Important message: will retire after October 2022.
For up-to-date information, please visit

IGS Workshop 1999 Summary Recommendations


[IGSMAIL-2359] IGS Workshop Summary
Thu Jul 8 04:01:35 PDT 1999

IGS Electronic Mail      Thu Jul  8  4:01:35 PDT 1999      Message Number 2359

Author: Tim Springer and Yehuda Bock
Subject: IGS Workshop Summary

Dear IGS colleagues,

Below is a summary of the IGS analysis center workshop, which took place from
8-10 June 1999 at the Scripps Institution of Oceanography in La Jolla,

The appendix contains an e-mail with the AC and AAC action items which
resulted from the workshop. This e-mail was distributed to the ACs and
AACs on 16 June 1999 and is included for completeness.

We would like to remind all authors to provide a written
contribution by 31 August 1999. Full length papers are preferred but
an extended abstract summarizing your presentation is also acceptable.
All submissions for publication in the workshop proceedings should be
sent to the IGS Website at:

The workshop at Scripps was most enjoyable and stimulating and on behalf
of the IGS I would like to thank my co-convenor, Yehuda for his great
effort and Myra Medina for the local organization which was quite excellent!

Summary of the 1999 IGS AC workshop
The 1999 IGS analysis center workshop dealt with 2 major topics.
 - Real- and near-real time products and applications
 - Long-term stability and accuracy of GPS Reference Frame

The workshop was opened with a welcome address by John Orcutt, Director
of IGPP, and an introduction by Yehuda Bock.
After the opening the status and challenges facing the IGS were
addressed by Chris Reigber, Ruth Neilan, Carey Noll, Tim Springer,
Angie Moore and Yehuda Bock.

Topic 1: Real- and near-real time products and applications
The position paper "Moving IGS products towards real-time" written by
Gerd Gendt, Peng Fang, and Jim Zumberge, proposed the generation of
both more rapid and frequent IGS products for (near-) real-time usage.
The generation of these "ultra rapid" products should start on
3 October 1999 (GPS week 1030). The products, which will be delivered
every 12 hours (2x per day), will contain a 48 hour orbit arc from which
24 hours are real orbit estimates and 24 hours are orbit predictions.
The latency of this product will be only 3 hours.
This position paper, which was available before the workshop, may
be obtained from the IGSCB at the working location of:
Note that once the proceedings of the workshop are complete, it will be
moved to IGS Publications page at:

One of the "forces" moving the IGS towards real-time products is
GPS meteorology. Applications of IGS products for ground and space
based meteorology were presented by Seth Gutman, Mike Bevis, and
Chris Rocken.

Another "force" moving the IGS towards real-time are the low Earth
orbiter (LEOs) missions where again meteorology is of primary interest.
Mike Watkins presented the status and recommendations of the IGS LEO
working group. This working group, with support from other IGS components,
will soon issue a call for participation
for a "LEO pilot project" where the focus will be mainly on the data
gathering issues (1 sec tracking data in hourly batches in near-real-time).
Precise orbit determination (POD) of the LEO satellite (either OERSTED or
SUNSAT, two satellite missions currently underway with GPS receivers) will also
be of interest. This pilot project is planned for October 1999.
Jim Zumberge and Angie Moore discussed two issues closely related to the
LEO missions, i.e., the LEO ground network and suggestions for a new
GPS binary data format.

Topic 2: Long-term stability and accuracy of GPS Reference Frame
The position paper "Achieving mm site positions and mm/yr site velocities
for geodynamics" written by Tom Herring, Geoff Blewitt, and Remi Ferland
was unfortunately not available before the workshop. However, the interesting
presentation by Tom Herring, and also the presentation given by
Mike Heflin, showed that there are significant
geographically correlated variations in the GPS derived time series.
Tom's prime conclusion was that if mm/yr velocities are expected we should
"call him in 10 years". However, a presentation given by Hadley Johnson
on site stability issues indicated that we might have to wait 20 years.

The geographically correlated effects may represent actual
deformations although Tom Herring concluded that it is likely that they
are caused by remaining deficiencies in IGS orbit modeling. These clearly
are a cause of concern and require further attention.
It also became clear (once again) that the feedback from the ACs to the
station operators is still far from optimal.
With respect to site stability calibration it would be helpful to have two
receivers at a site. which was also recommended at the
Network Workshop in Annapolis.

