Development of Crop Yield Forecast Model for rabi Sugarcane in Different Districts of South Gujarat, India

Abstract

The present study was under taken to investigate the feasibility of estimating productivity of rabi sugarcane crop based on weather variables using past weather and yield records of different districts of South Gujarat. In the modeling used, it was considered as an independent variable. The generated weather variables were developed using weighted accumulation of fortnightly data on weather variable, weights being the correlation coefficient of the weather variables, in respective fortnights with the dependent variable, i.e. the yield of sugarcane yield. Forecasting models were developed using from a time series of 26 years (1995-96 to 2020-21) of weather and yield data of sugarcane crop in four Navsari, Bharuch, Surat and Tapi districts and their performance have been validated against the observed yield during 2018-19, 2019-20 and 2020-21. Results indicated that coefficient of determination (R2) values were 0.67 to 0.76 during mid-season (F1) stage and 0.76 to 0.91 during pre-harvest (F2) stage in four districts. Hence, these models can be used for forecasting sugarcane yield in preharvest stage which is very useful to government authorities to plan the sugarcane production more efficiently.