The Development of the Combination Model Constituent of EGM96

This section summarizes the development of the combination model to degree and order 70 of EGM96. After the satellite only solution has been calibrated, the normal equations for surface gravity and altimetry are added to the satellite only solution (EGM96S).

• TOPEX : Include TOPEX data from cycles 11 to 84 (1993 & 1994) (No Poseidon data included in the solution).
• GEOSAT: Include data from Nov. 86 through January 87.
• ERS-1 : Include data from the 35 day repeat phase of the mission, specifically data from cycles 6, 8, 11, 14, and 17.
  
• Geometric tide corrections from an updated version of the tide model from Schrama and Ray [1994].
• Altimeter data weighting based on surface variability (ie. data over regions with high variability has a higher sigma in the solution).
• Improved high degree geoid corrections were used. EGM96 used the high degree geoid corrections derived from the EGM X04, interim test model, a quadrature solution, to degree 460 - instead of using JGM-3/OSU91a to degree 360.
  
• For TOPEX & ERS-1, we solve for a single static sea surface topography solution (QSST) to degree and order 20, since the data we have selected are temporally coincident. We also solve for a 10x10 time varying SST using annual & semiannual terms.
• The weight of ERS-1 in the solution is selected such that the QSST derived from TOPEX alone does not change by more than 2 cm over -66 to + 66 deg latitude.
• Altimeter data ionosphere correction for GEOSAT from IRI95.

• The surface gravity received full weight in EGM96. This is contrast to JGM-2, where the surface gravity normal equations were downweighted by a factor of four. This was previously found to be necessary because of the degradation of the orbit fits to various satellite test arcs when the surface gravity were added to the solution [Nerem et al., 1994; Lerch et al., 1992; Lerch et al., 1990] The high quality and global coverage of the surface gravity data used in EGM96 made such a downweighting unnecessary.
• A separate set of 5x5 harmonics was adjusted for the surface gravity, as in JGM-2, in order to accomodate any long-wavelength reference frame discords in the surface gravity data.

back to table of contents.
back to egm96 home page.