Date: Fri, 15 Nov 2002 19:00:44 +0100 From: Jean-Paul Berthias To: Pascal Willis Cc: "John C. Ries" , Bruce J Haines , Yoaz Bar-Sever , nzelensk@geodesy.gsfc.nasa.gov, sluthcke@xyz.gsfc.nasa.gov, krchoi@csr.utexas.edu, Martine Feissel Subject: Re: DORIS phase center determination/ all satellites Hello Pascal, I am really puzzled by your results. We have not tried to solve for the DORIS phase center so I cannot compare with your results. I am surprised that you can get even a 2 cm error bar on the offset, I would have assumed that it was not as well determined. By the way, how did you compute the sigma on the mean ? Looking in a similar direction, we have noticed something strange in the mean radial offset between our daily GPS and daily DORIS orbits. The mean is clearly zero when there is no eclipse, and in period of eclipse, the mean goes down linearly to reach almost 2 cm until the time of the flip, then goes back up in a symmetrical way. Our interpretation was that this is the result of a correlation between our empirical parameters which degrades observability, but we did not have the time to investigate it further. Do you see something similar if you compare your DORIS orbits with the GPS orbits ? As far as the impact of a long term frequency drift is concerned, I do not see how it would get in the Doppler data without getting in the time-tag, as the same polynomila is used for both. In the old format a frequency drift evaluated at the center of the pass was removed from the data, while in the new format the value derived from the polynomial is removed from each measurement. This does not introduce any long term difference. Anyway, I will have to think more about what all this should do to the orbit. If other users could have a look at the new files (the "willis" files) that could help us understand what is going on ! Best regards, Jean-Paul