Abstract

The energy budget of the modern-day Southern Hemisphere is poorly simulated in both state-of-the-art reanalyses and coupled global climate models. The ocean-dominated Southern Hemisphere has low surface reflectivity and therefore its albedo is particularly sensitive to cloud cover. In modern-day ... Show moreThe energy budget of the modern-day Southern Hemisphere is poorly simulated in both state-of-the-art reanalyses and coupled global climate models. The ocean-dominated Southern Hemisphere has low surface reflectivity and therefore its albedo is particularly sensitive to cloud cover. In modern-day climates, mainly because of systematic deficiencies in cloud and albedo at mid- and high latitudes, too much solar radiation enters the ocean. Along with too little radiation absorbed at lower latitudes because of clouds that are too bright, unrealistically weak poleward transports of energy by both the ocean and atmosphere are generally simulated in the Southern Hemisphere. This implies too little baroclinic eddy development and deficient activity in storm tracks. However, projections into the future by coupled climate models indicate that the Southern Ocean features a robust and unique increase in albedo, related to clouds, in association with an intensification and poleward shift in storm tracks that is not observed at any other latitude. Such an increase in cloud may be untenable in nature, as it is likely precluded by the present-day ubiquitous cloud cover that models fail to capture. There is also a remarkably strong relationship between the projected changes in clouds and the simulated current-day cloud errors. The model equilibrium climate sensitivity is also significantly negatively correlated with the Southern Hemisphere energy errors, and only the more sensitive models are in the range of observations. As a result, questions loom large about how the Southern Hemisphere will actually change as global warming progresses, and a better simulation of the modern-day climate is an essential first step. Show less