Quantifying Local Radiative Feedbacks at the Sea Surface

Abstract The spatial pattern of sea surface temperature (SST) change, which is important in determining the Earth’s energy budget under climate change, disagrees in observations and simulations. The spatial inhomogeneity of local feedback has been identified as an important contributor to the evolution of SST pattern; however, traditional regression methods are inaccurate in quantifying local feedbacks in the tropics. In this paper, we quantify the magnitude of local feedbacks at the sea surface using SST patch perturbation experiments, with a comparison of the results with those calculated using simple linear regression. The spatial patterns of local feedbacks calculated from these two methods are generally consistent, while the regression method is inaccurate for tropical latent heat and sensible heat feedbacks. The results show that local feedback at the sea surface is consistently negative, which prevents the SST from experiencing unlimited warming. Tropical ascent regions show the strongest negative local feedback in both zonal and meridional directions, and local latent heat feedback and cloud radiative feedback are the most important contributors to the spatial inhomogeneity of surface feedback. Local radiative feedbacks calculated using linear regression are consistent in simulations and observations, and the spatial pattern of local feedbacks is also consistent with future SST change pattern predicted by climate models, indicating that the El Nino-like future SST change pattern in climate models is likely plausible. However, climate models appear to overestimate the negative local radiative feedback compared to reanalysis and observation-based satellite data, which might potentially contribute to the differences between observed and model-simulated SST change patterns during the last several decades.