Agrometeorological wheat yield prediction in rainfed Potohar region of Pakistan

Abstract

The reliable early estimates of production had always been the prime concerns of growers on one hand and planners as well as policy makers for import/export on the other hand. This study represents a linear regression model making use of meteorological parameters at critical stages of crop’s life cycle to predict the wheat yield about two months earlier than the harvesting. A statistical based software “Statistical Package for Social Sciences” (SPSS) and MS-excel were employed as working tools. Decadal (ten days) agrometeorological data for Rabi season (for the period 1993-2011) being observed at Regional Agromet Centre, Rawalpindi have been utilized. The parameters studied for correlation were mainly rainfall (amount and days), air temperature (minimum, maximum, mean), heat units (on phenological basis), relative humidity, wind speed, sunshine duration, reference crop evapotran-spiration etc. Finally, minimum temperature, sunshine duration and rainfall amount in January (tillering and stem extension phase) proved to be the reliable predictors for the final yield. The correlation coefficients for these parameters on individual basis resulted within the acceptable range where as in aggregate it remained 0.87, an optimistic value.