Files, results and documents related to GOCE HPF WP7000.

GOCE HPF WP7000 Contribution from WP7220, 7230, 7240, 7310, 7420 and 7430.

Development phase 1. Space-wise software tools development: WP7220, 7230 and 7240.

Staff: C.C.Tscherning , Ewa Hove and M.Veicherts .

WP7220: Energy conservation.

Task: Computation of T along the orbit using energy conservation.
The program stat2pot has been modified as to handle End-to-end simulated data, which in contrast to CHAMP state-vector data are given in the inertial frame. However a number of tests have been performed, comparing the derived potential with the normal gravity potential of the earth (without the centrifugal term). See Fig 1. Fig 2. Fig 3. Fig 4.
The variations are far too large, but can be explained using the simulated common-mode accelerations.

Work with new test dataset. March 2005.

1 month of state-vector data has been used to calculate values of height anomalies at satellite altitude, using only that the friction was linear with time. The result can be seen in an output file from geocol, where values of EGM96 to degree 50 have been subtracted. The differences are shown for the first 100000 s and in for 1500000 s.
The differences have periods corresponding to half a revolution, and a probably related to the X residual accelerations after the drag compensation and to the projection of the Z and Y accelerations on the velocity vector direction. There are however also some strange, as yet unexplained, daily jumps.
When calibrated (common-mode) accelerations are received, the calculations will be repeated with corrections for the non-inertial accelerations.
Contribution to HPF ADD documents.
Partly tested software for energy conservation (stat2pot).

WP7230: Gridding.

Task 1: Computation of anomalous quantities.
The computation is done using the FORTRAN program GEOCOL. . Extensitive tests are reported in a paper published in the Sapporo proceedings. The IAG SC 7 data-set has been used for the tests. The tests have beeen repeated with the noise-free end-to-end simulator data with satisfactory results.
As the basis for the testing three data-sets have been created containing the differences between EGM96 to deg. 360 and deg. 24 of Tnn and Trr, /cct/goce/testm/emg96_24_360_trr.dat and egm96_24_360_tnn.dat, 2005-04-03 and /cct/goce/testm/egm96_24_360_tee.dat.
Task 2: Compute gridded values.
Reports on initial tests are found in the paper quoted above, and in a paper with D.Arabelos. Missing are the computation and use of the second order derivative with respect to the longitude.

A plan for further testing has been written.

A program to prepare job-files for gridding, createjg has been modified in order to create jobs for gridding different quantities.
T and Trr and Txx. .
A manual for setting up gridding jobs has been prepared.
Various service programs have been writen to support the gridding. One program can be used to create a global file of coordinates of gid-points.

Contribution to HPF ADD documents.
Tested software for gridding (geocol), see above.

WP7240: Harmonic coefficients estimate by FSC.

Task 1: Global covariance function parameters.
Such preliminary parameters have been determined using the programs EMPCOV and COVFIT. .
Task 2: Use regionally gridded data as input to sphgric.
Tests doing this have been reported in a paper with D.Arabelos.
Task 3: Determine corrections to spherical harmonic coefficients and error estimates.
This has also been tested in the above mentioned paper with data from IAG SC 7. Further tests with a more realistic noise have been reported in a paper with the team from POLIMI , published in the proceedings of GOCE04.

It seems that the error-estimates of the low-degree coefficients are up to a factor 10 too low. It seems to be dependent on the a-priori error-estimate of the gridded data.

The computation of error-covariances have been implemented in the program sphgric .

The tests have been repeated with data from the end-to-end simulator.
A draft manual for preparing runs with sphgric has been prepared. In the preparation a program genjob.f is used.

Task 4: Perform integration procedure, applied to gridded data.
Task 5: Test collocation solution by comparison with integration solution.
Reports on both tasks can be found in the above mentioned paper from GOCE04, and in papers from the same workshop by POLIMI.
The tests have been repeated with data from the end-to-end simulator. Preliminary results have been presented at an internal progress meeting at POLIMI in June, 2005. At the same occation it was agreed to use a standard set of file names. The document was revised at a meeting in Como on Dec. 22, 2005.
Contribution to HPF ADD.
Preliminary tested software, EMPCOV, COVFIT and SPHGRIC.

Following phases.

Not yet commenced.

Software used in WP7000


geocol17 , general program for gridding and creation of reference grids. A record of updates and planned changes is found here
Input examles: ex1.inp , ex2.inp , ex3.inp and ex4_17b.inp
The 3 first examples does not need extra input, but ex4 does:

Grid-file definitions.

cgengri , creates grid-files, see note


createjg , creates input files to geocol for gridding, see manual Last update 2006-01-05.
Input example , Output to the screen , Geocol-jobs for gridding T
0.5 deg. grid file , 0.5 deg. grid with Equator , 0.72 deg. grid
T -EIGENcg03c values 5 sec spacing , Data base with grid-cell variances of Trr-EIGENcg03c , EGM96-EIGENcg03 degree-variances from 2 to 360 in mgal**2

Coefficient estimation using FSC.

genjob , creates part of job-file for sphgric, see manual
Input and output from job.
"True" data used to compute error rms for each parallel.
job-file and out-put file

Subtraction of coefficient sets.

coefsu , and job-file
output and differences .


The program converts state-vector data and common-mode accelerations with attitudes to values of the anomalous potential. Stat2pot as pdf-file.
Stat2pot job-file
Stat2pot out-put .

Service progrms.

pickn.f picks n'th record of gravsoft file.

Last update 2006-01-09 by cct.