Date: Thu, 21 Nov 2002 09:56:18 +0100
From: Jean-Paul Berthias
To: Pascal Willis
Cc: Nikita Zelensky , ries@csr.utexas.edu,
bjh@cobra.jpl.nasa.gov, Yoaz.Bar-Sever@jpl.nasa.gov,
sluthcke@xyz.gsfc.nasa.gov, krchoi@csr.utexas.edu, feissel@ensg.ign.fr,
Gilles Tavernier
Subject: Re: JASON DORIS phase center and SAA effect
Hello,
What Pascal and Nikita have seen when adjusting the Z offset of the DORIS
antenna phase center is perfectly normal: this is again our "old friend", the
frequency drift. Actually, if my computations are correct, a Z offset in the
on-board antenna phase center has exactly the same Doppler signature as an
altitude biais on the station height, with a scale ratio between the two eqal to
(R/a)^3 * (1-a/R*cos(alpha))/(1-R/a*cos(alpha)) where R is the Earth radius, a
the orbit semi-major axis and alpha the angle between the radius vector to the
station and the orbit plane (alpha = 0 for overhead passes).
What this means is that solving for a Z offset in the antenna height using
Doppler data is identical to solving for a constant altitude offset on all the
stations, which amounts pretty much to a scale factor on the global set of
coordinates. As we know that there are large pass dependant frequency drifts on
Jason, the estimated offset will be the average value of these drifts on the set
of xstations included in the solution. Thus, when only stations outside of the
SAA are included, the observed offset corresponds to the average drift outside
of the SAA. When all the stations are included, the offset corresponds to the
global average over the whole network.
The drift which is removed from the delivery data is obtained by adjusting a
linear frequency model over 10 days (or more precisely a quadratic time model
adjusted with pseudo-range data only from master beacons, equivalent to one
measurement every few hours). This model is a global average which includes both
the slow drift outside of the SAA and the rapid drift over the SAA. The
difference between the actual "local" drift and this long term average is still
present in the data. This difference induces the perturbations that you see in
station altitude, antenna Z offset and scale parameter. The further away you are
from a global configuration averaged over 10 days, the more discrepencies you
will observe.
My recommendation is to leave the DORIS phase center where it is as we have no
way of sorting out the different effects. On the other hand, Pascal's results
for the other satellites have to be carefully analyzed, in particular for the
Spot satellites where the offsets seem to consistently point in the same
direction.
Best regards,
Jean-Paul