water of the climate system while the atmosphere dominates the rapid response of the climate system and more directly impacts human activity.
The ocean’s direct impact on the atmosphere is primarily through sea-surface temperature and ice-cover. Thus, it is fortunate that temperature records are among the longest oceanic time-series and have the best spatial coverage. Data sets include sea-surface temperatures from ocean vessels, long coastal sea-level and temperature records, and shorter or more scattered time-series of temperature and salinity from surface to sea-floor. Increasingly long time-series of directly measured ocean currents are becoming available, particularly in the tropics. The TAO (Tropical Atmosphere-Ocean) array in the Pacific, sometimes called the world’s largest scientific instrument, measures equatorial temperatures, winds and currents around one-quarter of the earth’s circumference (e.g., McPhaden et al., 1998). The array has given us detailed portraits of El Niño-Southern Oscillation (ENSO) cycles and equatorial general circulation.
Over longer times other aspects of oceanic circulation, chemistry and biology become important to climate. For example, the heat storage available to the atmosphere is strongly dependent on circulation and salinity stratification of the upper ocean. The depths of the ocean become involved as the thermohaline circulation (THC) and wind-driven circulation interact to reset surface conditions. There are “overturning circulations” at many scales, from the global THC (see Plate 4) to the shallow, near-surface cells of overturning lying parallel to the equator. Direct measurements of the circulation of the deep ocean are still sparse, and indirect means are often used to infer the circulation. Water density (from measured temperature and salinity) can be combined with dynamical constraints and atmospheric observations of air-sea interaction to estimate global ocean circulation (e.g., Ganachaud and Wunsch, 2000; Reid, 1994, 1998, 2001). The results are consistent with the limited direct measurements of currents, and also with the patterns of observed chemical tracers in the ocean. The tracers include natural dissolved gases and nutrients, dynamical quantities such as potential vorticity and potential density, and chemical inputs from human activity. Transient chemical tracers, injected into the atmosphere and subsequently absorbed by the ocean, provide particularly useful images of the ocean circulation. Bomb radiocarbon, tritium and chlorofluorocarbons (CFCs), for example, allow verification and quantitative assessment of the pathways of high-latitude sinking, equatorward flow in boundary currents, and interaction with the slower flow of mid-ocean regions (Broecker and Peng, 1982; Doney and Jenkins, 1994; Smethie and Fine, 2001).