MODELS FOR STUDYING RELATIONSHIP BETWEEN RAPESEED & MUSTARD YIELD AND WEATHER VARIABLES.

Various statistical models to study the effect of weather variables on rapeseed & mustard yield at different stages of crop and to forecast its yield for Faizabad District have been attempted. Time series data on yield for rapeseed & mustard crop of Faizabad district of Uttar Pradesh for 22 years (1990-91 to 2011-12) have been used. Weekly weather data for the period (1991-2012) on six weather variables viz. Minimum Temperature, Maximum Temperature, Relative Humidity at 7 hour, Relative Humidity at 14 hour, Sun-shine hours and Wind-Velocity of Faizabad district have been utilised for the present study. Various models have been tried to know the relationship between rapeseed & mustard crop yield by generating new weather indices from weekly data. In all, eight statistical models were developed. The model involving total of value of weather variable over weeks and linear functions of respective correlation coefficients between adjusted crop yield for trend effect and weekly data and weather variables as the independent variables has been found to be the best one (model VII & VIII) for studying the effect of individual weather variables. A forecast model has been developed to forecast the pre-harvest crop yield. The results revealed that forecast at 21st week of the crop period i.e. during third week of February before one and half months of its harvest was found to be reliable one as percent standard error of forecast was obtained to be approximately 4 percent. [ABSTRACT FROM AUTHOR]