Hazard evaluation of mountain flood disaster based on multi-source remote sensing information - A case study in city of rain, Ya'an

Abstract

Because the hazard of flood disaster is mainly related to the natural attribute of flood disaster, this paper makes full use of multi-source remote sensing information to reflect the meteorological and underlying surface conditions in Ya'an City, constitutes an evaluation index system according to selecting the rainfall, runoff depth, the basin storage capacity, elevation, slope and water system, from the two aspects of disaster-inducing factors and disaster-pregnant environment; trains historical flood occurrence sites by using BP neural network, obtains the weights of each research index; overlays analysis each index according to the index model, based on the GIS technology; and obtains the torrential flood hazard distribution map of Ya'an City as well as comparing it with the occurrence sites of historical flood disasters. The results show that the flood hazard level is the highest in the eastern part of Ya'an, the lowest level area of the flood hazard is the southern part. The evaluation results well reflect the actual situation of regional flood spatial distribution, which can provide a scientific basis for the development of reasonable flood control and disaster mitigation planning in Ya'an City.