Abstract: Impacts of climate change on European agricultural production, land use and the environment depend on its impact on crop yields. However, many impact studies assume that crop management remains unchanged in future scenarios, while farmers may adapt their sowing dates and cultivar thermal time requirements to minimize yield losses or realize yield gains. The main objective of this study was to investigate the sensitivity of climate change impacts on European crop yields, land use, production and environmental variables to adaptations in crops sowing dates and varieties’ thermal time requirements. A crop, economic and environmental model were coupled in an integrated assessment modelling approach for six important crops, for 27 countries of the European Union (EU27) to assess results of three SRES climate change scenarios to 2050. Crop yields under climate change were simulated considering three different management cases; (i) no change in crop management from baseline conditions (NoAd), (ii) adaptation of sowing date and thermal time requirements to give highest yields to 2050 (Opt) and (iii) a more conservative adaptation of sowing date and thermal time requirements (Act). Averaged across EU27, relative changes in water-limited crop yields due to climate change and increased CO2 varied between -6 and + 21% considering NoAd management, whereas impacts with Opt management varied between + 12 and + 53%, and those under Act management between 2 and + 27%. However, relative yield increases under climate change increased to + 17 and + 51% when technology progress was also considered. Importantly, the sensitivity to crop management assumptions of land use, production and environmental impacts were less pronounced than for crop yields due to the influence of corresponding market, farm resource and land allocation adjustments along the model chain acting via economic optimization of yields. We conclude that assumptions about crop sowing dates and thermal time requirements affect impact variables but to a different extent and generally decreasing for variables affected by economic drivers.

Abstract: Highlights • The parameters to be fixed were consistent across sites. • Model calibration must be performed separately for each specific case. • Possible to reduce model parameters from 66 to 45. • Strong model reductions must be avoided. • The error term for the training data were characterised by timing (phase shift). Abstract Proper parameterisation and quantification of model uncertainty are two essential tasks in improvement and assessment of model performance. Bayesian calibration is a method that combines both tasks by quantifying probability distributions for model parameters and outputs. However, the method is rarely applied to complex models because of its high computational demand when used with high-dimensional parameter spaces. We therefore combined Bayesian calibration with sensitivity analysis, using the screening method by Morris (1991), in order to reduce model complexity by fixing parameters to which model output was only weakly sensitive to a nominal value. Further, the robustness of the model with respect to reduction in the number of free parameters were examined according to model discrepancy and output uncertainty. The process-based grassland model BASGRA was examined in the present study on two sites in Norway and in Germany, for two grass species (Phleum pratense and Arrhenatherum elatius). According to this study, a reduction of free model parameters from 66 to 45 was possible. The sensitivity analysis showed that the parameters to be fixed were consistent across sites (which differed in climate and soil conditions), while model calibration had to be performed separately for each combination of site and species. The output uncertainty decreased slightly, but still covered the field observations of aboveground biomass. Considering the training data, the mean square error for both the 66 and the 45 parameter model was dominated by errors in timing (phase shift), whereas no general pattern was found in errors when using the validation data. Stronger model reduction should be avoided, as the error term increased and output uncertainty was underestimated.