Global Mean Dynamic Topography


ENACT CLS TASKS The RIO-03 Combined Mean Dynamic Topography (CMDT)
The direct method The synthetic method CMDT versus OCCAM, Levitus and Le Grand CMDT RIO-03 References

The direct method

Description

The altimeter provides the sea level height η relative to the reference ellipsoid. It is the sum of the geoid G plus the dynamic topography h (once removed other oceanic and atmospheric effects). By averaging altimetric heights over a given period, a Mean Sea Surface (MSS) can be estimated. The MSS is reference to the Earth ellipsoid. We use the MSS CLS01 η_93-99 that has been computed with altimetric data averaged over the 1993-1999 period. It is the sum of the geoid G plus the mean dynamic topography h_93-99. Thus, by subtracting the geoid G, we obtain can the Mean Dynamic topography h_93-99 this is the so called "direct method":

Results

Unfortunately, geoid accuracy is poor, but can be used at large scale. The EIGEN-2 CHAMP geoid is used here, at spherical harmonics degree 30 (wavelenth shorter than 660 km are filtered out). The cumulative error field show that errors are still larger than 10 cm in the tropics.

Cumulated errors of the EIGEN-2 geoid at degree 30

The corresponding filtering is also applied to the CLS01 MSS, in order to have consistent scales.

The two surfaces are then subtracted, to provide a large scale mean dynamic topography but unfortunately, with the same inaccuracy than the geoid:

MDT from the CLS01 MSS minus EIGEN-2 geoid at degree 30

	The errors of such "direct MDT" are illustrated below with, on the left, the difference between the CLS01 MSS and the EGM96 geoid. (using the direct method with the EGM96 geoid instead of the EIGEN-2 geoid); and on the rigtht, the Levitus climatology MDT. These two surfaces are expanded in spherical harmonics, and here only to degree and order 14, in order to filtered out geoid noise at short and medium wavelength. Differences on these large scales still reach the 15 cm rms.

The Levitus climatology provide realistic ocean circulation patterns where data are avalaible (at low and mid-latitudes), but rather poor quality at high latitude. The TDM is computed using Levitus climatology dynamic height to 1500 dbar, and extended toward the coast:

	The second of step of our method is thus to merge the direct method MDT and the Levitus climatology MDT at degree 30:

The merging is performed by weighting this two surfaces: full contribution of the Levitus climatology at low and mid-latitude, then increasing contribution of the direct method MDT toward the poles. Once the two filtered (at degree 30) surfaces are merged, the high wavelength contribution of the Levitus climatology MDT is added back. We thus obtain our guess based on Levitus climatology at low and mid-latitudes, and where the circulation of the subpolar gyre and the Antarctic Circumpolar Current are enhanced by gravimetric information:

Latest Update: 10/02/2004