Forecasting of crop yield using weather parameters - two step nonlinear regression model approach

Abstract

Concept of the paper is firstly to remove the trend of crop yield and then to develop the forecasting models using detrended yield. Not much work is available or development of forecast models or modelling due to their non-linear behaviour. For that, in this paper, methodology developed for forecasting using nonlinear growth models, which will help in forecasting yield, pest and disease incidences etc with high accuracy. Crop yield forecast models for wheat crop have been developed (using non-linear growth models, linear models and weather indices approach with weekly weather data) for different districts of Uttar Pradesh (UP). Weather Indices (WI) were obtained using above two approaches. Weather indices based regression models were developed using weather indices as independent variables while character under study such as crop yield was used as dependent variable for wheat crop, i.e. two step non-linear forecast model. Technique of forecasting using non-linear approach and using weather indices will enrich the knowledge in developing customized models on forecasting for different types of crops and for different locations. The approach provided reliable yield forecast about two months before harvest.