Mixed cropping constitutes an interesting cropping system for the cultivation of energy plants and the prevention of water pollution due to increased water and nutrient efficiency compared to conventional sole cropping. In a three-year field experiment, the sole crops maize (Ma) and sorghum (So), as well as mixtures of maize with scarlet runner bean (MaB) and sorghum with lupine (SoLu), were grown in a rainout shelter under optimal and suboptimal water availability with the aim of investigating their hydrological and agronomic characteristics. The targets of the present contribution are (1) analysing and quantifying the soil water balance of the different cropping treatments, (2) verifying that a well-proven hydrological model can be used to simulate hydrological processes under those specific experimental conditions (rainout shelter), and (3) examining whether mixed cropping is more efficient in using the growing factor water, evaluated by the water use efficiency (WUE), than sole cropping. The hydrological process analysis is based on measurements of groundwater table and irrigation. The observational data show that the evapotranspiration rate (ETR) of Ma was smaller than that of MaB. In contrast, the ETR of SoLu was similar to So. The experimental model reproduction with the SWAP model allowed a further investigation of the soil water balance by simulating spatially and temporally subtle hydrological process patterns. A bi-objective parameter optimization was used to improve the model performance by calibrating soil hydraulic model parameters against (1) soil water potential observed in the rainout shelter experiment, and (2) measured soil water retention curves at the study site. The attained Nash-Sutcliffe efficiencies for the predicted soil water potential range from 0.43 to 0.77. These measures prove that the usage of a hydrological model in the specific environment of a rainout shelter is possible. The simulation results showed that capillary rise is important at the study site, serving partially as compensator of drought stress. The WUE estimated by simulated ETR and measured yield differed among the crops, but was independent of the water availability. In individual cases, the WUE in mixed cropping was higher than in sole cropping. A general more efficient usage of the growing factor water by means of mixed cropping could not be stated from the results.
