Project description in Danish - projektbeskrivelse paa dansk
Project description in English.


C.C.Tscherning is the manager of the project.
On Oct. 2001 Eva Howe started as a Pd.D.-student working on the project.
The results from her work can be seen on her home-page.

On July 1 2002, Lars Stenseng stared work as a research associate, with a contract for 6 months. Since spring 2003 he works at KMS and since March 2005 at the Danish Spacecenter.
In the project participates scientists from the Danish Spacecenter (Rene Forsberg) and from AUT (D.Arabelos).
Eva Howe finished her PhD in September, 2005, and was then employed as a post-doc for 5 months, finishing January 2006.

The main activity in the project is the analysis of CHAMP data, which have been obtained following the acceptence of an AO proposal in 2001. and the preparation of the level 2 data-processing for GOCE.



A number of dedicated programs have been developed for the analysis of the CHAMP data:
stat2pot.f for the conversion of the state-vector to height anomalies at satellite altitude.
readacc.f for the reading of the CHAMP acceleration files.
correl.f for the calculation of biases, scale-factors and tilts of the accelerometer data.
ftrans.f for the fourier-analysis and low-pass filtering of the residual data (output from correl.f).
Up to Sept. 1 1 month of CHAMP data has been processed, i.e. corrected for friction, sun, moon etc. This is data presented as height anomalies at satellite altitude.
The data was converted to anomalous potential values, and the contribution of EGM96 to degree 24 was subtracted. The gravity disturbance was calculated and the data were up-ward continued to a common altitude of 440 km.
The data were gridded using geogrid to grids of 1.5 and 1.0 degr. equiangular grids. This was used as input to the program sphgrid , and used for the computation of spherical harmonic coefficients. Results are available from Lars Stenseng

The first results were presented at a meeting of the Danish Students in Geophysics, Nov., 1, 2002. The poster includes an plot of the differences in terms of geoid heights downward continued to sea-level between our solution and EGM96. The standard deviation of the differences is 0.8 m, corresponding very well to the differences of 0,08 m seen at satellite height.

In total 3 solutions were made, using griddings witj 0.75 deg., 1.0 deg and 1.5 deg. spacing.The smoothing done by LSC influences the degree-variances, see the image of the degree-variances.
The deffifferences per degree in terms of gravity anomalies is shown here.


The work is done withing the framework of the EGG-C.
The GOCE test data set produced by IAG SC7 has been analyzed. An agreement was found at the 0.001 EU level. A test conducted in a 2 deg x 2 deg. area at the Equator showed that it is possible to predict Tzz without loss of precision from data in the satellite frame. Use of correlated noise improved the results slightly.
However, when a more realistic noise with standard deviation of 55 mEU was used, the results became much worse, see Draft paper on gridding using GEOCOL, of 24 jan. 2003 and a final paper published in the proceedings IAG Sapporo, 2003

Some key GOCE products are mean values of free-air gravity and of height anomalies (geoid heights). A paper has been written: Proposal for the precise definition of mean values of gravity field quantities, Jan 2003 .

One of the problems with gravity data analysis is the aliasing. This has been investigated using Fast Spherical Collocation, in a report "A fast study of the aliasing of spherical harmonic coefficients, 2003-02-20.
Final version: A study of the aliasing of spherical harmonic coefficients. Festschrift zum 70. Geburtstag Helmut Moritz, pp. 201-205, Graz, 2003.
A more specific study of the aliasing of the coefficients using GOCE Tzz data is in progress. A detailled plan for the collaboration with D.Arabelos has been made. Revised detailled instructions are available.
A joint paper was presented at IUGG/IAG, 2003. revised version, Dec., 2003
The implementation of the FSC method has been completed, see the FORTRAN program sphgric.f and a number of tests have been completed, using:
(1) "horizontal" derivatives , and using
(2) mixed radial and horizontal derivatives
The results were merged into one paper presented at the 2. GOCE Users Workshop, 2004. Tscherning, C.C.: Simulation results from combination of GOCE SGG and SST data. 2. Int. GOCE user Workshop, ESRIN, March, 2004.
Furthermore a study was completed using T and Trr data,
Migliaccio, F., M. Regguzzoni, F. Sanso and C.C.Tscherning: An enhanced space-wise simulation for GOCE data reduction. Paper prepared for 2. Int. GOCE user Workshop, ESRIN, March, 2004
In April 2004, the implementation of the calculation of error covariances in sphgric.f was completed, see