Evaluation of the CropSyst Model during Wheat-Maize Rotations on the North China Plain for Identifying Soil Evaporation Losses

The North China Plain (NCP) is a major grain production zone that plays a critical role in ensuring China's food supply. Irrigation is commonly used during grain production; however, the high annual water deficit [precipitation (P) minus evapotranspiration (ET)] in typical irrigated cropland does not support double cropping systems (such as maize and wheat) and this has resulted in the steep decline in the water table (~0.8 m year-1 at the Luancheng station) that has taken place since the 1970s. The current study aimed to adapt and check the ability of the CropSyst model (Suite-4) to simulate actual evapotranspiration (ETa), biomass, and grain yield, and to identify major evaporation (E) losses from winter wheat (WW) and summer maize (SM) rotations. Field experiments were conducted at the Luancheng Agro-ecosystem station, NCP, in 2010-2011 to 2012-2013. The CropSyst model was calibrated on wheat/maize (from weekly leaf area/biomass data available for 2012-2013) and validated onto measured ETa, biomass, and grain yield at the experimental station from 2010-2011 to 2011-2012, by using model calibration parameters. The revalidation was performed with the ETa, biomass, grain yield, and simulated ETa partition for 2008-2009 WW [ETa partition was measured by the Micro-lysimeter (MLM) and isotopes approach available for this year]. For the WW crop, E was 30% of total ETa; but from 2010-11 to 2013, the annual average E was ~40% of ETa for the WW and SM rotation. Furthermore, the WW and SM rotation from 2010-2011 to 2012-2013 was divided into three growth periods; (i) pre-sowing irrigation (PSI; sowing at field capacity) to emergence period (EP), (ii) EP to canopy cover period (CC) and (iii) CC to harvesting period (HP), and E from each growth period was ~10, 60, and 30%, respectively. In general, error statistics such as RMSE, Willmott's d, and NRMSE in the model evaluation for wheat ETa (maize ETa) were 38.3 mm, 0.81, and 9.24% (31.74 mm, 0.73, and 11.89%); for wheat biomass (maize biomass) they were 1.25 Mg ha-1, 0.83, and 9.64% (0.78 Mg ha-1, 0.96, and 7.96%); and for wheat grain yield (maize grain yield) they were 0.65 Mg ha-1, 0.82, and 9.87% (0.2 Mg ha-1, 0.99, and 3.79%). The results showed that CropSyst is a valid model that can be use with a reliable degree of accuracy for optimizing WW and SM grain yield production and water requirement on the NCP.