Abstract: Crop models of crop growth are increasingly used to quantify the impact of global changes due to climate or crop management. Therefore, accuracy of simulation results is a major concern. Studies with ensembles of crop models can give valuable information about model accuracy and uncertainty, but such studies are difficult to organize and have only recently begun. We report on the largest ensemble study to date, of 27 wheat models tested in four contrasting locations for their accuracy in simulating multiple crop growth and yield variables. The relative error averaged over models was 24-38% for the different end-of-season variables including grain yield (GY) and grain protein concentration (GPC). There was little relation between error of a model for GY or GPC and error for in-season variables. Thus, most models did not arrive at accurate simulations of GY and GPC by accurately simulating preceding growth dynamics. Ensemble simulations, taking either the mean (e-mean) or median (e-median) of simulated values, gave better estimates than any individual model when all variables were considered. Compared to individual models, e-median ranked first in simulating measured GY and third in GPC. The error of e-mean and e-median declined with an increasing number of ensemble members, with little decrease beyond 10 models. We conclude that multimodel ensembles can be used to create new estimators with improved accuracy and consistency in simulating growth dynamics. We argue that these results are applicable to other crop species, and hypothesize that they apply more generally to ecological system models.

Abstract: Climate change is anticipated to affect European agriculture, including the risk of emerging or re-emerging feed and food hazards. Indirectly, climate change may influence such hazards (e.g. the occurrence of mycotoxins) due to geographic shifts in the distribution of major cereal cropping systems and the consequences this may have for crop rotations. This paper analyses the impact of climate on cropping shares of maize, oat and wheat on a 50-km square grid across Europe (45-65°N) and provides model-based estimates of the changes in cropping shares in response to changes in temperature and precipitation as projected for the time period around 2040 by two regional climate models (RCM) with a moderate and a strong climate change signal, respectively. The projected cropping shares are based on the output from the two RCMs and on algorithms derived for the relation between meteorological data and observed cropping shares of maize, oat and wheat. The observed cropping shares show a south-to-north gradient, where maize had its maximum at 45-55°N, oat had its maximum at 55-65°N, and wheat was more evenly distributed along the latitudes in Europe. Under the projected climate changes, there was a general increase in maize cropping shares, whereas for oat no areas showed distinct increases. For wheat, the projected changes indicated a tendency towards higher cropping shares in the northern parts and lower cropping shares in the southern parts of the study area. The present modelling approach represents a simplification of factors determining the distribution of cereal crops, and also some uncertainties in the data basis were apparent. A promising way of future model improvement could be through a systematic analysis and inclusion of other variables, such as key soil properties and socio-economic conditions, influencing the comparative advantages of specific crops.

Abstract: Assessment of climate change impacts on crops in regions of complex orography such as the Iberian Peninsula (IP) requires climate model output which is able to describe accurately the observed climate. The high resolution of output provided by Regional Climate Models (RCMs) is expected to be a suitable tool to describe regional and local climatic features, although their simulation results may still present biases. For these reasons, we compared several post-processing methods to correct or reduce the biases of RCM simulations from the ENSEMBLES project for the IP. The bias-corrected datasets were also evaluated in terms of their applicability and consequences in improving the results of a crop model to simulate maize growth and development at two IP locations, using this crop as a reference for summer cropping systems in the region. The use of bias-corrected climate runs improved crop phenology and yield simulation overall and reduced the inter-model variability and thus the uncertainty. The number of observational stations underlying each reference observational dataset used to correct the bias affected the correction performance. Although no single technique showed to be the best one, some methods proved to be more adequate for small initial biases, while others were useful when initial biases were so large as to prevent data application for impact studies. An initial evaluation of the climate data, the bias correction/reduction method and the consequences for impact assessment would be needed to design the most robust, reduced uncertainty ensemble for a specific combination of location, crop, and crop management.