Evaluation of continental precipitation in 20th century climate simulations: The utility of multimodel statistics

In support of the Intergovernmental Panel on Climate Change (IPCC), simulations of 20th century climate have been performed recently with some 20 global coupled ocean-atmosphere models. In view of its central importance for biological and socioeconomic systems, model-simulated continental precipitation is evaluated relative to three observational estimates at both global and regional scales. Many models are found to display systematic biases, deviating markedly from the observed spatial variability and amplitude/phase of the seasonal cycle. However, the pointwise ensemble mean of all the models usually shows better statistical agreement with the observations than does any single model. Deficiencies of current models that may be responsible for the simulated precipitation biases as well as possible reasons for the improved estimate afforded by the multimodel ensemble mean are discussed. Implications of these results for water resource managers also are briefly addressed.