Research on intelligent climate prediction methods of rainstorm concentration in Guangxi
In this study,using the daily precipitation of 79 meteorological observation stations in Guangxi and the data of the atmospheric circulation indices and sea surface temperature indices data of the National Climate Center from 1961 to 2023,we constructed a method for calculating the rainstorm concentration in Guangxi,and established a climate prediction model for the concentration of rainstorm based on the stepwise regression method,the particle swarm neural network,and random forest algorithm.The results showed that there are three areas of high rainstorm concentration in Guangxi,namely,the northeastern part of Guangxi centered on the northern part of Guilin and Liuzhou,the mountainous area in the western part of Guangxi centered on"Donglan,Bama,Fengshan"area,as well as the coastal area.The anomaly of rainstorm concentration basically reflects the severity of flooding and drought disasters.The climate prediction model of rainstorm concentration based on the stepwise regression method,particle swarm optimization neural network and random forest algorithm is established.According to the climate prediction experiments in 2020-2023,the most effective prediction was made by the particle swarm-neural network algorithm,followed by the random forest algorithm,and finally by the stepwise regression method.
万方数据
维普
