水力发电2024，Vol.50Issue(9)：22-29,43,9.
微博数据与地理信息数据结合的洪涝淹没概率估计方法
Estimation of Flood Inundation Probability by Combining Weibo Data and Geographic Information Data
摘要
Abstract
In recent years,extreme weather has caused frequent and continuous heavy rainfall,resulting in more frequent and irregular flood disasters,which seriously affects people's safety and economic development.The assessment of flood inundation probability can strengthen the emergency management ability of flood disasters.Traditional remote sensing data are limited by re-entry periods and meteorological conditions,while the hydrological models are limited by data input and parameter settings.Considering the limitations of the above data and model methods,and the characteristics of time-sensitive and massive of Weibo data,a method is proposed to estimate the probability distribution of flood inundation in disasters by integrating digital elevation model and its derived data,rainfall data,water system data and Weibo data in disasters.The inverse distance attenuation function is used to generate a real-time updated inundation probability map of the whole map according to the water depth point,and then the weight allocation of the geographic information data is combined with the Gaussian function to obtain the final spatial continuous inundation probability map after comprehensive superposition.The"7·20"Henan rainstorm event in 2021 is taken as an example,and the results show that the accuracy of radar images is 92.75%,and the accuracy of official media and Weibo image information verification is 93.33%.关键词
洪涝淹没/概率估计/微博数据/反距离/高斯Key words
flood inundation/probability estimation/Weibo data/inverse distance/Gaussian分类
天文与地球科学引用本文复制引用
崔志美,黄维,黄志都,邬蓉蓉..微博数据与地理信息数据结合的洪涝淹没概率估计方法[J].水力发电,2024,50(9):22-29,43,9.基金项目
广西电网有限责任公司电力科学研究院研究项目(GXKJXM20222160) （GXKJXM20222160）
国家重点研发计划青年科学家项目(2021YFF-0704400) （2021YFF-0704400）
