Evaluation of production functions to estimate rapeseed yield in different Interval times

In this research evaluated production functions has been modified by Rase, 2004 And Tafteh et al.2013, this methods evaluated to estimate rapeseed yield under deficit irrigations with in the monthly intervals, quadruple growth stages and total growth period in the Qazvin plain. For this purpose, randomized complete block design was applied. The irrigation treatments (60-90-120-150 mm evaporation from pan class A) with three replications were used during 2010 - 2012. The results showed that Treatments T11 and T12 with 2390 and 1320 yield (kg/ha) respectively have a maximum and minimum yield. Also results show that the best estimation was occurred in the total growth stage and other intervals were decreased the accuracy of the models. So the minimum error was in the total growth stage in the both methods. Statistical analysis shown that Tafteh et al. (2013) method in the total growth stage has the lowest root mean squared error (RMSE) and Normalized root mean squared error (NRMSE) respectively equal to 423 and 0.211. Also this method has a maximum of agreement index (d) and correlation coefficient (R2) Respectively equal to 0.76 and 0.645. Therefore Tafteh et al. (2013) method is suitable way for estimation of rapeseed yield in deficit irrigations.