Zuheir Altamimi and Remi Ferland showed and discussed the quality of the
ITRF-97. Clearly the ITRF-97 is again a major improvement w.r.t. its
predecessor (ITRF-96). Remi Ferland showed results which indicate that
the VLBI and GPS based polar motion estimates will be in much better
agreement in the ITRF-97 than they are in the ITRF-96, where a 0.3 mas
bias is observed. It was decided that the IGS will adopt the ITRF-97
reference frame starting with GPS week 1021, 1 August 1999.

Presentations by Jim Ray, Markus Rothacher, and Arthur Niell dealt
with the problems of mixing different receivers and antennas in the
global network. The antenna phase center offsets and
variations, of both the receiver and transmit (satellite) antennas, represent
one of the major remaining error sources in GPS data analysis.
Jim Ray showed that the used of the different code observables
(CA vs P1) causes problems for the IGS clock estimates and combination;
for more details about this problem see IGSMAIL #2320.

Important in view of the long-term stability was the presentation
by John Rush from NASA on the microwave spectrum issues and new GPS
frequencies. Three important action items were suggested:

IGS Contact the International Telecommunication Union, and, each
IGS member contact their local radio administration to encourage:
 (1) The preservation of the radionavigation satellite service band
     1559-1567 MHz for use by radionavigation satellite services, with
     no sharing of any portion of the band with MSS users.
     (WRC-00 Agenda Item 1.9)

 (2) Addition of the radionavigation satellite service allocation to
     cover the band 1164-1188 MHz.
     (WRC-00 Agenda Item 1.15.1)

 (3) The addition of an allocation for use in the space-to-space direction
     for radionavigation satellite services in the 1215-1260 MHz and
     1559-1610 MHz bands.
     (WRC-00 Agenda Item 1.15.2)

The workshop concluded with an interesting tour along the posters of all
the IGS analysis centers followed by presentations summarizing the status
and activities of all the IGS working groups.

Workshop recommendations
Besides the action items for the ACs and AACs (see appendix) and the
three action items proposed by John Rush the following recommendations
resulted from the workshop.

R1: The operational, regional and global data centers are asked to optimize
    the data transfer of the hourly RINEX-files in support of the ultra-
    rapid products

R2: The IGS LEO WG will issue a call for participation for a LEO pilot
    project which will start in October 1999.

R3: The IGSCB, in cooperation with the Infrastructure Committee, should
    enhance the newly developed receiver and antenna name list by adding
    the tracking techniques of each receiver. This is important in view of the
    available observables and their effect on the IGS clock estimates.

R4: The IGS ACs will have to decide and advertise whether the IGS clock
    estimates will be based on CA- or P1-code observations. The IGS
    will have to provide a table containing the CA-P1 code biases.

R5: Tom Herring will draft a letter for the IGS Governing Board
    to request VLBI tracking of the GPS satellites in order to determine
    the L1 and L2 phase center offsets of the GPS satellite antennas.

R6: Improve the feedback from the AC to the station operators in case of bad
    station performance. Better advertise the availability of feedback to the
    station operators. This is the responsibility of the Network Coordinator.

Kind Regards,
         Tim Springer
         Yehuda Bock

Appendix: E-mail send to ACs and AACs on 16 June 1999.
AC and AAC action items:

Date: As soon as possible
1) ACs start submitting ERP files with their predicted orbits.
   These ERP files should include at least the PM offsets and their rates for
   the day of the predicted orbits. The inclusion of more values is allowed
   if this is convenient for the AC. Of course the ERP files should be
   consistent with the predicted orbits.
   The standard IGS ERP format should be used.

2) ACs start submitting clock files with their final orbits and the clock
   files should contain the stations coordinates in the header (see new
   format description below).

3) The IGS Analysis Center Coordinator (ACC) will contact the IERS, Duncan
   Agnew, and Hans-Georg Scherneck (et. al.) about setting up some kind of
   ocean loading service to provide the ACs, AACs, and other IGS customers
   with ocean loading corrections for the different IGS sites.

4) The ACs are encouraged to (better) advertise the availability of
   accuracy codes in the orbit (SP3) files. The ACs developing and
   distributing their own GPS processing software are encouraged to add
   the capability to automatically use the orbit accuracy codes to their

Date: July 4, 1999. GPS week 1017
1) All ACs, AACs, RINEX files, site-logs, and IGS.SNX should contain the
   new receiver and antenna names! All IGS products from this week on should
   use the new receiver and antenna names. The "old but still valid" names
   become "invalid" for all IGS products after this date.

