Understanding Intermodel Variability in the Response of Global-Mean Precipitation to CO2 Forcing

Monday, May 12, 2014 - 07:00
Add to Calendar

Though climate models consistently predict an increase in global mean precipitation with increasing carbon dioxide, the amount of increase varies considerably among models, even when normalized by global mean surface temperature change. We attempt here to develop a better physical understanding of the intermodel variability in hydrological sensitivity by analyzing pre-industrial and abrupt quadrupled CO2 simulations in an ensemble of models participating in Phase 5 of the Coupled Model Intercomparison Project (CMIP5). The precipitation response is separated into a component originating directly from increased CO2 and a component that is driven by global surface warming, by regressing global mean precipitation changes onto temperature changes for each model. A substantial portion of the spread in total global hydrological sensitivity originates from the direct response to CO2, with models projecting precipitation decreases from this component ranging from approximately -6 to -3 W/m2. The spread in the temperature-mediated component, ranging from approximately 1.8 to 2.7 W/m2/K, is further explored through analysis of relevant atmospheric quantities. Much of this spread is explained by intermodel differences in the temperature-mediated changes in combined longwave and shortwave net atmospheric cooling. Moreover, intermodel differences in shortwave clear sky atmospheric heating alone appear to make a large contribution to the total spread in the temperature-mediated precipitation response. These findings point toward the potential importance of intermodel divergence in the treatment of radiation, particularly shortwave. The spatial pattern of surface warming and its relation to global mean precipitation sensitivity across models is also investigated. Models with enhanced warming in the southern hemisphere subtropical oceanic regions and suppressed warming in the northern high latitudes, relative to global mean warming, tend to have larger temperature-mediated precipitation responses. This implies that the spatial pattern of warming, possibly through its influence on global evaporation, may play as important a role as atmospheric radiative cooling on the intermodel spread in global hydrological sensitivity. Further analysis is currently underway to better elucidate the physical mechanisms that link intermodel variability in hydrological sensitivity with atmospheric radiation, spatial patterns of climate change, and their potential interaction.

Presentation file(s):