Abstract: We assessed the weather data aggregation effect (DAE) on the simulation of cropping systems for different crops, response variables, and production conditions. Using 13 process-based crop models and the ensemble mean, we simulated 30 yr continuous cropping systems for 2 crops (winter wheat and silage maize) under 3 production conditions for the state of North Rhine-Westphalia, Germany. The DAE was evaluated for 5 weather data resolutions (i.e. 1, 10, 25, 50, and 100 km) for 3 response variables including yield, growing season evapotranspiration, and water use efficiency. Five metrics, viz. the spatial bias (Delta), average absolute deviation (AAD), relative AAD, root mean squared error (RMSE), and relative RMSE, were used to evaluate the DAE on both the input weather data and simulated results. For weather data, we found that data aggregation narrowed the spatial variability but widened the., especially across mountainous areas. The DAE on loss of spatial heterogeneity and hotspots was stronger than on the average changes over the region. The DAE increased when coarsening the spatial resolution of the input weather data. The DAE varied considerably across different models, but changed only slightly for different production conditions and crops. We conclude that if spatially detailed information is essential for local management decision, higher resolution is desirable to adequately capture the spatial variability for heterogeneous regions. The required resolution depends on the choice of the model as well as the environmental condition of the study area.

Abstract: In estimating responses of crops to future climate realisations, it is necessary to understand and differentiate between the sources of uncertainty in climate models and how these lead to errors in estimating the past climate and biases in future projections, and how these affect crop model estimates. This paper investigates the complexities in using climate model projections representing different spatial scales within climate change impacts and adaptation studies. This is illustrated by simulating spring barley with three crop models run using site-specific observed, original (50•50 km) and bias corrected downscaled (site-specific) hindcast (1960-1990) weather data from the HadRM3 Regional Climate Model (RCM). Original and bias corrected downscaled weather data were evaluated against the observed data. The comparisons made between the crop models were in the light of lessons learned from this data evaluation. Though the bias correction downscaling method improved the match between observed and hindcast data, this did not always translate into better matching of crop models estimates. At four sites the original HadRM3 data produced near identical mean simulated yield values as from the observed weather data, despite differences in the weather data, giving a situation of ‘right results for the wrong reasons’. This was likely due to compensating errors in the input weather data and non-linearity in crop models processes, making interpretation of results problematic. Overall, bias correction downscaling improved the quality of simulated outputs. Understanding how biases in climate data manifest themselves in crop models gives greater confidence in the utility of the estimates produced using downscaled future climate projections. The results indicate implications on how future projections of climate change impacts are interpreted. Fundamentally, considerable care is required in determining the impact weather data sources have in climate change impact and adaptation studies, whether from individual models or ensembles. No Label

Abstract: Crop growth simulation models can differ greatly in their treatment of key processes and hence in their response to environmental conditions. Here, we used an ensemble of 26 process-based wheat models applied at sites across a European transect to compare their sensitivity to changes in temperature (−2 to +9°C) and precipitation (−50 to +50%). Model results were analysed by plotting them as impact response surfaces (IRSs), classifying the IRS patterns of individual model simulations, describing these classes and analysing factors that may explain the major differences in model responses.   The model ensemble was used to simulate yields of winter and spring wheat at sites in Finland, Germany and Spain. Results were plotted as IRSs that show changes in yields relative to the baseline with respect to temperature and precipitation. IRSs of 30-year means and selected extreme years were classified using two approaches describing their pattern.   The expert diagnostic approach (EDA) combines two aspects of IRS patterns: location of the maximum yield (nine classes, Figure 1) and strength of the yield response with respect to climate (four classes), resulting in a total of 36 combined classes defined using criteria pre-specified by experts. The statistical diagnostic approach (SDA) groups IRSs by comparing their pattern and magnitude, without attempting to interpret these features. It applies a hierarchical clustering method, grouping response patterns using a distance metric that combines the spatial correlation and Euclidian distance between IRS pairs. The two approaches were used to investigate whether different patterns of yield response could be related to different properties of the crop models, specifically their genealogy, calibration and process description.   Although no single model property across a large model ensemble was found to explain the integrated yield response to temperature and precipitation perturbations, the application of the EDA and SDA approaches revealed their capability to distinguish: (i) stronger yield responses to precipitation for winter wheat than spring wheat; (ii) differing strengths of response to climate changes for years with anomalous weather conditions compared to period-average conditions; (iii) the influence of site conditions on yield patterns; (iv) similarities in IRS patterns among models with related genealogy; (v) similarities in IRS patterns for models with simpler process descriptions of root growth and water uptake compared to those with more complex descriptions; and (vi) a closer correspondence of IRS patterns in models using partitioning schemes to represent yield formation than in those using a harvest index.   Such results can inform future crop modelling studies that seek to exploit the diversity of multi-model ensembles, by distinguishing ensemble members that span a wide range of responses as well as those that display implausible behaviour or strong mutual similarities.   The full manuscript of this study is currently under revision (Fronzek et al. 2017).