Advances in Large-Scale Ocean Dynamics From a
Decade of Satellite Altimetric Measurement of Ocean
MS 300-323, Jet Propulsion Laboratory, California Institute of Technology,
4800 Oak Grove Drive, Pasadena, CA 91109, USA.
Tel: (1) 818-354-8167; fax: (1) 818-393-6720; email: firstname.lastname@example.org
Centre National d’Etudes Spatiales, 18, Ave. E. Belin, Toulouse, 31401, France.
Tel: (33) 5 61 27 48 72; fax: (33) 5 61 28 25 95; email: email@example.com
The past decade has seen the most intensive observations of the global ocean surface topography from satellite altimeters. The Joint U.S./France TOPEX/Poseidon (T/P) Mission has become the longest radar mission ever flown in space, providing the most accurate measurements for the study of ocean dynamics since October, 1992. The European Space Agency’s ERS-1 and –2 Mission also provided altimetric
observations from 1991-2000. The combined data from T/P and ERS have higher spatial resolution and greater coverage than the individual missions. Major advances in large-scale ocean dynamics from these observations are reviewed in the paper, including the ocean general circulation and its variability, the evolution of the El Niño Southern Oscillation cycles as well as the emerging decadal variability, the response of the ocean to wind forcing, assimilation of altimeter data by ocean general circulation models and the estimation of deep ocean circulation, global sea level rise, and tidal models and mixing.
Keywords: satellite altimetry, ocean circulation, TOPEX/Poseidon, El Niño, tides
about 10 cm over a scale of a few hundred 1. INTRODUCTION
kilometers. There are important oceanographic
The utility of a spaceborne radar altimeter for the signals at all these scales, dictating a better
study of ocean circulation and dynamics was first measurement system for making substantial
demonstrated by Seasat (Fu , 1983). Despite the progress in ocean dynamics.
short 3-month duration of the satellite mission, a
wealth of dynamic phenomena in the ocean was The Joint American/French TOPEX/Poseidon
revealed in the observations, including tides, Mission (T/P hereafter) was the first altimetric
eddies, and boundary currents. Following Seasat, mission optimally designed for the study of the
the U.S.Navy launched Geosat in 1985. With dynamics of ocean circulation. The mission
capabilities similar to Seasat, Geosat collected 4 concept was developed in the early 1980s.
years’ worth of data and generated a wide range Realizing the outstanding problem in orbit
of applications in the study of global ocean determination, NASA and CNES initiated a 10-
dynamics (Fu and Cheney, 1995). While the year effort to improve the modeling of the earth’s
results from Seasat and Geosat were tantalizing, gravity field as part of the mission. To make
the limitations caused by the rather large precise tracking of the satellite’s positions, the
measurement errors were disappointing. For satellite carries three tracking systems: laser
example, the uncertainty in determining the retroreflectors, DORIS receivers, and GPS
satellite’s orbits creates an error in ocean surface receivers. The altimeter was the first with dual
topography on the order of 1 meter over a scale of frequencies (13.6 and 5.3 GHz) for removing the
thousands of kilometers. The errors caused by the path delays caused by the ionopsheric free
lack of accurate measurement of the tropospheric electrons. Also carried was a 3-frequency
water vapor content lead to topography errors microwave radiometer (18,21, 37 GHz) for
removing the path delays caused by the water topography. The difference between the height of vapor in the atmosphere. The satellite’s orbit sea surface measured by altimetry and the height configuration was dictated by optimal of the geoid is the ocean dynamic topography performance in orbit determination and tidal from which the surface geostrophic velocity is sampling. The reader is referred to Chelton et al. then determined. The top panel of Figure 2 (2001) for a description of the principle of radar shows the dynamic topography derived from altimetry and to Fu et al. (1994) for the details of Geosat (Nerem et al. 1990), to be compared with the T/P satellite and instruments.
T/P was launched in 1992 and has provided 10
years’ worth of high-quality data so far.
Displayed in Figure 1 is the comparison of the sea
level measurements made by T/P and a tide gauge
at the Christmas Island in the equatorial Pacific
Ocean. Note that the rms differences are close to
2 cm. During the past decade, ESA also
Fig. 1 Monthly sea level anomalies from T/P (black) and a tide gauge (blue) at the Christmas Island (Robert Cheney, personal communication, 2002).
