The general circulation is observed using a variety of techniques. The most traditional, dating back centuries, are based on surface drift and water column temperature. Other than at the sea surface, direct current measurements have generally been too sparse for inferring large scale patterns of circulation. The geostrophic method, based on an assumption of balance between the Coriolis force (proportional to horizontal velocity) and horizontal pressure gradient, is the most widely used method of estimating flow. In satellite altimetry (surface height measurement), this concept is applied directly at the sea surface. Within the ocean, the pressure of a surface relative to the geoid cannot be measured accurately enough to infer geostrophic velocity so the pressure distribution is calculated using vertical profiles of seawater density, which depends on temperature, salinity and pressure, and an assumed balance between the vertical pressure and the overlying seawater mass. Thus vertical profiles of temperature and salinity as a function of pressure are required.
Temperature profiles have been measured using thermometers for centuries, and in useful numbers since the early 20th century. Useful salinity observations in the 20th century with high precision salinity observations using the conductivity method became available in the 1950's. Mechanical and expendable bathythermographs have provided upper ocean temperature coverage since the 1960's. Much more precise electronic conductivity (salinity)-temperature-depth instruments (CTD's and predecessor STD's) have been used for detailed vertical resolution of temperature and salinity since the late 1960's. Levitus [1992] documents the number of hydrographic station (temperature and salinity observations at discrete depths using thermometers and water samples), CTD and bathythermograph profiles available from the National Oceanographic Data Center for each year since 1900. Stations were collected in large numbers starting in the 1930's, with massive programs in local coastal regions of many countries; in the Pacific these include around Japan, off California, western Canada, and southern Australia. These comprise the longest available time series. The maximum number of stations was collected in the 1960's, and seem to have been replaced in numbers to some extent by bathythermographs and CTD stations. The World Ocean Circulation Experiment has collected 2800 high quality CTD/oceanographic stations to the ocean bottom in the Pacific in the period 1991-1994, with approximately 500 more stations to be collected.
Expendable bathythermograph (XBT) profiles are collected on merchant ships, and have been used to great effect for inventories of upper ocean heat content and evolution in the Tropical Ocean Global Atmosphere (TOGA) project in the Pacific, and in the mid-latitude North Pacific as the Transpac program. In WOCE, XBT's are being employed as inexpensive oceanographic station replacements, being used for time series in a mode of small station separation to provide mesoscale-resolving sections, although the general lack of concomitant salinity measurements is a serious deficiency. The WOCE high resolution XBT program started in 1986 and has now been expanded to cross all the principal basins in the Pacific Ocean. Temperature time series from moorings are now being collected at many sites in the tropical Pacific as part of TOGA (TOGA TAO array), and being used as input to operational models. Surface temperatures from surface drifters are being used increasingly in these models as well, from locations all over the world.
Velocity is measured directly with moored current meters, surface drifters, subsurface floats, acoustic doppler current profilers (ADCP), and electromagnetic measurements. In the 1980's current meters were deployed across the central North Pacific's subtropical gyre; the data have been used to look at the response of the ocean to winds, transport of the Kuroshio and its recirculations, and the distribution of eddy energy and statistics [ Schmitz et al., 1982; Schmitz, 1988; Niiler et al., 1985]. Current meters are now deployed on some of the TOGA/TAO moorings, relaying information in real time. WOCE current meter deployments are focused on transports through confined regions such as western boundary currents and throughflows; in the Pacific these include the Samoan Passage, and the western and deep western boundary currents in the mid-latitude South and North Pacific.
Surface drifters designed for maximum ability to follow the current have been deployed in great quantities in the world's oceans as part of TOGA and WOCE, drogued at 15 meters depth. A significant number of drifters with deeper drogues have been deployed in the South Atlantic, and northern North Atlantic and North Pacific. The drifters include surface temperature measurements as mentioned above, and a fraction, deployed in regions of minimal ship traffic, also include atmospheric pressure sensors.
Subsurface float experiments were conducted primarily in the Gulf Stream region in the North Atlantic prior to recent years. Until recently, all floats were acoustically tracked. In the Pacific, large deployments have been made in the central South Pacific [ Rossby, 1991; Hautala and Riser, 1993] and recently in the western North Pacific [ Riser, 1994]. With the development of a pop-up float which does not require acoustic tracking, subsurface Lagrangian methods have spread to other regions of the Pacific, and most of the mid-latitude and tropical Pacific Ocean is now being sampled as part of WOCE. Experiments in the Southern Ocean have also been initiated. These float experiments taken together provide an extraordinary new picture of mid-depth flow.
Satellite measurements are beginning to revolutionize our view of the general circulation. With the current altimetry mission, called TOPEX/POSEIDON, a global view of the sea surface topography with impressively low measurement error is now possible, and the next few years should see significant interaction between the communities studying in situ and satellite measurements of the circulation. A goal will be to better quantify the scales on which the ocean circulation varies in time and space.