Climate feedback with latitude diagnosed from radiation budgets, temperatures and cloudiness

Abstract

Abstract Estimates of climate feedback (λ, in Wm-2K-1) and equilibrium climate sensitivity typically focus on the global mean values, and are constrained by the time-evolution of the climate system in response to forcing, evaluated from historical observations, palaeo-reconstructions and/or numerical simulations. Here, a framework is presented to evaluate the climate feedback with latitude, λ(ϕ), from the dependence of the radiative response to surface temperature using spatial observations of the climatological mean state. Assuming local cloud type is insensitive to climate state, but allowing cloud amount to vary, we find significant latitudinal variation in climate feedback: hemispheric maxima occur in the subtropics of λ(ϕ)=1.8±0.2 Wm-2K-1 (±1-sigma); while hemispheric minima occur in the midlatitudes of λ(ϕ)= -0.10±0.15 Wm-2K-1 (Northern hemisphere) and λ(ϕ)=0.15±0.22 Wm-2K-1 (Southern hemisphere). The global mean climate feedback estimated from spatially averaging λ(ϕ) comes to λ(ϕ) =1.05±0.13 Wm-2K-1, in good agreement with the latest temporal response-to-forcing derived climate feedback estimate of 1.16±0.4 Wm-2K-1 from the Intergovernmental Panel on Climate Change Assessment Report 6. Our approach diagnosing climate feedback is complementary to existing temporal response-to-forcing estimates as it is based around different underlying assumptions, for example our approach does not require accurate quantification of historic radiative forcing. We anticipate that the presented framework for evaluating latitudinal climate feedback will prove useful across a number of applications, including: analysing and comparing complex Earth system models; exploring Arctic amplification of anthropogenic warming; and exploring palaeo-climate cycles linked to orbital insolation changes at northern mid-latitudes.