Dual crop coefficient modelling applied to the winter wheat–summer maize crop sequence in North China Plain: Basal crop coefficients and soil evaporation component

The dual crop coefficient (Kc) approach to estimate crop evapotranspiration (ETc) separately considers soil evaporation (E) and plant transpiration (T) by computing a soil evaporation coefficient (Ke) and a basal crop coefficient (Kcb), respectively, with Kc=Ke+Kcb. This approach may be more precise than the single Kc approach particularly when the crops incompletely cover the ground. The SIMDualKc model, which is adopted in this study, is an irrigation scheduling simulation model that uses a daily time-step for performing two separate soil water balances, one for the soil evaporation layer from which Ke is computed, and the other for the entire root zone, thus allowing to compute the actual Kcb adjusted to the soil moisture conditions (Kcbadj). The standard Kcb is corrected to the climate, crop density and height. Two years of field experimental data relative to winter wheat and summer maize were used for model calibration and validation using soil water content data observed with time-domain reflectometry (TDR) in a silt loam soil. Field data also include E measured with microlysimeters placed along the crop rows. The calibration procedure consisted in adjusting the basal crop coefficients, the soil evaporation parameters used to compute Ke, and the soil water depletion fraction for no stress (p) to achieve the best fit of the observed soil water content data. The calibrated Kcb values for winter wheat were 0.25 for the initial and the soil frozen period, 1.15 for the mid-season and 0.30 at harvesting. For the summer maize, the initial, mid season and end season Kcb were respectively 0.2, 1.10 and 0.45. Model results have shown a good agreement between model predictions and field observations of the soil water content of both crops, with root mean square errors of estimates (RMSE) of about 0.01m3m−3 for both the calibration and validation. The modelling efficiency EF and the index of agreement dIA were larger than 0.96 and 0.99, respectively, thus indicating good performance of modelling with SIMDualKc. Model estimates of E using Ritchie's approach were compared with microlysimeter data; for winter wheat a RMSE=0.37mmd−1 was obtained, while for maize RMSE of 0.45 and 0.49mmd−1 were obtained for both years of observations. Results for soil evaporation allow confirming the appropriateness of using Ritchie's model to estimate soil evaporation of a cropped soil. E averaged 124mm for wheat, representing 29% of ETc, and 146mm for summer maize, i.e. 41% of ETc. In conclusion, results show that the model is appropriate to simulate the soil water balance adopting the dual Kc approach and may be further used to develop improved irrigation schedules for the winter wheat–summer maize crop sequence in North China.