Same soil - different climate: crop model inter-comparison with lysimeter data of translocated monoliths

Crop model inter-comparisons have mostly been carried out to test the predictive ability under the past range of climatic conditions and for soils of the same site. Unknown is, however, the ability of individual crop models to predict effects of changes in climatic conditions on soil ecosystems beyond the range of site-specific variability. The objective of this study was to test the predictive ability of agro ecosystem models using weighable lysimeter data for the same soil under changed climatic conditions and to compare the simulated crop growth and soil-ecosystem response to climate change between these models. To achieve the objective, data were analyzed from the network of TERENO SOILCan lysimeters for a soil-ecosystem at the original site (Dedelow) and data from the lysimeters containing Dedelow soil monoliths that were transferred to Bad Lauchstädt and Selhausen. For Bad Lauchstädt, this transfer was to a drier and a warmer and for Selhausen to a warmer and wetter site as compared to the original location of the soils at Dedelow. Data time series from the cropped arable soil lysimeters included drier and wetter years and a site-specific crop rotation under comparable management conditions. Identical soil properties and crop growth and boundary conditions were provided for all models after a calibration for the original site at Dedelow, predictions were made for sites Selhausen and Bad Lauchstädt using boundary conditions from those sites. The overall simulation performance of the models was separated in a crop-related part, ecosystem productivity (grain yield, biomass, leaf area index) and in an environmental part, ecosystem fluxes (evapotranspiration, net drainage, soil moisture). When moving soil to a drier region, the crop models’ agronomic and environmental part were well predicted, when moving to wetter regions, only the environmental part of the models seemed to be well predicted. The results suggest considering climate change scenarios, more attention to soil properties and testing of model performance for conditions beyond the calibrated range and site-specific variability will help improving the models.