Principal Component Technique for Pre-harvest Crop Yield Estimation Based on Weather Input

Abstract

Forecasting of crop production is one of the most important applications of statistics in agriculture. Such predictions before harvest are needed by the national and state governments for various policy decisions relating to storage, distribution, pricing, marketing, import- export, etc. Therefore, a methodology for the estimation of wheat yield, ahead of harvest time, is developed specifically for wheat growing districts in Haryana (India). The Haryana state, having a total geographical area of 44212 sq. km, was divided into four zones for pre-harvest crop yield forecasts. An attempt has been made in this paper to estimate the yield of the wheat crop using principal components of the weather parameters spread over the crop growth period. Principal component analysis has been used for the purpose of developing zonal yield forecast models because of multicollinearity present among weather variable. The results indicate the possibility of district-level wheat yield prediction, 4-5 weeks ahead of the harvest time, in Haryana. Zonal weather models had the desired predictive accuracy and provided considerable improvement in the district-level wheat yield estimates. The estimated yield(s) from the selected models indicated good agreement with State Department of Agriculture (DOA) wheat yields by showing 2-10 percent average absolute deviations in most of the districts except for the Rohtak district observing 12.81 percent average absolute deviation from the real-time data.