Evaluation of Preharvest Forecasting Models for Prediction of Wheat Yields in Hoshiarpur Shiwalik Foothills

Abstract

An attempt was made to evaluate four of the preharvest forecasting models for prediction of wheat yield in Shiwalik foothills under limited irrigation conditions by taking four of the biometrical characters of wheat at the time of the time of harvest as predictor variables namely, number of plants and ears/m row length, average plant height and number of spikelets per ear. The models under study were able to explain over 80% of variation in wheat yield. The highest variation (88%) was explained by the multiple linear regression model while the lowest SE of estimate was obtained by using the square root linear regression models.