launched ERS-1 and ERS-2 satellites, both of
which carried radar altimeters. Although the ERS
altimeter performance is not as good as T/P, the
data from the two ERS altimeters have provided
spatial coverage to complement the T/P
observations. The unprecedented continuous 10-
year observation of the global ocean topography
has revolutionized the way we study the global
oceans. Significant advances have been made in a
wide range of subjects in ocean dynamics. This
paper presents a summary of the highlights from
this remarkable progress. Fig. 2 Top: surface dynamic topography from the Geosat data (Nerem et al., 1990. Middle: from the T/P data (Stammer et al., 2000). Lower: from a 2. OCEAN GENERAL CIRCULATION numerical model assimilating the T/P data (Stammer et al., 2000) Satellite altimetry provides a unique approach to
the determination of the ocean dynamic
the middle panel showing the result from T/P, and
the lower panel showing the computation from an
ocean model that assimilates the T/P data
(Stammer et al., 2000). The improvement of T/P
over Geosat is quite substantial. However, the
spatial resolution is limited by the errors in the
present geoid models that amount to 10 cm at
1000 km scales. The newly launched GRACE
satellite is expected to provide better geoid
models that will have only 1 cm error at 300 km
scales. Note that the model has provided a
smoother estimate of the topography that has no
voids due to lack of observations. Constrained by
the surface observations, the numerical models
also make estimates of the circulation of the deep Fig. 4 Standard deviation of the ocean surface
ocean, as shown in Figure 3 for a depth of 2000 m topography (in cm) computed from the data from T/P, (Tong Lee, personal communication, 2002). Such ERS-1 and –2 (Ducet et al. 2000) constrained model estimates allow the
determination of the flux of mass, heat, and
dissolved greenhouse gases in the ocean as well finding has motivated the improvement in the as the air-sea fluxes. spatial resolution of numerical ocean models to as
fine as 1/16 degree for resolving the details of
4. EL NINO AND LA NINA
The much improved orbit determination for T/P
has enabled the best global observations of the
large-scale ocean variability. The global view of
the 1997-98 El Niño as it evolved in real time was
the best demonstration of this powerful new
capability. For the first time, altimetry
observations were used by the US NOAA climate Fig. 3 Current velocity vectors at 2000 m prediction models for improving forecast skills. estimated from a numerical model assimilating Shown in Figure 5 are the yearly averaged sea the T/P data. surface anomalies for 1997-2000, depicting the evolution of El Niño into La Nina, followed by the development of a slow change in the Pacific, possibly a new phase of the Pacific Decadal 3. MESOSCALE VARIABILITY Oscillation. What satellite alimetry has done best to date is to 5. GLOBAL MEAN SEA LEVEL RISE illustrate how the global oceans are changing with time. Shown in Figure 4 is a map of the standard After averaging over the entire globe, the errors in deviation of ocean topography computed using the T/P observations are mostly cancelled out due data from T/P, ERS-1 and ERS-2 (Ducet et al., to their random nature. The errors in the 2000). The standard deviation is dominated by computation of the global mean sea level change the mesoscale fluctuation of ocean currents and are generally less than 5 mm over a 10-day span eddies. The map reveals the details of the (Nerem and Mitchum, 2001), allowing a credible mesoscale ocean variability not seen before. Of estimate for the slow change of the global mean particular interest is the structure around the sea level. Figure 6 shows the comparison Grand Banks, resembling the path of the North between the estimates from T/P and in-situ Atlantic Current. This feature was incorrectly temperature observations (Cabanes et al., 2001). mapped from previous observations. The new
1997 1998 -1.5
1994 1996 1998 2000 2002
residual Fig. 5 Yearly averaged surface topography anomalies computed from the T/P data from
Fig. 6 Top: Global mean sea level in cm determined
from T/P (Nerem et al., 2001). Lower: Global mean sea
The 2.7 mm/year rise estimated from T/P is level from T/P (dotted), after being smoothed (solid),
from estimation based on in-situ temperature data largely consistent with the thermal expansion of
(dashed), and the residuals ( dash-dotted, T/P – in situ). the global oceans over a period of eight years.
6. TIDES AND TIDAL MIXING
Ocean topography changes on the order of 1 meter due to the tides. If not corrected for, the tidal signals would overwhelm most other oceanographic signals. Specifically designed for sampling ocean tides with the satellite orbit’s inclination of 66 degrees, T/P observations have led to the development of the best ocean tide models. The accuracy of such models is better than 3 cm in an rms sense globally. The energy fluxes derived from these models revealed an unexpected finding: close to 30 % of the tidal dissipation occurs in the deep ocean (Egbert and Ray, 2001), in contrast to the old notion that most of the tidal energy dissipates over the shallow seas. Such deep ocean dissipation (Figure 7) is via the conversion of external tides into internal tides over ocean bottom topographic features. 2Fig. 7 M dissipation rate (in mW/m) derived from a tide 2 The tidal dissipation provides a powerful source model constructed from the T/P data (Egbert and Ray, 2001). of mixing in the deep ocean and is important to
the understanding of the formation of the Geosat Follow-on (launched in 1998) has also thermocline and ocean general circulation. This been producing useful data along the Geosat new finding has triggered a series of efforts of ground tracks.
including tidal mixing in ocean models.
