Agrometeorological rice yield modelling in Karnal

District-level Agromet rice yield model was developed using rice yields of past fourteen years (1981–1994) and meteorological data such as minimum-maximum temperature and sunshine hours in the Karnal district of Haryana state. The Growing Degree Days (GDD), Temperature Difference (TD) and Accumulated Sunshine hours (ASH) were calculated and integrated over three different crop growth phases to study their influence on district-level rice yield. The three growth phases considered for analysis were Active Vegetative Phase (AVP), Reproductive Phase (RP) and Maturity Phase (MP). A two step linear statistical technique was adopted for multiple linear regression analysis. In the first step, best possible subset of independent variables were selected by leaps and bounds technique. In the second step, the multiple regression for each selected subset were carried out and variance, regression coefficients and residuals were computed. The selected subset of independent variables constitute TD at AVP and RP, GDD at RP and ASH at AVP and RP, which resulted in best multiple regression model with R² 0.842 and SEOE 0.663. This model explains about 84 per cent variability in the district-level rice yields. The model predicted 2.537 t ha^?1 rice yield for the kharif 1995 season.