Read here. As all empirical climate research demonstrates, climate models are atrocious prediction tools. The latest example is a peer-reviewed study by a team of 9 researchers from the U.S., the UK and Australia that found climate models to be worthless regarding precipitation levels and intensity.
"Stephens et al. determined that "the character of liquid precipitation (defined as a combination of accumulation, frequency, and intensity) over the global oceans is significantly different from the character of liquid precipitation produced by global weather and climate models," noting that "the differences between observed and modeled precipitation are larger than can be explained.....say their results imply that state-of-the-art weather and climate models have "little skill in precipitation calculated at individual grid points," and that "applications involving downscaling of grid point precipitation to yet even finer-scale resolution has little foundation and relevance to the real earth system," which is not too encouraging a result, considering it is the "real earth system" in which we live and for which we have great concern." [Graeme L. Stephens, Tristan L'Ecuyer, Richard Forbes, Andrew Gettlemen, Jean-Christophe Golaz, Alejandro Bodas-Salcedo, Kentaroh Suzuki, Philip Gabriel, John Haynes, J. 2010: Journal of Geophysical Research]
Additional peer-reviewed postings. Other climate-model postings.