The Long-term Change Of Wheat Yield And Soil Moisture In Jingnan Plateau

Along with the development of social economy, human beings have increasing demand for food. How to effectively increase the crop yield has become a huge task for people. Due to the limited personal abilities, the author only discusses the relationship between winter wheat yield and moisture in southern Shanxi in this paper so as to figure out the relationship between precipitation, soil moisture and water consumption and the yield of winter wheat. The author also establishes the mathematical model of moisture and yield to provide theoretical support for specifically increasing the water use efficiency of winter wheat and improving agricultural measures and to offer new ideas and methods for yield predication, thus contributing to the agricultural economic production and agricultural decision-making.Based on the field observation and the analysis of the long-term observed data, the following conclusions are drawn:1. Divide the wheat yield into tendency yield and meteorological yield by the application of moving average method. The correlation analysis has shown that the correlation coefficients between the precipitation of fallow period, of sowing-wintering period, of wintering-reviving stage, of reviving-elongation stage, of elongation-boot stage and of grain filling stage and the meteorological yield are0.664**,-0.118,-0.022,0.075,0.064,-0.254, respectively. The precipitation of fallow period has a significant correlation with meteorological yield (0.664**).2. The analysis on soil moisture storage rate during the fallow period indicates that the precipitation stored in the soil of experimental plot in this period is only42.11%, averagely. The analysis on the water consumption of wheat indicates that the water supplied by soil accounts for37.59%of the total during the growth period, which shows that the precipitation of fallow period is not fully used or stored in the soil and effective measures for preserving soil moisture should be taken to manage soil water.3. Through the data analysis from the perspectives of water supply and water consumption as well as the aspects of precipitation, soil moisture and water consumption, we can conclude that the moisture content of the soil layer which is most influential to the yield of each growth stage can be used as the input data of yield prediction. But in order to ensure the orthogonality of the data, we need to use the sum of these soil water contents as input data. A desirable prediction method is to use the thought of regression analysis and BP neural network to predict yield and then find out that BP neural network prediction has a smaller relative error.