development; and how all of these efforts can be made most useful for decision making in this decade and beyond.
Examples of the types of strategic questions to be addressed include: What is the appropriate balance between improving resolution and adding complexity as computing power improves? What are the advantages and disadvantages of different approaches to projecting regional climate change (e.g., embedded regional models, statistical downscaling, etc.)? What are the benefits and tradeoffs associated with multimodel versus unified modeling frameworks? What opportunities might exist to develop better interfaces and integration between Earth system models and models of human systems? What observations and process studies are needed to initialize climate predictions on both regional and global scales, advance our understanding of relevant physical processes and mechanisms, and validate model results? What critical infrastructure constraints, including high-performance computing and personnel issues, currently limit model development and use? What steps can be taken to improve the communication of climate model results (e.g., presentation of uncertainties) and ensure that the climate modeling enterprise remains relevant to decision making? What modeling approaches and activities are likely to provide the most value for the investments required?