[RADS] Reference frame offset made permanent
remko at altimetrics.com
Thu May 1 17:52:02 CEST 2008
Dear RADS users,
** FIELD 3801: Reference frame offset made permanent **
In March I implemented the reference frame offsets for each of the
satellites. I attach my previous e-mail below, to remind you of it.
Making it "permanent" means that (on popular request) the reference
frame offset is now automatically removed when computing the sea level
anomaly (field 0). This is accomplished in the system-wide getraw.nml
file, so (except for downloading it or doing a CVS update) no
additional downloads are needed.
If you REALLY don't want to apply the correction, or because you like
your biases between the missions, feel free to complain or simply add
to your getraw.nml file. Yes ... it's that simple.
Have a good May Day!
From: remko at altimetrics.com
Subject: RADS: Reference frame offsets
Date: 5 March 2008 14:29:32 EST
** FIELD 3801: Reference Frame Offset **
While you were all happily downloading the new data, I slipped in a
new (static) field for each altimeter. It is intended to model the
reference frame differences between the various altimeters, which
includes differences between the orbits as well as some other
geographical differences in the altimeter-dependent models (like SSB).
These models are based on 5 spherical harmonic coefficients: C00 (a
constant bias), C10, C11, S11 (shifts in Z, X and Y direction), and
C20 (a difference in flattening). Because it is eventually arbitrary
which offset to trust as reference, I chose TOPEX. So all these
coefficients are 0 for TOPEX and the others are relative to that.
The reference frame correction can be subtracted from SSH. This is not
done by default, but you can establish it by adding the following
lines to your getraw.nml file:
This basically says: please use field 3801 in the editing of SSH and
Below are illustrations of the reference frame offsets. For example:
ERS-1 and ERS-2 both give too high SSH over Europe compared to TOPEX,
and both have the same signature, except for an offset between the two.
Envisat is merely offset from TOPEX. And Jason-1 also has a small C20
The Geosat correction field is extremely preliminary, since there is
no coincident altimeter data to separate the natural variations in
these terms from reference frame errors.
If you already have some mean sea surface computed, you may like to
remove the offsets from your grids. Thus I have appended the reference
frame grids (in netCDF format) for each satellite. Those are the grids
that I used to produce the plot below. These grids will follow by
If there is consensus that this is a correction that should be applied
by default, I can simply incorporate the two lines above in the system-
wide getraw.nml file.
Let me know you thoughts about this new correction that is not
available in regular GDRs.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 266683 bytes
Desc: not available
Url : http://www.deos.tudelft.nl/pipermail/rads/attachments/20080501/0aec0647/attachment-0001.pdf
More information about the RADS