Ensemble Numerical Experiments with Earth System Models
摘要
Currently, the ensemble approach to climate modeling is widely used. This, in particular, allows one to estimate the characteristics of uncertainty associated with changes in the state of the Earth’s climate system. Often, an already existing ensemble of modern Earth system models is used, for example, the CMIP (Coupled Model Intercomparison Project) ensembles of different generations. It is also reasonable to set up special ensemble numerical experiments with climate models, in which some model parameters or integration conditions are systematically varied. The aim of this paper is a review of the most frequently used methods for constructing and post-processing ensemble numerical experiments with Earth system models, as well as of examples of using the ensemble approach for climate modeling problems. The effect of various sources of uncertainty on the total uncertainty of the modeling results is discussed.