USING CERES-WHEAT MODEL TO SIMULATE GRAIN YIELD PRODUCTION FUNCTION FOR FAISALABAD, PAKISTAN, CONDITIONS

Using computer simulation model as a management tool requires model calibration and validation against field data. A three-year (2008-2009 to 2010-2011) field study was conducted at the Postgraduate Agricultural Research Station of the University of Agriculture, Faisalabad, Pakistan, to simulate wheat grain yield production as a function of urea fertilizer applications using Crop Environment REsource Synthesis (CERES)-Wheat model. The model was calibrated using yield data for treatment of urea fertilizer application at the rate of 247 kg-urea ha(-1) during growing season 2009-2010 and was validated against independent data sets of yield of two years (2008-2009 and 2010-2011) for a wide variety of treatments ranging from no urea application to 247 kg-urea ha(-1) application. The model simulations were found to be acceptable for calibration as well as validation period, as the model evaluation indicators showed a mean difference of 8.9%, ranging from 0.05 to 15.38%, root mean square error of 356 having its range from 242 to 471 kg ha(-1), against all observed grain yield data. The scenario simulations showed maximum grain yield of 4100 kg ha(-1) for 350 kg-urea ha(-1) in 2008-2009; 4600 kg ha(-1) for 300 kg-urea ha(-1) in 2009-2010 and 5200 kg ha(-1) for 340 kg-urea ha(-1) in 2010-2011. Any further increase in urea application resulted in decline of grain yield function. These results show that model has the ability to simulate effects of urea fertilizer applications on wheat yield; however, the simulated maximum grain yield data need field-based verification.