Title: Understanding the land-ocean contrast in the responses of temperature and the hydrological cycle to climate change

Abstract: Observations and climate models show a pronounced land-ocean contrast in the responses of surface temperature and the hydrological cycle to global warming: Land temperatures increase more than ocean temperatures, low-level relative humidity increases over ocean but decreases over land, and the water cycle has a muted response over land in comparison to ocean regions at similar latitudes. A comprehensive physical understanding of these land-ocean contrasts has not been established, despite the robustness of the features and their importance for the regional and societal impacts of climate change.

Here we investigate land-ocean contrasts in temperature, relative humidity, and precipitation minus evaporation (P-E) under climate change using both idealized and full-complexity models. As in previous studies, we find enhanced surface warming over land relative to the ocean at almost all latitudes. In the tropics and subtropics, the warming contrast is explained using a convective quasi-equilibrium (CQE) theory which assumes equal changes in equivalent potential temperature over land and ocean. As the CQE theory highlights, the warming contrast depends strongly on changes in relative humidity, particularly over land. The decreases in land relative humidity under warming can be understood using a conceptual model of moisture transport between the land and ocean boundary layers and the free troposphere.

Changes in P-E over ocean are closely tied to the local surface-air temperature changes via a simple thermodynamic scaling; the so-called "rich-get-richer" mechanism. Over land, however, we show that the response has a smaller magnitude and deviates substantially from the thermodynamic scaling. We examine the reasons for this land-ocean contrast in the response of P-E by analyzing the atmospheric moisture budget. Horizontal gradients of surface temperature and relative humidity changes are found to be important over land, with changes in atmospheric circulation playing a secondary role. An extended thermodynamic scaling is introduced and is shown to capture the multimodel-mean response of P-E over land. The physical mechanism behind the extended scaling is discussed.

Speaker's website: http://web.mit.edu/byrnem/www/