Abstract

Due to anthropogenic climate change, cover crop water consumption in winter could potentially increase drought stress for a succeeding crop. Simulation of cover crop evapotranspiration (ET) losses could be a tool for farmers to make smart management decisions. In Germany, the model AMBAV is used by the German Meteorological Service (DWD) to advise farmers in irrigation management. We compared measured ET of phacelia (Phacelia tanacetifolia), oilseed radish (Raphanus sativus var. oleiformis) and white mustard (Sinapis alba) cultivated in a container experiment with simulated data and conducted a sensitivity analysis to identify the meteorological and crop-specific parameters, which had the strongest effect on simulated ET. In general, measured ET exceeded simulated ET. Different statistical criteria showed that AMBAV performed best for the simulation of evaporation from a bare soil surface. Model performance was also strongly influenced by the irrigation regime in the container experiment. However, the sensitivity analysis showed that changes in irrigation hardly influenced simulated ET. We recommend optimization of the model for irrigated agriculture. Furthermore, we identified temperature and humidity as the most important meteorological and leaf area index as the most important crop-specific parameter for ET simulations with AMBAV. Since farmers' management decisions depend on the accuracy of ET simulations, they should be aware that even small regional deviations of meteorological conditions and soil cover can significantly affect model predictions.