An Investigation of the Effective Climate Sensitivity in GFDL’s New Climate Models CM4.0 and SPEAR

Abstract

Abstract Despite a relatively low climate sensitivity indicated by atmospheric-only simulations with uniform sea surface temperature (SST) warming, GFDL’s new climate model CM4.0 participating in CMIP6 and the seasonal-to-decadal prediction system SPEAR, both of which use an identical atmospheric model AM4.0, produce relatively high effective climate sensitivity (EffCS). The substantial increase in CM4.0’s EffCS is found to be caused by additional positive forcing associated with the CO2 fertilization effect on vegetation, enhanced positive feedback due to stronger reduction in Southern Hemisphere (SH) sea ice concentration (SIC), and clouds whose feedback depends on SST warming patterns. Compared to a SPEAR run using a static vegetation model (SPEAR-SV), CM4.0 produces roughly 30% larger EffCS, among which roughly 1/3 of the increase is due to dynamical vegetation with the rest due primarily to changes in SIC. Although cloud feedback does not explain the key feedback differences among CM4.0, SPEAR, and SPEAR-SV, it is the primary cause of the models’ increase (less negative) in TOA net feedback during the later period of their quadrupling CO2 simulations due to changes in their SST warming patterns. Moreover, CM4.0’s SST warming pattern and its effects on cloud feedback appear to be the leading cause of CM4.0’s EffCS increase compared to the earlier generation GFDL model ESM2M, which produces one of the lowest EffCS values among CMIP5 models. In comparison, CM4.0’s enhanced reduction in SH SICs plays a slightly less important role in its increase in EffCS compared to ESM2M.