Numerical experiments are carried out using the GFDL General Circulation Model to assess climate sensitivity associated with CO2 increase and surface warming. This work is motivated by the calculation by Cess and Potter (1988, JGR), who proposed that surface temperature perturbations may be used as a surrogate for climate change induced by CO2 increase.

We compare climatic changes due to CO2 increase in slab-ocean simulations with changes forced by surface warming in prescribed-surface-temperature simulations with fixed CO2 (Cess-type experiments). We found that slab-ocean and Cess-type experiments give the same rates of change per degree surface warming for the global atmosphere temperature and circulation strength. However, the global precipitation increases almost twice as slowly in slab-ocean runs (1.5%/K) when compared to Cess-type runs (2.8%/K). Therefore, we caution that Cess-type experiments may not be suitable for studying global precipitation change under climate change.