Research on Analysis and Evaluation of Extreme Weather Based on Dynamic Evaluation Model

Through the analysis of extreme weather data in the United States, this paper establishes a dynamic evaluation model of weather, analyzes the temporal and spatial law of extreme weather, carries out time series analysis and multiple regression analysis of economic loss of extreme weather, and gives the prediction of economic loss of extreme weather. Firstly, this paper reasonably determines the starting position of the time stage, establishes the data dynamic evaluation model of T -Year return period, takes the data information of tornadoes in the Texas area of the United States as an example, divides the categories of tornadoes. Then, the indexes for studying the temporal and spatial law, namely trend, period and space, are determined, and the data set is fitted to obtain the trend model of extreme tornadoes in this area, so as to obtain the periodic change law of extreme tornadoes in Texas area of the United States. Secondly, the spatial frequency change data set can be obtained by counting the occurrence frequency of extreme weather in other states of the United States. Combined with the spatial location of American States, the bilinear interpolation model is introduced to obtain the spatial change law of the event. Finally, based on the above analysis, the ARIMA model is used to make a scientific and reasonable prediction of the occurrence frequency of the event in the future.