A study on pre-harvest forecast of sugarcane yield using climatic variables

A suitable statistical model has been developed for forecasting the yield of the sugarcane in Coimbatore district (1981-2004) using the yield data and fortnightly weather variable viz. average daily maximum and minimum temperature, relative humidity in the morning and evening and total fortnightly rainfall. The forecast model was developed using generated weather variables as regressors in model. 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 yield. The data for a period of (1981-2001) was used to develop the forecast model. The validation of the model was done using the data from (2002-2004). The results revealed that the forecast model developed was able to explain 87% of variation in the sugarcane yield. And it is possible to forecast sugarcane yield successfully two months before harvest.