Regional and global climate projections increase mid-century yield variability and crop productivity in Belgium

The impact of mid-century climatic changes on crop productivity of winter wheat, maize, potato and sugar beet was assessed for a temperate maritime climate in the Flemish Region, Belgium. Climatic projections of multiple regional and global climate models (RCMs from the EU-ENSEMBLES project and GCMs from the Coupled Model Intercomparison Project phase 3) were stochastically downscaled by the LARS-WG weather generator for use in the crop models AquaCrop and Sirius. Primarily positive effects on mean yield were simulated. Crops benefitted from elevated CO2, and from more radiation interception if the cropping period was adapted in response to higher temperatures. However, increased productivity was linked with increased susceptibility to water stress and greater inter-annual yield variability, particularly with adapted management. Impacts differed among and within ensembles of climate models, and among crops and environments. Although RCMs may be more suitable for local impact assessments than GCMs, inter-ensemble differences and contingent wider ranges of impacts with GCM projections found in this study indicate that applying RCMs driven by a limited number of GCMs alone would not give the full range of possible impacts. Further, this study suggests that the simulated intermodel variation can be larger than spatial variation within the region. These findings advocate the use of both GCM and RCM ensembles in assessments where temperature and precipitation are central, such as for crop production.