wenchuan earthquake to a certain extent, especially for the fitting with the strongly felt area (Ⅵ degrees). This study provides an effective way for us to know quickly the earthquake felt range and correct for the disaster information of the earthquake. And it will play an important role in making scientific and rational decisions during the earthquake emergency.

Key words: Wenchuan earthquake, Internet, disaster acquisition, earthquake emergency, earthquake felt area, spatial interpolation, GIS

摘要：

通过搜索和分析互联网上有关汶川地震的信息得出, 汶川地震后3h为有感信息密集发布时段, 信息总量为4 515条.分析信息的发布时间、地域和有感强度分布特征, 对信息按照无感、有感和强有感进行分类, 提取分布于199个地区的2 289条有感信息作为数据源进行有感范围的空间插值拟合处理.空间插值选用反距离权插值、径向基函数插值和克里格插值3种方法, 通过对不同插值结果的比较分析发现, 克里格插值方法在一定的插值模型和参数下对有感范围的拟合效果与汶川地震实际影响范围基本吻合, 尤其是强有感范围基本反映了地震灾区范围.该项研究为震后快速获取地震有感范围提供了有效途径, 同时也为地震灾害速判提供了动态修正依据, 为地震应急提供第一时间的灾情信息.

关键词: 汶川地震, 互联网, 灾情获取, 地震应急, 地震有感范围, 空间插值, GIS