A BIMODAL CLIMATE RESPONSE CONTROLLED BY WATER VAPOR TRANSPORT IN A COUPLED OCEAN-ATMOSPHERE BOX MODEL

The importance of the hydrological cycle as a controlling factor on the magnitude of the thermohaline circulation is illustrated in a simple one-hemisphere coupled ocean-atmosphere box model. The ocean model includes differential surface heating and evaporation, horizontal and vertical exchange of heat and salt between boxes, and a simply parameterized thermohaline circulation. Surface heat fluxes and evaporation a re determined through the coupled ocean and energy balance atmosphere models which treat fluxes of long-and short-wave radiation and sensible and latent heat. Two parameters represent the most important physics: mu controls the magnitude of the thermohaline circulation; epsilon controls the strength of the hydrological cycle. For fixed mu, two regimes are distinguished. One, associated with small values of epsilon, has weak latitudinal water vapor transport in the atmosphere, a strong thermohaline circulation with sinking in high latitudes, upwelling in low latitudes, and strong latitudinal transport of heat by the ocean. The second regime for larger epsilon is characterized by strong latitudinal water vapor transport which, by reducing the surface salinity in high latitudes, shuts down the thermohaline circulation and has reduced ocean and net latitudinal heat transport. The bimodal response in the model is shown to be the consequence of a shift in the mechanism of supply of salt to the high-latitude surface ocean from predominantly thermohaline transport, a nonlinear process, to or from predominantly eddy mixing transport, a linear process. In climatological terms, the bimodality represents two distinct climate regimes, one with an active ocean meridional circulation and relatively warm ocean and atmosphere temperatures in high latitudes, and the other with a less active ocean circulation and an increased latitudinal temperature gradient in atmosphere and ocean. The regime with an active thermohaline circulation tends to be less stable than the other, exhibiting over a range of e a "halocline catastrophe" to perturbations in surface salinity. In many respects the model supports current concepts concerning the role of bimodal ocean physics and atmospheric water vapor transport in glacial to interglacial climate changes and in the more rapid events such as the Younger Dryas.