Estimating climate sensitivity using two‐zone energy balance models

Abstract Estimates of 2 × CO 2 equilibrium climate sensitivity (EqCS) derive from running global climate models (GCMs) to equilibrium. Estimates of effective climate sensitivity (EfCS) are the corresponding quantities obtained using transient GCM output or observations. The EfCS approach uses an accompanying energy balance model (EBM), the zero‐dimensional model (ZDM) being standard. GCM values of EqCS and EfCS vary widely [Intergovernmental Panel on Climate Change range: (1.5, 4.5)°C] and have failed to converge over the past 35 years. Recently, attempts have been made to refine the EfCS approach by using two‐zone (tropical/extratropical) EBMs. When applied using satellite radiation data, these give low and tightly constrained EfCS values, in the neighborhood of 1°C. These low observational EfCS/two‐zone EBM values have been questioned because (a) they disagree with higher observational EfCS/ZDM values and (b) the EfCS/two‐zone EBM values given by GCMs are poorly correlated with the standard GCM sensitivity estimates. The validity of the low observational EfCS/two‐zone EBM values is here explored, with focus on the limitations of the observational EfCS/ZDM approach, the disagreement between the GCM and observational radiative responses to surface temperature perturbations in the tropics, and on the modified EfCS values provided by an extended two‐zone EBM that includes an explicit parameterization of dynamical heat transport. The results support the low observational EfCS/two‐zone EBM values, indicating that objections (a) and (b) to these values both need to be reconsidered. It is shown that in the EBM with explicit dynamical heat transport the traditional formulism of climate feedbacks can break down because of lack of additivity.