Regression Models of Winter Wheat Yield Components in Zhengzhou Area

Winter wheat yield prediction plays an important role in ordinary meteorological service and its accuracy is associated with forecasting techniques. It's one of main techniques analyzing relationship of yield components to meteorological elements and then establishing regression model about yield components.Winter wheat yield in Zhengzhou Area is studied based on the date of winter wheat growth stage, 1000-grain weight, spike grain number and the daily meteorological data at Zhengzhou Agro-meteorological Experimental Station from 1980 to 2006. By calculating the impact indexes of light, temperature and precipitation, the relationship between these indexes and 1000-grain weight, spike grain number are analyzed, and key factors are selected by analyzing the change of correlation coefficient to establish the prediction model of 1000-grain weight and spike grain number. The result indicates that 1000-grain weight shows a positive relationship and a negative relationship with the average temperature from the 15th day to 19th day and from the 29th day to 33rd day during the heading stage, respectively. At the same time, the average sunlight hour from the 9th day to 13th day shows a negative effect. To gain a high 1000-grain weight, the mean temperature is supposed to be between 18.8℃ and 22.8℃, and the highest temperature doesn't exceed 29.5℃. The average temperature from the 21st day to 25th day during the reviving stage shows a positive effect obviously for spike grain number. By analyzing the relationship between yield components and climatic elements respectively, the estimate of 1000-grain weight and spike grain number are possible in advance.