Research On Peanut Yield Forecast Technology In Henan Province

Crop yield forecast is an important part of meteorological service for agriculture.With the implementation of the national rural revitalization strategy,peanuts are playing an increasingly important role in international trade and people's life.The research on quantitative prediction technology of peanut yield can provide objective and quantitative prediction conclusions according to the future weather situation changes,which can not only provide accurate information for the government and decision-making units,but also improve the level and quality of meteorological services for rural revitalization.Based on the historical yield data and meteorological data of 13 main peanut producing areas in Henan Province from 1985 to 2017,and on the basis of summarizing the previous crop development period forecast and yield forecast techniques,this paper determined the suitable technical methods of peanut yield forecast by screening and comparing,which are the influence index of key meteorological factors and the similar year type method of historical yield.The meteorological factors from 1985 to 2014 were used to forecast the peanut yield in each region,and the data from 2015 to 2017 were used to carry out extrapolation test.On this basis,the accuracy weighted method was used to integrate the above two forecasting methods,and a technical method of peanut yield prediction based on digitization and integration technology was obtained.The main conclusions are as follows:(1)The yield of peanut was greatly affected by meteorological conditions,especially the temperature at branching stage and late grain filling stage(fruit-filling stage)had significant correlation with the relative meteorological yield of peanut.(2)The prediction effect of similar year model and key meteorological factor influence index model for peanut yield in all regions of Henan Province was relatively good,but the prediction accuracy of a few years did not reach 90.0%;(3)The forecast accuracy of production based on provincial decision is better than that of regional modeling,and the forecast accuracy is more than 95%,which indicates that the statistical method has a better fitting effect on the overall macro situation and the forecast results are more reliable.(4)The accuracy and stability of the peanut yield forecasting model were improved by the accuracy weighting,which indicated that the integrated forecasting model was more suitable for business application.