Performance evaluation of CERES-Wheat model in Guanzhong Plain of Northwest China

Winter wheat (Triticum durum Desf.) is one of the most important food crops in Guanzhong Plain of Northwest China. Irrigation and fertilizer are essential for wheat growth in the region because of inadequate precipitation and unfertile soil in the region. However, in most cases optimal yields are not gained due to unreasonable irrigation and fertilizer applications. A decision-supporting tool is required to facilitate local farmers to evaluate irrigation and fertilizer uses and their impact on yield. In this study, CERES (Crop Estimation through Resource and Environment Synthesis)-Wheat model, a process-based model, was adapted for predicting yield of winter wheat calibrated using plot experiments in the region. Water and nitrogen (N) limitation conditions were included in the experiments conducted in Yangling on small test plots during the growing years of 2009–2010, 2010–2011 and 2011–2012. Results showed that the simulated aboveground biomass, grain yield, leaf area index (LAI), canopy N, cumulative evapotranspiration (ET), water use efficiency (WUE) and nitrogen partial fertilizer productivity (NPFP) deviations from the measured were acceptable with normalized root mean square error (NRMSE) below 21%. LAI rose as N application increased and the synergistic impact of irrigation and N was exerted on LAI. ET and canopy N at maturity increased with irrigation although N application was different for 2010–2011 and 2011–2012. The response of both ET and canopy N to fertilizer N showed the similar trend. Besides, the impact of irrigation on ET was affected by different N applications under low irrigation conditions. WUE and NPFP were exerted a great influence on by N fertilizer and the impact of irrigation on them was not significant. Model sensitivity analysis indicated that the grain yield was very sensitive to field capacity among the factors except for the date-related factors and planting date had a significant influence on grain yield by comparison with the factors related to the date. Therefore, the model can be used for the region to assess and optimize winter wheat yield with respect to irrigation and N applications as well as other influential factors.