Robust Impacts of Climate Change in Europe and Why Study Scale is Important for Adaptation

Friday, 18 December 2015
Poster Hall (Moscone South)
Chantal Donnelly, Jafet Andersson, Jonas Olsson, Thomas Bosshard, Wei Yang, Peter Berg and Berit Arheimer, Swedish Meteorological and Hydrological Institute, Norrköping, Sweden
Impacts of climate change on water resources in Europe have been studied using multiple climate, hydrological and downscaling models and at multiple scales. Although results seem to differ largely between these studies, robust qualitative results have emerged at the European scale. Generally, a drying trend coupled with more intense extremes (floods & droughts) is observed in southern Europe, whereas a wetting trend coupled with less intense extremes is observed in Northern Europe. The location of the change between wetting and drying leads to uncertainty of climate change impacts in central Europe. Also, temperature-related hydrological processes lead to more robust predictions than precipitation-related processes as climate models are more consistent for temperature. For European hydrology, this leads to more robust predictions in regions with snow-dominated hydrology and where evapotranspiration dominates the water cycle. Robust predictions of changes to the seasonality of discharge are seen in snow-dominated regions (Fennoscandinavia, Alps) while robust predictions of decreases in runoff are seen for the Iberian peninsula. While uncertainty in the projections mostly comes from climate uncertainty, predictions of impacts on soil moisture and low flows can be largely dependent on the choice of hydrological model where methods to estimate evapotranspiration and runoff differ widely. While the regional-scale results show some robustness, there can be large local-scale differences even where the same catchment is considered at different scales. Furthermore, understanding of uncertainties due to correction and downscaling procedures are only beginning to emerge. Our evaluations of bias-correction indicate that uncertainties vary considerably depending on the variable. The scale of uncertainty due to bias-correction can be similar to the projected climate changes and to the uncertainty from the climate models. In this presentation we put forward a new bottom-up method for climate change impact studies where the regional-scale results and their known uncertainties are considered together with local-scale studies designed specifically for the changes to which a system must be adapted. We illustrate the value of the new method in a number of real case studies for various societal sectors.