Applying a site based crop model to estimate regional yields under current and changed climates

A method of upscaling a site-based crop model to obtain regional and national results on spring wheat productivity under changing climate is presented. The model, CERES-Wheat, was calibrated and validated first at sites, and in the upscaling phase it was run across a regular 10×10 km grid over Finland. In the grid the model was run both for the present-day (1961–1996) climate and scenarios of future climate for 2050. Regional averages were computed for the years 1981–1996 to be comparable to yield observations in the farm yield statistics. CERES-Wheat does not consider crop stress caused by poor soil aeration under wet conditions; however, this was found to be crucial to obtain satisfactory simulation results at sites. The model was modified accordingly, and the new version labelled ‘CERES-Wet-Wheat’. The results indicate that CERES-Wet-Wheat was able to detect both at the site and regional scale yield variations dependent on climate, even though the approach did not consider variations in crop management, pests and diseases and soil dependent differences in initial conditions. A regional approach to estimate future yields is useful especially at northern latitudes where crop suitability is likely to shift under changing climate. However, in a regional assessment there are many uncertainties associated with the yield estimates, including uncertainties in projections of future climate, model errors and assumptions and observation errors. For this reason, the underlying assumptions need to be clear when the estimates in yield changes are interpreted.