Date: Thu, 18 Dec 2003 10:23:11 +0100 (Paris, Madrid)
From: Martine Feissel-Vernier 
To: Ramesh Govind , John Ries ,
    Pascal Willis ,
    Laurent Soudarin ,
    Jean-Michel Lemoine ,
    Suryia Tatevian ,
    Serguei Kuzin 
Cc: Gilles Tavernier ,
    Jean-Jacques Valette ,
    Jean-Pierre Granier ,
    Carey Noll ,
    Edouard Gaulué , feissel@ensg.ign.fr
Subject: Re: Participation in the IDS-GRACE Campaign

Dear colleagues

As far as I know, a part of the expected products have already been
provided to the IDS Data Centers: the LCA contribution and a part of IGN
one. Thank you Laurent and Pascal.


I react here to Pascal's message of Dec 3 (PW) and John's one of Dec 7
(JR) following my Dec 3 message.

- (JR) Gravity fields: GGM01S vs GGM01C

   John warns us that the use of GGM01C might give worse results than
   GGM01S. It is a bit late to change our recommendations. In any case,
   results referred to either field will be analyzed.

- (PW) The order of the gravity field considered.

  I understand the question is a limited model (70*70) versus the complete
  one. I assume that the analysts make their choice as their best
  compromise between practicability and accuracy. If some centers want to
  submit two parallel solutions with the two versions, they are welcome to
  do so.

- (JR) Don't use Jason at all.
  (PW) Ignore a list of 24 stations in using Jason

  Again, it is a bit late to change recommendations. Considering that 24
  stations is almost 1/2 of the network in the same region, and taking
  into account John's remark that the other are deteriorated too, we may
  indeed expect difficulties. We'll see whether the effect can be seen
  globally.

- (PW) Polar motion daily rates estimated or not

  From parallel correspondence between PW and M. Rothacher forwarded to me
  by PW, I understand that MR would accept solutions without estimated
  polar motion daily rates. I therefore maintain my proposal not to
  estimate them for the time being. Indeed, this does not mean that
  solutions based on the other choice would not be analyzed.

- (PW) Reference epoch of weekly TRFs

  We were requesting that this epoch be the central date of the data
  involved. This epoch would be considered in internal analyses of the
  solutions. However, in the comparisons between solutions we'll have to
  assume that all solutions for a given week are referred to the same
  epoch, that will be the MJD of the middle of the week. It's the
  analyst's responsibility to define his reference epoch, knowing that
  among other constraints.

- (PW) Compatibility with the IERS Campaign.

  As Pascal says, the preparation of this campaign is only starting, while
  the currently planned schedule is very tight, with a submission deadline
  next spring and a final discussion in October 2004! I doubt that by this
  time we shall be in a position to provide a unified IDS solution that
  would make sense. However, IDS might participate in two ways:
  1. with some analysis centers providing solutions - individual center
     solutions will be accepted, as well as multitechnique ones, e.g.
     Doris+SLR;
  2. the analysis coordinator making in parallel comparisons and tests on
     these submisisons, using the experience gained from the IDS-GRACE
     campaign.

- (PW) Why a ccc.snx.readme file?

  To give an example of what I mean, let's consider the directory
  /doris/products/sinex_series/ignwd. It currently hosts four different
  solutions, each one having its .dsc file. As a user, I would like to
  avoid going down a directory with hundreds of files (it takes time
  to have all lines on the screen), then reading these four files (several
  pages each) to decide what series I want to download. I would prefer to
  find a file at the upper level which tells me in a few lines what are
  the major characteristics of each of the solutions and their start &
  end dates. This is the purpose of the ccc.snx.readme file. We also have
  to think of geophysicists looking for Doris-derived stations motions,
  for example. By the way, this file should also include the email address
  of the person to be contacted for more information.

Best regards
Martine