Analysing uncertainties in climate change impact assessment across sectors and scenarios

Many models have been developed to explore the likely consequences of climate change. These models tend to focus on single physical or socio-economic sectors and their processes, and neglect the many feedbacks that occur between the different components of the real world. To overcome this problem, models are increasingly being combined in integrated assessment platforms (IAPs), of which the CLIMSAVE IAP is an example, modelling cross-sectoral impacts, adaptation and vulnerability to climate change in Europe by combining 10 different meta-models that focus on specific sectors. Where models are combined in this way, however, attention must be given to the potential errors and uncertainties that integration might introduce. We present a quantitative uncertainty analysis of selected outputs of the CLIMSAVE IAP based on creating and sampling from probability density functions (PDFs) of each of the IAP’s input variables to take account of model and scenario uncertainty. We find limited uncertainties in aggregate outputs of the IAP, which allow specific impacts to be predicted with definable levels of confidence. However, we also find substantial overlap between different socio-economic scenarios at the European scale, suggesting that changes to socio-economic conditions cannot reliably overcome climate-related uncertainty. Nevertheless, there is evidence that particular adaptation actions may significantly alter the impacts of climate change, especially at local or national scales. Copyright Springer Science+Business Media Dordrecht 2015