DECADAL CLIMATE PREDICTION

摘要: The importance of improved information about near-term (from 1 year to several decades in advance) regional climate for many societal applications has prompted considerable research in the field of decadal climate prediction that addresses those time scales. Meehl et al.(2009a) outlined the problems and issues involved with decadal climate prediction, and reviewed some of the first results from the few model studies that had been performed up to that time. The purpose of this paper is to provide scientists and possible users of such near-term climate information with an update of research in this rapidly evolving field since the Meehl et al.(2009a) paper, including recent multimodel studies and Coupled Model Intercomparison Project phase 5 (CMIP5) results. CMIP5 includes, as part of the standard multimodel experiments to be run for analysis and comparison, a set of decadal climate prediction experiments—both hindcasts and predictions (Taylor et al. 2012). This multimodel ensemble represents a major contribution to climate science in general and decadal climate prediction in particular in allowing this coordinated set of initialized hindcasts and predictions to be made available to the international climate science community to advance our understanding of the decadal climate prediction problem.For example, in part due to results from the initialized decadal climate predictions from CMIP5, the assessed range of near-term global warming (2016–2035) was less than the uninitialized simulations in the IPCC AR5 (Kirtman et al. 2013). To provide an example of a group of users who could conceivably make use of decadal climate predictions …