微博数据与地理信息数据结合的洪涝淹没概率估计方法OACSTPCD

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%.