Applicability Evaluation Of AquaCrop Model In The Guanzhong&Weibei Area

AquaCrop model is a new crop simulated model driven by water issued by FAOorganization in2009. This model changes the problems of detailed and complete datarequirements, complicated operation, poor transparence of operation process and unstablesimulated results. As a crop model focused on practical application, it achieves a good balancein simulation accuracy, intuition of simulation process and stability of numerical simulations.Before applying this model, the parameters must be calibrated and adjusted based on thespecific areas and crops, in addition, the applicability evaluation of this model must beperformed based on the measured data to achieve model localization. Localized AquaCropmodel can not only effectively simulate productivity of local crop to provide theory guide formaking agricultural policy, but also provide technical support for the optimization of localirrigation scheduling and agricultural strategies.Based on the research status and existing problems of domestic crop models especially theAquaCrop model, this study was aimed at calibrating the model parameters according to thecollected field experiment data, and evaluating the applicability of AquaCrop model on theproduction of winter wheat in the Guanzhong area and summer corn in the Weibei dryhighland. The main results are as follows:(1) AquaCrop model could simulate the production of winter wheat in the Guanzhongarea very well. This model could be applied to provide feasible monitoring and prediction onthe production of winter wheat of Guanzhong area. Moreover, it could be used to optimize theirrigation system.When using AquaCrop model to simulate the production of winter wheat in Guanzhongarea, especially under the condition of sufficient simulated water, the parameter of soilmoisture threshold should experience proper calibration and testing. After this step, the modelcould simulate the yield, aboveground biomass and changes during growth stage of theGuanzhong winter wheat under different irrigation conditions in high precision. In this study,the adjusted determination coefficient (Adj.R2) between simulated yield and measured yieldwas0.840, with relative error being-2.763%-10.684%, while the adjusted determinationcoefficient (Adj.R2) between simulated biomass and measured biomass was0.918, withrelative error being-4.210%-8.801%. Moreover, the root mean square error (RSME) between simulated and measured biomass during the whole growth period of winter wheat was0.541-1.042t/hm~2, with normalized root mean square error being7.087%-13.649%andconformity index being0.986-0.994.(2) By calculating and testing the AquaCrop model parameters according to the springmaize parameters recommended by Hsiao et al., the model could preferably simulate the yield,aboveground biomass and changes in evapotranspiration and biomass during growth stage ofspring maize in the Weibei dry highland under conventional tillage measures. This modelcould be used to simulate the production of spring maize in the Weibei dry highland underconventional tillage measures and optimize the agricultural measures.Under conventional tillage measures, the relative error between simulated value andmeasured value of spring maize yield and aboveground biomass was-2.479%-4.394%and-5.742%-4.571%, respectively. Moreover, the root mean square error (RSME) betweensimulated and measured biomass during the whole growth period of spring maize was0.417-1.187t/hm~2, with normalized root mean square error being3.435%-13.146%andconformity index being0.989-0.999.Under mulching measures, the relative error between simulated value and measuredvalue of spring maize yield and aboveground biomass was11.182%and11.150%,respectively. The root mean square error (RSME) between simulated and measured biomassduring the whole growth period of spring maize was2.257t/hm~2, with normalized root meansquare error being18.157%and conformity index being0.971.The simulation result under conventional tillage measures was better than that ofmulching measures. This was mainly because that the data collected was limited and thecalibration and testing of the model parameters under mulching measures could not beconducted, resulting in comparatively large deviation between simulated and measuredvalues.（3）AquaCrop model can be applied to simulate the yeild and aboveground biomass ofwinter wheat in the Weibei dry highland. But the accuracy of the simulation was greatlyaffected by the initial soil moisture content. Their relationship is still unclear, so furtherstudies are needed.