2) The 4-character id used for both the site-log and the RINEX file name
   becomes the official 4-character id for the respective
   site-receiver-antenna combination. This 4-character id should be used to
   label ALL results from this site-receiver-antenna combination, e.g.,
   SINEX, clock files, etc. etc. New 4-character id's will be the
   responsibility of the IGSCB (of course stations may propose a 4-char id).
   The IGSCB will notify the IGS about changes in the 4-char id's by sending
   an IGSMAIL.

Date: August 1, 1999. GPS week 1021
1) All ACs, and AACs start using the ITRF'97 as terrestrial reference frame
   for their IGS products. Zuheir Altamimi will make available a special
   IGS SINEX file containing the ITRF'97 positions and velocities of the
   51 reference stations (ITRF97_IGS_RS51.SNX and ITRF97_IGS_RS51.SSC).
   Remi Ferland will work together with Zuheir to ensure the correct list
   and naming (4-char) of the 51 sites.
   All ACs and AACs are encouraged to test the new reference frame and
   to study the changes between the ITRF'96 and ITRF'97 results.
   Please report your results to the Analysis Center Coordinator.
   The ACC will (try to) summarize the observed differences in the IGS
   products and document and distribute them using the IGSMAIL.

Date: October 03, 1999. GPS week 1030
1) The ACs are encouraged to start submitting "ultra-rapid" products
   two times per day according to the scheme as outlined in the
   position paper of the IGS workshop:
     title:   Moving IGS products towards real-time
     authors: Gerd Gendt, Peng Fang, Jim Zumberge

   To avoid naming conflict I would like to propose to use the following
   naming convention for the products. Use the first two characters of
   your AC and add a "u" as third character, e.g., cou, emu, esu, gfu...
   The combined products will be labeled "igu" as proposed in the position

This completes my list of must urgent AC action items. Early next week I
hope to distribute (after consulting with the organizers at SIO and the
IGSCB) the complete list of recommendations and action items of the
IGS workshop. Two important items on that list from an AC point of view
are the upcoming call for participation in the LEO pilot project,
and P1-CA code biases and its effect on the IGS clock estimates.

I hope to have informed/remembered you sufficiently.

Kind regards,
         Tim Springer

cc: IGS ACs

Clock format change to allow for station coordinates in header.
Author: Jim Ray
Based on discussions at the IGS Analysis Center Workshop, held last week
at Scripps Institution of Oceanography, Analysis Centers were asked to
include station coordinates in their clock solution files.  These
coordinates should correspond to the values used in the clock analysis
(whether fixed or adjusted).  This will be a great convenience for users
of these files.

This change can be accommodated in the existing "SOLN STA NAME / NUM" header
record by using 35 previously blank spaces (for stations on or near the
Earth's surface).  Note that the geocentric coordinates should be given in
millimeter units.  The relevant portion of the changed format is:

 |SOLN STA NAME / NUM | For each station/receiver included in    |            |
 |                    | the clock data records, as well as the   |            |
 |                    | analysis reference clock even if it has  |            |
 |                    | zero values and is not included in the   |            |
 |                    | data records (number given in the        |            |
 |                    | previous header record), include one     |            |
 |                    | record with the following information:   |            |
 |                    | - 4-character station/receiver name      |   A4,1X,   |
 |                    |   designator                             |            |
 |                    | - Unique station/receiver identifier,    |    A20,    |
 |                    |   preferably the DOMES number for        |            |
 |                    |   fixed stations                         |            |
 |                    | - Geocentric XYZ station coordinates     |   I11,X,   |
 |                    |   corresponding to the analysis clock    |   I11,X,   |
 |                    |   values reported (in millimeters!)      |    I11     |
 |                    |                                          |            |
 |                    | * REQUIRED for data types "Ax"           |            |
 |                    |                                          |            |

An example of such a record would look like:

GOLD 40405S031           -1234567890 -1234567890 -1234567890SOLN STA NAME / NUM


Earlier this month, clarifications were added to deal with reporting
analysis clock reference values when the time scale has been re-aligned.

All these changes are backwards compatible.  The complete documentation
is available at .



 Tim Springer                      Astronomical Institute, University of Berne 
 springer at            Sidlerstrasse 5, CH-3012 Bern, Switzerland      Tel: +41 31 6318592    Fax: +41 31 6313869

[Mailed From: Tim Springer <tim.springer at>]