In the long run, satellite altimetry should be part
of an operational global ocean observing system 7. CONCLUSIONS AND OUTLOOK
run by operational agencies such as the U.S. The decade-long altimetric observations of the NOAA and others. Before this goal becomes a global ocean surface topography have reality, a bridging mission currently called Ocean revolutionized the way we study the ocean. Surface Topography Mission (OSTM) is being Never before has a dynamic variable of the global planned for launch in 2006 to continue the oceans been monitored routinely over such a long precision altimetry data stream towards 2010. period of time. This capability has broken new OSTM is a collaborative effort among NASA, ground for routine estimate of the physical state CNES, NOAA and EUMETSAT. It represents of the ocean. The success of ocean altimetry has the transition of satellite altimetry from a provided a key motivation for the Global Ocean research/development mission to an operational Data Assimilation Experiment. The paradigm of mission that will be conducted indefinitely in the oceanography has shifted from exploration to future. As an experimental payload being quantification, a crucial step towards the ability to considered for OSTM, a wide-swath ocean make climate prediction. The challenge in the altimeter has been designed for measuring ocean future is the sustenance of this important data topography over a swath of 200 km in width at a stream. resolution of 15 km x 15 km, covering nearly
100% of the global ocean surface between +/- 66 The follow-on to T/P, called Jason-1, was degree latitudes every 10 days. This new launched on December 7, 2001. This mission has instrument will be able to resolve the Rossby the same payload and flies in the same orbit as radius of deformation at all latitudes and go a T/P. During the first eight months of the Jason-1 long way towards understanding and monitoring mission, T/P and Jason-1 fly over the same of eddies, fronts, and boundary currents that have ground tracks separated in time by only 60 not been properly sampled by conventional seconds. Comparisons of the two measurements altimeters.
taken under nearly identical sea conditions are
extremely useful in the calibration and validation ACKNOWLEDGEMENT
of Jason-1 in an effort to make its measurements
consistent with those of T/P for building a long The research presented in the paper (LLF) was time series record from the two satellites. After carried out by the Jet Propulsion Laboratory, the completion of this Calibration Phase, T/P is California Institute of Technology, under contract planned to be moved to a new orbit that will with the National Aeronautic and Space produce ground tracks interleaving those of Administration, and (YM) by Centre National Jason-1 (i.e., the old T/P tracks). In this Tandem d’Etudes Spatiales. Support from the Mission, the two satellites will be able to sample TOPEX/POSEIDON Project and Jason-1 is the global ocean with twice the spatial resolution acknowledged.
and improve the knowledge of the global
mesoscale variability as well as coastal tides.
ESA launched ENVISAT in March, 2002. Its
payload includes a dual-frequency altimeter and a Cabanes, C., A. Cazenave, and C. Le Provost, dual-frequency microwave radiometer for water-2001: Sea level rise during past 40 years vapor corrections. Flown in the same sun-determined from satellite and in situ observations. synchronous orbit (in which the solar tides are Science, 294, 840-842.
aliased to the mean topography) as ERS-1 and
ERS-2, ENVISAT altimeter data will Chelton, D.B., J. Ries, B. Haines, L.-L. Fu, and P. complement Jason-1 in a similar manner as ERS-Callahan, 2001: Satellite altimetry. In “Satellite 1 and 2 did to T/P. Additionally, the U.S. Navy’s Altimetry and Earth Sciences: A Handbook for
Techniques and Applications” edited by L.-L. Fu
and A. Cazenave, pp. 1-131, Academic Press, San Fu, L.-L., E. J. Christensen, C.A. Yamarone, M. Diego, 423 pp. Lefebvre, Y. Menard, M. Dorrer, and P. Escudier,
1994: TOPEX/POSEIDON Mission Overview. Ducet, N., P.Y. Le Traon, and G. Reverdin, 2000: Journal of Geophysical Research, 99, 24369-
Global high-resolution mapping of ocean 24381.
circulation from TOPEX/Poseidon and ERS-1
and –2. J. Geophys. Res., 105, 19477-19498. Nerem, R.S., and G.T. Mitchum, 2001: Sea level
change, In “Satellite Altimetry and Earth Sciences: Egbert, G.D., and R.D. Ray, 2001: Estimates of A Handbook for Techniques and Applications” M2 tidal energy dissipation from edited by L.-L. Fu and A. Cazenave, pp. 329-349, TOPEX/Poseidon altimeter data. J. Geophys. Res., Academic Press, San Diego, 423 pp. 106, 22475-22502.
Nerem, R.S., B.D. Tapley and C.-K. Shum, 1990: Fu, L.-L., 1983: Recent progress in the application Determination of the Ocean Circulation using of satellite altimetry to observing the mesoscale Geosat Altimetry. J. Geophys. Res., 95, 3163-
variability and general circulation of the oceans. 3180.
Reviews of Geophysics and Space Physics, 21, 1657-
1666. Stammer, D, et al., 2000: The global ocean state
during 1992-1997, estimated from ocean Fu, L.-L., and R. E. Cheney, 1995: Applications observations and a general circulation model, part of satellite altimetry to ocean circulation I: methodology and estimated state, The ECCO studies:1987-1994. Review of Geophysics, 32, Report #4, 53 pp. Scripps Institution of Supplement,. 213-223. Oceanography, La jolla, CA, USA.