sea-surface temperatures (SSTs) and climate The atmosphere interacts with its boundary layer, which on Earth is 30 per cent land and 70 per cent ocean. Not surprisingly, interaction between the atmosphere and the ocean is important both for the climatology and the variability of weather and climate. This interaction has only recently acquired the status it deserves in climatology, as is evidenced by the now completed long-term international collaborative study entitled ‘Tropical Ocean Global Atmosphere’ (TOGA).
Interaction between the oceans and the atmosphere occurs through the transfer of heat, mass, and momentum. Transfer can occur in either direction. In the case of heat, both latent heat resulting from evaporation and sensible heat are important, and both depend on factors such as wind speed, the sea-surface temperature (
SST), and the temperature of the overlying air. SSTs may result from ocean processes such as advection of cold or warm water in ocean currents, from the influence of the radiation budget (which largely depends on latitude, season, and cloudiness), or from the action of the atmosphere itself through wind-driven evaporation.
Despite this complexity, regions of climatological interaction between the ocean and atmosphere can be readily identified. The cold ocean currents on the western side of continents, for example, are associated with desert conditions, examples being the Benguela Current and the Namib Desert, the Humboldt Current and the Atacama Desert as well as the western margins of North Africa, North America, and Australia. If seen as an ocean-to-atmosphere process, these cool waters stabilize the overlying atmosphere by cooling the near-surface air. Convective rainfall is thus unlikely. The atmosphere also helps to maintain the cool water since alongshore or offshore winds driven by the resident subtropical anticyclones in these locations drive upwelling regimes in the ocean which ensure a supply of cold, deeper water.
The exchange of heat from the ocean to the atmosphere is important in driving the large convective regions of the tropics, particularly over Indonesia where the SST is some 30 °C. Here the flux of latent and sensible heat from the ocean to the atmosphere is sufficient to generate convection throughout the depth of the atmosphere. This response is possible since the relationship between air temperature and the amount of water vapour the air can hold (and hence the subsequent release of latent heat when clouds are formed) is highly non-linear. In the eastern tropical Pacific, on the other hand, where SSTs may be half those near Indonesia, skies are clear and subsidence in the atmospheric column is dominant.
Differences between land and sea temperatures are vital in driving the monsoon circulations of the globe. Water responds more slowly than land to heating since it allows heat to be mixed more easily, it delivers radiation through a depth of several metres, it is more reflective, and evaporation cools the surface. Seasonal heating and cooling of oceans lags several months behind that of land and is of a smaller amplitude. These differences are revealed in the monsoon circulations. Similarly, during winter the high-latitude oceans are characterized by low-pressure centres, with the Aleutian and Icelandic lows centred in the northern margins of the Pacific and Atlantic Oceans, with a high-pressure cell over Asia. The picture reverses in summer, with low pressures over land.
Much of the interannual variability in the tropical atmospheric circulation and rainfall is related to variations in SST. For some time this connection was overlooked, largely because variations in sea-surface temperature were regarded as being too small to influence the atmosphere, but also because global data sets of historical SSTs had yet to be developed.
The best-known example of air–sea interaction is the
El Niño–Southern Oscillation (
ENSO) phenomenon. The atmospheric component was identified in the 1920s, but it was not until the late 1960s that Bjerknes argued for ENSO as a fully coupled ocean–atmosphere system. It is now regarded as a fairly simple, yet robust system. Nevertheless, the issue of whether the SST anomalies cause the rainfall anomalies or simply reflect changes in the atmospheric circulation has often been raised. In part, this question has been examined in general circulation model (GCM) experiments forced by SST changes. The most comprehensive to date have been undertaken at the UK Meteorological Office, where an atmosphere-only GCM has been forced with observed SSTs from 1903 to 1993. These experiments show that much of the variability of the tropical atmosphere may be explained by variations in tropical SST, particularly those in the tropical Pacific. The mid-latitude atmosphere, on the other hand, appears to be highly chaotic, varying independently of SSTs.
An important application has resulted from the understanding of ocean–atmosphere interaction. Since changes in SST take place slowly, and since the atmosphere in much of the tropics is linked with SSTs, it is possible to predict seasonal rainfall categories for several parts of the tropics. The Sahel, eastern and southern Africa, and north-east Brazil are among the regions in the tropics where seasonal rainfall forecasts with leads of several months are prepared. The Sahel, for example, is sensitive to SSTs in the Pacific and the Atlantic. The recent decade-long drought in the Sahel has been explained in terms of the changing SST structure in the Atlantic, whereby positive temperature anomalies in the southern part of the Atlantic led to dry summers in the Sahel. This change in the Atlantic may reflect changes in the thermohaline circulation, which is the largest mode of oceanic circulation. Tropical cyclone frequency in the tropical Atlantic is linked to ENSO events and Sahel rainfall amounts. Southern Africa rainfall, on the other hand, is sensitive to variations in sea-surface temperatures in the Pacific and Indian Oceans. Grain yield in southern Africa is related to the pattern of temperatures in the eastern tropical Pacific, half way round the globe! On even longer timescales, the cooling of the global oceans during ice ages is thought to have had a profound effect on the global atmospheric circulation. Included in these timescales is the role of the thermohaline circulation mentioned above.
R. Washington
Bibliography
Open University (1989) Ocean circulation. Pergamon Press, Oxford.
