[IGSMAIL-1174] CODE Rapid Orbits

[T.A.]

******************************************************************************
IGS Electronic Mail      Fri Dec 29  9:57:28 PST 1995      Message Number 1174
******************************************************************************

Author: T.A. Springer
Subject: CODE Rapid Orbits


Dear IGS colleagues,

Starting today (December 29, 1995) the IGS analysis center CODE 
(Center of Orbit Determination in Europe) makes its rapid orbits available 
as soon as they have been computed. To avoid confusion we will NOT submit 
these orbits to the CDDIS. They will be available on our anonymous ftp 
account only (see below). Furthermore the orbits will be submitted to the 
IGS Analysis Center Coordinator, Jan Kouba and his team at NRC,
to be included in the IGS preliminary orbit combination (coming soon).

The CODE rapid orbit processing starts at 10:00 MET (09:00 UTC) and,
depending on the number of stations available, will be available at 
approximately 12:00 MET (11:00 UTC). The delay is therefore in the order of
12 hours! Our rapid orbits do contain "broadcast" clocks for the satellites.


How to obtain our orbit products
--------------------------------

The rapid orbits (and other orbit products) are available through 
anonymous ftp using:
   ftp ubeclu.unibe.ch (or 130.92.6.11)

Use "anonymous" to login and your full e-mail address as password.
After login type the command:
   cd aiub$ftp

Thereafter you should change to the directory CODE
   cd code

Here you will find our orbits under the names:
   codWWWWD.eph --> Official CODE IGS orbit, 6-12 day delay
   codWWWW7.erp --> Weekly pole file belonging to the official CODE IGS orbits

   codWWWWD.ep1 --> CODE 1-day orbit 4-day delay
   codWWWWD.er1 --> Daily pole file belonging to the CODE 1-day orbits

   codWWWWD.eph_r --> CODE rapid orbit, 12 hour delay
   codWWWWD.erp_r --> Daily pole file belonging to the CODE rapid orbits

   where: WWWW : is the GPSWEEK (eg. 0819 for this week)
              D: is the DAY of WEEK (0=sunday,1=monday...6=saturday)


Precision of the CODE orbit products
------------------------------------

The precision of the CODE rapid orbit will depend mainly on the number
of available stations and their geographical distribution. The table below
list the orbit quality of the rapid orbits of last week (0832) and their
rms compared with our official CODE IGS solution. For comparison we also
listed the results of our normal (4-day delay) 1-day orbits between brackets.

 --------------------------------------------------------------
   #sta     doy   orbit rms(cm) wrt. official CODE IGS solution
 --------------------------------------------------------------
  26 (60)   353   59 (13) <-- only 6 NON-European stations.
  27 (59)   355   46 (13)
  33 (60)   354   28 (13)
  35 (58)   352   26 (13)
  37 (50)   356   23 (16)
  47 (60)   351   15 (13)
 --------------------------------------------------------------

Since the procedure for the computation of our rapid orbits is identical 
to our normal 1-day solution our orbit quality could be as good as 10 cm.
However the number of stations seems to be something between 30 and 40!
Therefore we estimate that our orbits will have a precision of 20 to 30 cm. 

We will also be generating orbit predictions for the next two days and 
will make them available on our "aftp" account (soon). They can be used for 
"real time" apllications. The predicted orbits seem to be of the same 
quality as the predictions we are generation routinely in our normal IGS 
processing, 80cm. Therefore we expect our "rapid predictions" to be better 
then 100 cm which is still a factor of 3 (!) better than the
broadcast orbits. Furthermore we are working on some software 
changes which should improve the quality of our orbit prediction 
significantly (50cm).
 
All the best to all of you for 1996,
         Tim Springer
         CODE analysis center




 ------------------------------------------------------------------------------
 Tim Springer                       Astronomical Institute, University of Berne 
 Tel:   +41 31 6313802              Sidlerstrasse 5, CH-3012 Bern, Switzerland  
 Fax:   +41 31 6313869              ftp: ubeclu.unibe.ch  -or-  130.92.6.11
 e-mail: springer at aiub.unibe.ch          after login: cd aiub$ftp
 ------------------------------------------------------------------------------


[Mailed From: Springer Tim <SPRINGER at aiub.unibe.ch>]