Date: Wed, 8 May 2002 11:35:34 +0100 From: Henno.Boomkamp@esa.int To: Markus Rothacher Cc: Martine Feissel , "John C. Ries" Subject: JASON Campaign Dear Markus, Just before Ottawa I have also been contacted by Martine Feissel about doing a combined JASON campaign between IGS LEO and IDS (international DORIS service). So, it seems that we now have three organisations interested in having a JASON campaign, and I think that everybody agrees that a single combined campaign will be more effective than three small campaigns. Most of the POD centres are involved with at least two of these organisations anyway. John Ries suggested quite sensibly that we might wait for a while until a clean orbit period has been identified, without too many orbit events. I think we can take a bit of time to plan things properly, and get the best results out of the JASON campaign. As far as I can see, we should agree upon the following aspects: (1) Campaign period (date and duration) I would suggest that we do a period of at least two repeat cycles, but probably three (i.e. about one month in total). Most of the centres will have no problem generating three cycles of JASON POD, even if this would need to be repeated at a later date for some reason. With three cycles, we will probably not find a single period without manoeuvres or another event, but with such a long dataset we can afford to reject a few hours around a manoeuvre, without losing much. Alternatively, we select three disconnected periods of 10 days each, but this does not seem very practical. (2) Analysis methods to be applied. The pairwise orbit comparisons are easy to do, and always provide very useful information. Then, to arrive at absolute precision estimates, there are possibilities for JASON that do not exist for CHAMP. In my current CHAMP analysis, I estimate a single, global proportionality factor "D" between SLR residuals and orbit error, and from this I derive at absolute orbit precision estimates. The trouble with this is that different orbits might in fact need different factors D, but with one single reference data set (SLR) you can not really estimate such individual factors without making further assumptions. However, with four or five reference data sets it becomes possible to extend the analysis, and estimate much more subtle relationships between the (relative) comparison results and the (absolute) residuals. I'll try to find some time to work out the details and produce a decent theoretical basis. In any case my suggestion would simply be that for JASON all available tracking data sets are re-processed for verification purposes, using fixed input orbits and exactly the same processing for all cases. In that way we generate an abundance of fairly independent information from SLR, DORIS, GPS, altimetry (and perhaps derived observables, altimetry crossovers, differenced GPS, ...). With some careful analysis, this will enable a more reliable absolute precision estimate than for CHAMP. Depending on the required processing time I can either offer to do this processing at ESOC, or we can distribute the processing over several centres. As long as the input orbits are fixed, and terrestrial reference frames can be guaranteed to be consistent, it does not seem very critical that different POD platforms are used for processing different data sets. What is critical is that the processing is identical for all input orbits, so that resiudals can be linked to the orbit comparisons in a reliable way. So, I would like to ask from you some suggestions on the following: - Campaign period ? Do you agree on 3 full cycles? - Independent verification ? Is it a good idea to simply process all data sets? - Do yo agree that we can distribute these checks over various centres without serious risk of inconsistencies? What cross-checks are needed, if any? - Would you offer to participate in the verification analysis (if so, what data sets ?) Note that this data processing will have to be repeated for every new contribution. It's not really a lot of work if you can more or less automate your system, but it will claim some computer resources. From the part of ESOC, I can offer to maintain the campaign webpages, to run the analysis of verification results from different centres (i.e. assimilate the data that goes into the webpages), and to process one or more of the verification tracking data sets. I suggest that we stay in touch about the plans. At the risk of adding another redundant organisation, I form the ad-hoc "JASON Campaign Committee" consisting of ourselves as the three representatives from CSTG, IDS and IGS LEO. I invite John Ries on this as well, due to his involvement in previous discussions on this subject. So, please "reply to all" with your comments. Best regards, Henno Boomkamp PS Markus - your latest CHAMP orbits seem to be in between CSR and GFZ, but the differences become smaller and smaller. I'll send you another mail as soon as the processing is ready.