基于夜间灯光影像的地震受灾区域和人口评估研究

发布时间：2019-02-12 12:25

【摘要】：灯光是人类活动的象征,一个光源以及光源强弱度代表人的存在或人类活动的强弱,同时灯光与电力设施和建筑物也息息相关,一个地区受强烈地震后震害信息将在灯光的破坏范围和程度上得到明显表现,特别是对救援主体-人能够提供最为直接的信息。因此夜间灯光影像对于地震灾害救灾决策具有至关重要的意义。本文以DMSP/OLS夜间灯光遥感影像作为数据源,选择了2011年日本大地震、2008年汶川大地震开展了夜间灯光影像的滤波、快速匹配、归一化的预处理研究有效消除了非地震因素影响,并对灯光变化检测采用了两时相、多时相和全年灯光数据插值的方法进行提取,研究结果表明3种方法得到了相似的效果,检测到日本大地震受灾严重的宫城县、福岛县和岩手县等以及汶川地震中灾害严重的都江堰、彭州市、什邡市、绵竹市、安县、汶川县等与实际统计结果相一致,这验证了之前的实验流程、处理方法、提取方法都是正确的。不同提取方法具有不同的适用范围和精度,证明了利用灯光数据来分析地震灾害的影响区域是可行的,但提取分析也受地震影响区域的经济发展、人口环境、自然地理环境,震前震后天空中的云量和云层厚度以及传感器探测能力等因素的影响。最后,以常态下2002-2010年中国人口统计数据与灯光数据建立了人口分布与灯光的定量关系模型,并对汶川地震灾区面积提取及受灾区域内人口进行估算和精度分析,根据受灾面积和人口比例判定了极重灾区、较重灾区和一般灾区,实现了对人口定量评估及其宏观分布的动态监测,进而为地震等灾害应急决策提供快速信息支持。
[Abstract]:Light is the symbol of human activity. A light source and its intensity represent the existence of human beings or the intensity of human activity. At the same time, lighting is closely related to power facilities and buildings. The earthquake damage information of a region after a strong earthquake will be obviously displayed in the scope and extent of light damage, especially the most direct information can be provided to the rescue main-person. So night lighting image is very important for earthquake disaster relief decision. In this paper, the DMSP/OLS night lighting remote sensing image is used as the data source, the 2011 Japan earthquake and the 2008 Wenchuan earthquake are used to filter and match the night lighting images. The normalized pretreatment study effectively eliminates the influence of non-seismic factors, and the 02:00 phase, multi-phase and year-round lighting data interpolation methods are used to detect the light changes. The results show that the three methods have similar results. Miyagi Prefecture, Fukushima Prefecture and Iwate Prefecture, which were badly affected by Japan's major earthquake, as well as Dujiangyan, Pengzhou, Shifang, Mianzhu, Anxian and Wenchuan counties, which were seriously affected by the Wenchuan earthquake, were detected to be consistent with the actual statistical results. This verifies that the previous experimental flow, processing methods and extraction methods are all correct. Different extraction methods have different scope of application and accuracy, which proves that it is feasible to use light data to analyze the affected areas of earthquake disasters, but the extraction and analysis of areas affected by earthquakes are also affected by economic development, population environment, natural geographical environment. The influence of cloud amount, cloud thickness and sensor detection ability in the sky after the earthquake. Finally, the quantitative relationship model between population distribution and lighting is established based on the population statistics and lighting data of China from 2002 to 2010, and the area of Wenchuan earthquake disaster area is extracted and the population in the affected area is estimated and analyzed accurately. According to the disaster area and the proportion of the population, the paper determines the most serious disaster area, the worst disaster area and the general disaster area, realizes the quantitative population evaluation and the dynamic monitoring of macroscopic distribution, and then provides the rapid information support for the emergency decision of earthquake and other disasters.
【学位授予单位】：河南理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：P315.9

【相似文献】