The ocean plays a key role in determining the global climate and its time evolution. To understand this role and subsequently develop techniques for predicting future climate, one must understand the dynamics of the global ocean circulation---the movement of water that transports mass, heat, salt, and other biogeochemical properties of the ocean that are closely linked to the processes of climate change. The only viable approach to observing the global ocean circulation with sufficient resolution and consistent sampling is the use of a satellite radar altimeter to measure the height of the sea surface---the sea level [ Wunsch and Gaposchkin, 1980; Stewart, 1983; Wunsch, 1992]. After removing the effects of the tides and atmospheric pressure from the observation, the deviation of the sea surface from the geoid, called the ocean dynamic topography, is readily related to the velocity of the surface geostrophic flow---a component of the surface flow on which the surface pressure force is balanced by the Coriolis force due to the Earth's rotation. Moreover, the ocean dynamic topography provides a strong constraint for determining the ocean circulation through the entire water column via the dynamic equations governing the fluid motion. Precise measurement of the shape of the global sea surface thus provides a powerful tool for studying the dynamics of the ocean circulation.
Progress in the application of satellite altimetry for ocean circulation studies from 1987--1994 is reviewed in this paper. The reader is referred to Douglas et al. [1987], Brown and Cheney [1983], and Fu [1983] for early development in the subject. The last review of the subject was conducted eight years ago [ Douglas et al., 1987]. The past eight years have seen the greatest strides in the development of satellite altimetry as an observational tool in oceanography. The U.S. Navy launched Geosat in 1985 with a primary objective of mapping the marine gravity field for military applications. Most of the data collected during Geosat's first 18-month primary mission (the Geodetic Mission) are still classified. However, the declassified portion of the data has led to significant advancement in marine geophysics [ e.g., McAdoo and Marks, 1992]. When its primary mission was achieved, Geosat was maneuvered into a 17-day repeat orbit, and an extended mission (the Exact Repeat Mission) for oceanographic applications started in October 1986 [ Born et al., 1987; Douglas and Cheney, 1990]. Although the mission was not designed specifically for oceanographic applications, the 2.5-year data set created the first opportunity for oceanographers to experiment with a multi-year global data set (see the March and October 1990 issues of the it Journal of Geophysical Research devoted to Geosat.)
The most serious Geosat error source has been the uncertainty in the radial orbit height, but significant progress continues to be made in this area. The Geosat orbit error was on the order of 2 m in the initial data release, but was later reduced to about 30--50 cm in a later release largely through gravity model improvements [ Cheney et al., 1991; Haines et al., 1994]. However, the orbit accuracy rapidly degrades to the 1-m level after mid 1988 through the end of the mission due to inadequate modeling of the drag force caused by the increased solar activities. By using the altimeter data as a constraint, Shum et al. [1990 a] achieved an orbit precision of about 20 cm. Finally, in 1993, the U.S. Navy released additional Geosat Doppler tracking data which have led to a further reduction in the orbit error to about 10 cm [ S. Nerem and R. Williamson, private communication]. Other important error sources include the altimeter signal delay caused by tropospheric water vapor and ionospheric free electrons [ Emery et al., 1990; Musman et al, 1990]. Based on the best orbit available, the present Geosat error budget is at the level of about 15 cm, an impressive accomplishment in view of the initial low expectations for the mission.
In 1991, two years after the end of the Geosat mission, the European Space Agency launched the ERS-1 (European Remote Sensing Satellite-1) altimeter. An identical satellite, ERS-2, will take its place in 1995. These two missions provide the type of continuous, long-term coverage needed for studies of interannual ocean variability and also provide altimeter data to operational users in near-real time [ Cheney and Lillibridge, 1992]. Because the ERS series of satellites carry a suite of sensors, a number of different ground track patterns have been followed, including 3-day and 35-day repeats. During most of its final year of operation, the ERS-1 satellite was flown in an orbit that yielded a dense global network of altimeter profiles spaced 15-km apart at the equator for geodetic studies. As was the case for Geosat, orbit error is the primary obstacle to be overcome for ERS-1 altimeter applications. The problem was compounded by early failure of one of the two onboard tracking systems. However, laser tracking together with advanced gravity models have now resulted in orbits with an accuracy of about 15 cm [ Lillibridge et al., 1993].
To be useful for studying the ocean circulation, especially at the basin scales, the accuracy of altimetry measurement must be less than 15 cm [ TOPEX Science Working Group, 1981]. To achieve this goal, a satellite mission with specially designed instrumentation and orbit configuration, and much improved knowledge in the Earth's gravity field for orbit determination was required. The TOPEX/POSEIDON Mission was the result of this conviction by scientists and engineers in both the U.S. and France [ Fu et al., 1991 a; The TOPEX/POSEIDON Science Working Team, 1991; Fu et al., 1994]]. TOPEX is an acronym standing for Ocean Topography Experiment, the name originally used by the U.S.; POSEIDON is the original French name for the mission. This joint U.S./France mission was launched in August, 1992 with an expected lifetime of three to five years. The performance of the mission has exceeded its specification. Orbit and altimeter accuracies are estimated to be 3.5 cm and 3.2 cm, respectively, resulting in an absolute accuracy of 4.7 cm for the determination of the geocentric sea level [ Fu et al., 1994]. For the first time, oceanographers have a global observing system that is providing data of sufficient accuracy and sampling for the study of the global ocean circulation. Preliminary results from TOPEX/POSEIDON can be found in the December 1994 issue of the Journal of Geophysical Research. Significant advancement in the knowledge of the large-scale ocean circulation is expected from the mission in the near future.
A brief summary of the accomplishments made in the past eight years is given below in categories based on oceanographic phenomenology. Only those papers addressing direct applications to the ocean circulation are reviewed. A large body of literature on techniques for altimetry correction and data analysis has thus been left out (see Chelton, [1988] for an overview). Of great importance to the utility of altimeter data is the correction for the tidal effects on the sea level observation. The reader is referred to Cartwright and Ray [1990] for estimating tides from altimeter data.