Study On Precipitation Evolution Characteristics And Rainstorm Forecasting In Zhengzhou

By analyzing the evolution characteristics of precipitation in zhengzhou,the annual precipitation and daily rainstorm can be predicted,which is helpful to reveal the time change rules and evolution characteristics of precipitation in zhengzhou,Guiding flood control and disaster reduction,improving the planning and management of water resources,and promoting regional sustainable development have certain theoretical significance and application value.According to the precipitation data of zhengzhou in 67 years(1951 ~ 2017),the trend coefficient method,R/S analysis method,wavelet,mann-kendall catastrophe test and chaos theory were used to analyze the time change and chaos.Cumulative anomaly curve,nonuniformity coefficient method,precipitation concentration(PCD),precipitation concentration period(PCP),and Z index were used to analyze the precipitation nonuniformity and drought or flood.The coupled models of wavelet decomposition-Elman neural networkARIMA(WEA)and CEEMD decomposition-Elman neural network-ARIMA(CEA)were used to predict the precipitation from 2013 to 2017.The results show that:(1)The precipitation of Zhengzhou in year,flood season and non-flood season shows a trend of decrease but not significant.In the future,the precipitation of year,flood season and non-flood season will also show a trend of decrease.The flood season and annual precipitation show the same cycle change scale and main cycle.The catastrophe law of flood season and annual precipitation is roughly consistent,and there are many catastrophe points.The catastrophe law of precipitation in non-flood season is different from year and flood season,and the number of catastrophe points is less.The chaos in flood season is the strongest,and the chaos in non-flood season is the weakest,the minimum independent variables needed to describe the precipitation system in flood season,month and non-flood season are 7,5 and 4,respectively.(2)The interannual variation law of annual and flood season precipitation is consistent,specifically,the duration of dry season is longer,while that of wet season is shorter,however,non-flood season shows a longer period of dry season and wet season.The nonuniformity degree of precipitation is,the first is the year,the second is the non-flood season,and the last is the flood season.The precipitation concentration showed significant interannual variation,the monthly precipitation is mainly concentrated in June to August,the precipitation in flood season is mainly concentrated in late July to late August,and the precipitation in non-flood season is mainly concentrated in late January.The law of drought and flood in year and flood season is similar,and the year of occurrence of drought and flood in different levels is basically the same,while the year of occurrence of drought and flood in non-flood season is much different from year and flood season.(3)Both WEA model and CEA model can be used to predict annual precipitation and maximum daily precipitation,however,WEA model is slightly better than CEA model in these two types of precipitation prediction.In WEA's annual precipitation prediction,the maximum,minimum and average relative errors are 17.11%,1.58% and 7.3%,respectively,CEA models were 26.19%,7.85% and 14.1%,respectively.In the prediction of maximum daily precipitation,the maximum,minimum and average relative errors of WEA model are 31.61%,2.86% and 16.3%,respectively,the CEA model is 35.94%,1.19% and 20.85% respectively.