Risk Trend Analysis And Risk Assessment Of Typhoon Disaster In South China

Typhoon disaster is one of the main natural disasters in coastal areas of China.Typhoon landing can bring abundant precipitation to the region and alleviate local drought.However,with landing,there are also gales,rainstorms and storm tides.Typhoon gale and precipitation often cause secondary disasters such as house collapse,landslide and debris flow,threatening people's life and property safety and economic development It was seriously affected.Among them,South China(in this paper refers to Guangdong Province,Guangxi Zhuang Autonomous Region and Hainan Province)is located in the south of China,and it is adjacent to the South China Sea and the Beibu Gulf sea area in the south,which is vulnerable to typhoon disasters from the Pacific Ocean,which brings a relatively big threat to the lives and production of local people.Therefore,this paper will take South China as the research area,based on the historical typhoon disaster data and South China disaster bearing body attribute data,analyze the trend risk of landing typhoon extreme disasters,and establish the typhoon risk pre assessment and dynamic assessment model,so as to realize the typhoon disaster grade pre assessment and annual dynamic risk assessment in South China.The research work of this paper mainly includes:(1)In recent years,the risk trend of typhoon landing in South China is analyzed.KS test method and one-dimensional information diffusion model are used to determine the probability distribution of risk variables of typhoon disaster,and further determine the risk threshold of extreme disasters.Then MK trend analysis method is used to explore the historical risk trend of typhoon extreme disasters in South China.Finally,R/S analysis method is used to predict the future evolution trend of typhoon extreme disasters.The results show that the risk of extreme gale and heavy rain is increasing.(2)The risk index of typhoon in South China is constructed by using the combination weight theory,and the pre evaluation of typhoon risk level is realized by using machine learning classifier.The vulnerability index and disaster prevention and mitigation index of South China disaster bearing body are constructed by principal component analysis method.Then,the combination weight is determined by combining subjective and objective view.The risk is quantified by risk formula,and the risk pre assessment index of typhoon in South China is established.The direct economic loss rate of typhoon is detrending,and the disaster level is divided by unsupervised clustering.Different machine learning classifiers are used to establish the pre assessment model of typhoon disaster level in South China.The machine learning classifier is most suitable for typhoon risk pre assessment by selecting independent test samples.The results show that Random Forest is most suitable for typhoon risk pre assessment in South China.(3)A dynamic risk assessment model of typhoon in South China is established based on the change of month and landing conditions.The time element of risk system is considered to ensure the elimination of the impact of systematic error.Firstly,one-dimensional information diffusion and two-dimensional information matrix are used to determine the probability distribution of typhoon risk intensity in different months and the typhoon vulnerability function in different landing sites in South China.Then,the probability distribution and vulnerability function are coupled to calculate the typhoon risk distribution in different months and landing conditions in South China in a year.The results show that the risk of typhoon in South China reaches the peak in July and September,and the risk of typhoon landing in Hainan will be greater.