On the use of the Fluctuation Dissipation Theorem to predict climate sensitivity

The availability of high spatial and spectral resolution infrared radiances from instruments such as AIRS places renewed emphasis on the development of analytic tools for the comparison of model output and climate date. It has previously been shown that the Fluctuation Dissipation Theorem (FDT) may allow estimation of climate sensitivity by calculation of the lag covariance of climate forcings and temperature. Here we use a simple climate model to test the utility of this approach. We analyze the behavior of the model subject to various stochastic forcings and for various climate sensitivities, modulated by either albedo or emissivity feedbacks. We compare the equilibrium model sensitivity to these forcings with the sensitivity derived from the FDT. The ability of the FDT to predict sensitivity in a reasonably short time is found to depend strongly on the heat capacity of the model ocean, and on strength of the thermal coupling between the model ocean surface and the deep ocean. We also investigate the robustness of the method in situations where an observed quantity both forces and responds to climate variations. The figure below shows the increasing error of the FDT estimate of climate sensitivity as the coupling between the surface and the deep ocean increases. In this case, the deep ocean is assumed to have a heat capacity equivalent to a depth of 1 km, while the surface layer is assumed to have a depth of 10 m.