Date: Sat, 12 Oct 2002 19:18:25 +0200
From: Escudier Philippe <Philippe.Escudier@cnes.fr>
To: Pascal Willis <Pascal.R.Willis@jpl.nasa.gov>,
    Berthias Jean-Paul <Jean-Paul.Berthias@cnes.fr>,
    "John C. Ries" <ries@csr.utexas.edu>
Cc: Gilles Tavernier <Gilles.Tavernier@cnes.fr>,
    Martine Feissel <feissel@ensg.ign.fr>,
    "Jayles, Christian" <Christian.Jayles@cnes.fr>,
    "Sengenes, Pierre" <Pierre.Sengenes@cnes.fr>
Subject: RE: satellite clock in the SAA region

Hi John,

I would like to come back on two points that you mentionned :
- The effect is clearly non linear and non quadratic when the satellite 
enters the SAA. At this time due to the radiation effect there is 
a rapid change in the frequency drift so that the slope goes from 
-2E-10 / day outside the SAA up to few +10-9/day in the SAA. 
This explains the results you got, which were very useful for us 
to get a better understanding of the physical nature of the effect :
 * adjusting a slope is somehow efficient for beacons located in 
the middle of the SAA (it helps recovering this important drift 
which otherwise go in the tropo coefficient and / or station 
altitude and orbit solution and /or residuals). 
 * adjusting a slope is not efficient for beacons located at the 
edge of SAA because there, the main effect is a sudden slope 
change during the pass, which is an effect going in the station 
latitude and/or orbit solution and/or residuals.
Even for station in the middle of SAA as the exact shape is not 
a linear drift this slope cannot recover the complete effect, 
that is why we consider that beside the workaround that you have 
proposed and which proves to be very efficient for POD 
(underweighting those beacons), the only way to efficiently 
compensate from this effect for those beacons will be to use 
a radiation model (which may be very simple, just a function 
of latitude and longitude when computed for Jason) that will 
have to be tuned with the data in a similar way you are doing 
for atmosphere model for instance.

- It seems that there has been some misunderstanding with 
this "saturation effect". This saturation should reduce 
significantly the effect and not enhance it (it seems that 
T/P USO evidenced an effect almost as large as Jason during 
the first months of the mission, we would be interested to 
know if you have some evidence of that).
At the present time we think that the explanation of this 
effect increase since January is due to the fact that during 
the last months we have been leaving the solar activity 
maximum which has induced a lowering of the SAA which has 
created a more severe radiation environment (this can be 
seen on other equipment of the satellite such as the star tracker).

Thanks a lot for the effort you have provided to identify, 
understand and compensate for this effect.

Philippe