A130-07
Historical radiative feedbacks accurately quantified using RFMIP simulations: coupled CMIP6 models show higher effective climate sensitivity than AMIP

Friday, 11 December 2020: 16:24
Virtual
Yue Dong, University of Washington, Atmospheric Sciences, Seattle, United States, Kyle Armour, University of Washington, Dept. of Atmospheric Sciences and School of Oceanography, Seattle, WA, United States, David Battisti, University of Washington, Atmospheric Sciences, Seattle, WA, United States, Timothy Andrews, Met Office Hadley Centre, Exeter, United Kingdom, Christopher James Smith, Center for International Climate and Environmental Research Oslo, Oslo, Norway, Piers Forster, University of Leeds, School of Earth and Environment, Leeds, United Kingdom, Cristian Proistosescu, University of Washington, JISAO, Seattle, WA, United States, David Paynter, NOAA GFDL, Princeton, NJ, United States and Hideo Shiogama, National Institute for Environmental Studies, Center for Global Environmental Research, Tsukuba, Japan
Recorded Presentation
Abstract:
Previous studies have shown that the radiative feedbacks within atmospheric global climate models (AGCMs) forced by historical observed sea-surface temperatures (SSTs) are generally more negative than those within fully-coupled GCMs forced by CO2 quadrupling. However, radiative feedbacks have generally not been accurately estimated within fully-coupled GCMs driven by realistic historical forcing, owing to the fact that historical effective radiative forcing (ERF) is not quantified for many GCMs.

As the Radiative Forcing Model Intercomparison Project (RFMIP) becomes available, a new set of simulations permit accurate quantification of ERF and thus also of radiative feedbacks within multiple fully-coupled GCMs. In this study, we make use of the RFMIP simulations in conjunction with fully-coupled historical simulation to provide estimates of historical radiative feedbacks for seven CMIP6 GCMs. Moreover, we make use of the amip simulations within AGCMs, permitting an evaluation of feedbacks when the same atmospheric models are driven by observed SSTs.

The results show a notable difference in the trend of net radiative feedback over the historical period between fully-coupled GCMs and their counterpart AGCMs forced by observed SSTs. In particular, over the past three decades, feedbacks become less stable within the coupled GCMs (corresponding to a higher effective climate sensitivity), while they become more stable within the AGCMs (corresponding to a lower effective climate sensitivity). This discrepancy in the evolution of feedback strength is traced to the differences in the pattern of SST trends simulated by the coupled GCMs and that observed, which highlight the largest discrepancies within the tropical Pacific and Southern Ocean.

<< Previous Abstract | Next Abstract >>