12.8 Key uncertainties and research priorities

Uncertainties in future climate projections are discussed in great detail in Working Group I Section 10.5 (Meehl et al., 2007). For Europe, a major uncertainty is the future behaviour of the NAO and North Atlantic THC. Also important, but not specific to Europe, are the uncertainties associated with the still insufficient resolution of GCMs (e.g., Etchevers et al., 2002; Bronstert, 2003), and with downscaling techniques and regional climate models (Mearns et al., 2003; Haylock et al., 2006; Déqué et al., 2007). Uncertainties in climate impact assessment also stem from the uncertainties of land-use change and socio-economic development (Rounsevell et al., 2005, 2006) following European policies (e.g., CAP), and European Directives (Water Framework Directive, European Maritime Strategy Directive). Although most impact studies use the SRES scenarios, the procedures for scenario development are the subject of debate (Castle and Henderson, 2003a, b; Grübler et al., 2004; Holtsmark and Alfsen, 2005; van Vuuren and Alfsen, 2006). While current scenarios appear to reflect well the course of events in the recent past (van Vuuren and O’Neill, 2006), further research is needed to better account for the range of possible scenarios (Tol, 2006). This might be important for Europe given the many economies in transition.

Uncertainties in assessing future climate impacts also arise from the limitations of climate impact models including (i) structural uncertainty due to the inability of models to capture all influential factors, e.g., the models used to assess health impacts of climate change usually neglect social factors in the spread of disease (Kuhn et al., 2004; Reiter et al., 2004; Sutherst, 2004), and climate-runoff models often neglect the direct effect of increasing CO₂ concentration on plant transpiration (Gedney et al., 2006), (ii) lack of long-term representative data for model evaluation, e.g., current vector-monitoring systems are often unable to provide the reliable identification of changes (Kovats et al., 2001). Hence, more attention should be given to structural improvement of models and intensifying efforts of long-term monitoring of the environment, and systematic testing of models against observed data in field trials or catchment monitoring programmes (Hildén et al., 2005). Another way to address the uncertainty of deterministic models is to use probabilistic modelling which can produce an ensemble of scenarios, (e.g., Wilby and Harris, 2006; Araújo and New, 2007; ENSEMBLES project, http://ensembles-eu.metoffice.com/).

Until now, most impact studies have been conducted for separate sectors even if, in some cases, several sectors have been included in the same study (e.g., Schröter et al., 2005). Few studies have addressed impacts on various sectors and systems including their possible interactions by integrated modelling approaches (Holman et al., 2005; Berry et al., 2006). Even in these cases, there are various levels (supra-national, national, regional and sub-regional) that need to be jointly considered, since, if adaptation measures are to be implemented, knowledge down to the lowest decision level will be required. The varied geography, climate and human values of Europe pose a great challenge for evaluation of the ultimate impacts of climate change.

Although there are some good examples, such as the ESPACE-project (Nadarajah and Rankin, 2005), national-scale programmes, such as the FINADAPT project, studies of adaptation to climate change and of adaptation costs are at an early stage and need to be carried out urgently. These studies need to match adaptation measures to specific climate change impacts (e.g., targeted to alleviating impacts on particular types of agriculture, water management or on tourism at specific locations). They need to take into account regional differences in adaptive capacity (e.g., wide regional differences exist in Europe in the style and application of coastal management). Adaptation studies need to consider that in some cases both positive and negative impacts may occur as a result of climate change (e.g., the productivity of some crops may increase, while others decrease at the same location, e.g., Alexandrov et al., 2002). Key research priorities for impacts of climate change, adaptation and implications are included in Table 12.5.