Study On Agricultural Drought Prediction In Henan Province

Drought is one of the most common,most common and widely distributed climate disasters in China's agricultural production.Agricultural drought prediction is not only an important basis for decision makers to make disaster reduction decisions,but also one of the non-engineering measures for drought defense.It is of great significance to reduce the agricultural losses by understanding the drought trends in advance,adjusting crop varieties,and formulating effective measures to combat drought and reduce disasters.Henan is China's main grain producing area,and accurate drought prediction is of great significance to improve Henan's comprehensive defense ability against drought.Based on the drought situation in Henan,Shangqiu,Xuchang,Zhumadian,Anyang and Mengjin were selected as five typical research areas to study the agricultural drought in Henan Province.The main research results are as follows:(1)Comprehensive analysis of the three most authoritative drought indexes PDSI,SPI and SPEI in the academic world.Finally,SPI and SPEI were selected as the drought index for studying agricultural drought in Henan Province.(2)For the typical representative area,the combination of MK test method and sliding T test method is used to test the trend and the mutation of precipitation and temperature.The study found that the annual precipitation in Henan Province of Henan Province showed an upward trend,and the trend was not significant.The annual precipitation in Henan,Yubei,Yuzhong and Henan areas of Henan Province showed a downward trend,and the trend was not significant.The annual average temperature in Henan Province has a significant upward trend.The annual average temperature in the eastern Henan region has changed suddenly around 1993.The annual average temperature in the southern Henan region has changed suddenly around 1996.The annual average temperature in the western Henan region has changed suddenly around 2006.(3)Using the SPI index and SPEI index to analyze the drought law in Henan Province,the study found that the SPEI index is more suitable for agricultural drought in Henan Province than the SPI index.Today,the SPEI index's drought classification standard still uses uniform standards across the country.This paper has improved this and formulated a classification standard that is more suitable for agricultural drought in Henan Province according to the agricultural drought situation in Henan Province.It innovatively proposes different time scales for different regions in Henan,different thresholds of SPEI index for agricultural drought.For identification,the correct rate can reach more than 90%.(4)Using the time series neural network,which is less frequently used in hydrological time series,NAR neural network.In time series prediction,NAR neural network is more predictive than traditional BP neural network and REB neural network..The precipitation and temperature of the five study areas were predicted,and the SPEI index was calculated based on the predicted precipitation and temperature data to predict the agricultural drought.The results show that the NAR neural network has a high accuracy rate for agricultural drought prediction in Henan Province within three years,and the NAR neural network can predict the agricultural drought in Henan.