[IGSMAIL-0414] Recommendations for Analysis Centers

[Jan]

******************************************************************************
IGS Electronic Mail      Tue Dec 14 13:47:53 PST 1993      Message Number 0414
******************************************************************************

Author: Jan Kouba/Jim Zumberge
Subject: Recommendations for Analysis Centers

Dear Colleagues,

As the official start of the IGS - January 1, 1994 - is fast
approaching, it is important that the consensus reached during the
Ottawa workshop be implemented by all Analysis Centers as soon as
possible and not later than January 1, 1994.  Below is a summary of
recommendations to be implemented:

1. All Analysis Centers fix or tightly constrain all their stations on
   the following 13 station list:

   Algonquin Park (ALGO)
   Fairbanks      (FAIR)
   Goldstone      (GOLD)
   Hartebeesthoek (HART)
   Kokee Park     (KOKB)
   Kootwijk       (KOSG)
   Madrid         (MADR)
   Santiago       (SANT)
   Tromso         (TROM)
   Tidbinbilla    (TIDB)
   Wettzell       (WETT)
   Yarragadee     (YAR1)
   Yellowknife    (YELL)

   ITRF92 coordinates and velocities are to be used for these sites (see
   IGSMAIL #263).  The fixed or constrained coordinates are to be
   updated (to account for the station velocities) at least once a
   month, or more frequently in order to stay well below the 10-mm error
   level.

   Note:  Whenever a change is made which can introduce an erp
   discontinuity, IERS requires an estimate of the shift.  The
   recommended procedure is to analyze several days with both the new
   and old strategies, and indicate the earth orientation parameters for
   both.

2. SP3 will be the format for all orbit products by January 1, 1994.
   Whenever possible, satellite clock information should be included in
   the SP3 files as well.

3. All Centers shall make an effort to submit their orbit and ERP
   products as soon as possible but not later than TWO WEEKS from the
   date of observation.  (Daily submissions or multiple submissions per
   week are recommended).

In addition, the following changes and additions, although not agreed
upon formally, would also be useful and appreciated by IGS customers:

a. SP3 files should contain accuracy estimate codes in their headers;
   include only ephemerides for 24 h (0h-23:45h UTC) as originally
   proposed by IGS, and if possible have the standard 15-min sampling
   interval.

   Note that accuracy codes are important not only for orbit combination
   but also for all users.  Intervals longer than 24 h are redundant and
   wasteful both in storage and communication.  A simple filter program,
   which can be made available through IGS, can concatenate any number
   of files!

b. In addition to MJD, pole x, y, UT1-UTC, and the corresponding
   sigmas, ERP files should contain number of stations, number of fixed
   (or tightly constrained) stations, and number of satellites for each
   day.  (This is an interim step only.  A more detailed discussion on
   the ERP format is currently in progress over e-mail.)

c. Summary files need not contain ERP, but as a feedback to both
   observers and analysts, they should contain statistics pertaining to
   data quality (for example, number of observations, noise sigmas each
   day, and (if possible for each station/satellite), station inclusion
   table, if not already implied in the statistics portion), and orbit
   quality checks (for example, orbit overlaps).

We hope this is not too imposing or restricting.  Any problems,
questions, suggestions and objections please refer to AC coordinator.

   Season's Greetings and Best Wishes for 1994!

      Jan Kouba (for AC coordinator) &  Jim Zumberge (for CB)