Evaluating and Improving Regional Climate Projections – a ...

1 Evaluating and Improving Regional Climate Projections a Workshop report from Dr Filippo Giorgi From 11-13 February 2009, representatives of the communities dealing with Regional modelling, downscaling and Climate impact assessments met in Toulouse, France, for a Workshop on Evaluating and Improving Regional Climate Projections . Dr Filippo Giorgi from the Abdus Salam International Centre for Theoretical Physics in Trieste, Italy, chairs the WCRP Task Force on Regional Climate Downscaling (TF-RCD) together with Dr Colin Jones from the Rossby Center in Norrk ping, Sweden. The Workshop was attended by about 50 people, with representation of a number of global and Regional modelling centres, statistical downscaling groups and impact assessment scientists. The meeting provided a forum for summarising the attributes of various downscaling methods and discussing the extent to which these methods address the needs of a broad and very inhomogeneous user community.

2 But, the main objective of the workshop was to start laying out a plan to set up a coordinated Regional Climate downscaling (RCD) experiment with the goals to 1) evaluate, intercompare and possibly improve different RCD techniques; and to 2) produce a new set of RCD-based Projections for use in impact assessment/adaptation work and for input to the upcoming IPCC Fifth Assessment Report (AR5). All participants enthusiastically shared the vision and agreed goals, which is already in itself a good sign. After extensive discussions, the following decisions were reached: 1) We should define a general framework for the community to refer to (something along the lines of what the Atmospheric Model Intercomparison Project (AMIP) and the Coupled Model Intercomparison Project (CMIP) have been for General Circulation Models, GCMs) but we should start with a more limited effort to make sure that achievable goals are set and achieved, specifically within the timeline of the AR5 which will be published in 2013.

3 2) In terms of the evaluation and benchmarking experiments it was decided to use the ERA-Interim product (1989-2007) to produce lateral boundary conditions for Regional Climate Models (RCMs) and driving data for statistical downscaling (in this case recognizing that additional data might be used because of the short length of the ERA-Interim), until more extensive re-analysis products become available. 3) The initial standard resolution will be 50 km to enhance the feasibility of the experiments and the involvement of a wider community. 4) Although a comprehensive set of domains will be considered in the matrix of experiments, in particular covering all developing country regions, the initial focus will be on Africa and more specifically on a domain covering the entire African continent. 5) In terms of GCM-driven experiments, the first priority will be on the long term scenario simulations, and more specifically the periods 1950-2005 historical run 2005-2100 scenario run (RCP: Representative Concentration Pathway) 2005-2100 scenario run.

4 6) Multiple GCMs should be used for driving the RCD in order to capture the inter-model uncertainty. A number of issues remained unresolved, and it is the task of the Task Force on Regional Climate Downscaling (TF-RCD), created in 2008 by WCRP through the Working Group on Coupled Modelling (WGCM), to follow up on them in the next few months leading to the Lund meeting (4-8 May 2009), where the TF-RCD will meet again. The consensus was evident in the room that the initial set up described above represented an optimal compromise and doable first step towards the definition of the broader framework. [fg] Links and References: Workshop on Evaluating and Improving Regional Climate Projections Task Force on Regional Climate Downscaling (TFRCD)