Climate models sensitive to tuning of cloud parameters



One of the largest sources of uncertainty in climate model projections of future warming is the effects of clouds. Clouds have both warming and cooling effects, and their interactions with aerosols are complex and not fully understood. To predict future climate using models, researchers commonly tune uncertain cloud parameters to fit current observations, then run the model to project future warming. However, as Golaz et al. show, the choice of parameterization can have a large effect on the simulated warming. The authors studied the effects of tuning cloud parameters in the CMIP5 GFDL CM3 model, a commonly used coupled climate model. They constructed two alternate configurations with plausible but different combinations of parameters. The different configurations showed only very slight differences in modern day climate, but their simulations of warming from preindustrial times to the present differed greatly. This indicates that climate model simulation results depend strongly on cloud parameterization. (Geophysical Research Letters, doi:10.1002/grl.50232, 2013)