From jimr@maia.usno.navy.mil Tue Dec  1 09:48:11 EST 1998
Received: (from jimr@localhost)
	by maia.usno.navy.mil (8.8.6 (PHNE_14041)/8.8.6) id JAA06579
	for gpst@maia; Tue, 1 Dec 1998 09:45:25 -0500 (EST)
From: Jim Ray (USNO 202-762-1444) 
Message-Id: <199812011445.JAA06579@maia.usno.navy.mil>
Subject: Circular 6 -- Report of 2nd meeting
To: gpst@maia.usno.navy.mil
Date: Tue, 01 Dec 1998 9:45:24 EST
X-Mailer: Elm [revision: 212.4]
Status: RO

Circular 6 (01 December 1998)

Subject:  Report of the 2nd Working Group meeting

Dear Colleagues,

Below is a brief report of the 2nd meeting of the IGS/BIPM Pilot 
Project held yesterday.  Please excuse me if I have been incomplete
or imprecise.

Best regards,
--Jim

----------------------------------------------------------------------------

The IGS/BIPM timing project participants met in a splinter session
associated with the 30th annual PTTI meeting.  The gathering was held at
13:00 on Monday, 30 November 1998, at the Hyatt Regency Hotel, Reston,
Virginia.

Gerard Petit opened the session by explaining that Claudine Thomas had
recently moved to a new position at the BIPM and that he is now handling
matters related to the BIPM Time Section.  Petit raised the question
whether the proper coordinator for the timing community in the Pilot
Project might be the Consultative Committee for Time and Frequency (CCTF)
rather than BIPM, since the CCTF is the governing body for timing issues.
The CCTF will meet next at BIPM on 20-22 April 1999.  The question was
not resolved.

Jim Ray then reviewed the progress of the Pilot Project over the previous
year, particulary since the first meeting held at BIPM in June.  Within
the IGS, plans for the formation of combined station clock products are
well developed and it is likely that they will become available in early
1999.  These will supplement the IGS's combined satellite clock products
which are already produced.  As in the past, the underlying time scale
for IGS clock products will continue to be aligned to GPS broadcast time.
The IGS would prefer to relate its clocks to UTC but this requires
calibrated links between IGS receivers located at timing labs and UTC(k).
Calibration is also required to enable the GPS carrier phase method to be
be useful for time transfer generally.  Thus the calibration issue will
assume an increasingly significant role in the next year.

Representatives from several participating groups then gave very brief
summaries of their recent activities or plans.

   Judah Levine and his colleagues at NIST and the University of Colorado
   are presenting a full paper on their results at the PTTI meeting.  So
   he focussed on some very recent results which seem to indicate a
   divergence between TWSTFT and carrier phase results beginning in
   September 1998.  Previously, the two techniques tracked one another
   closely.  Possible sources of error were discussed.  In addition,
   Levine showed some periods when two different approaches for analysis
   of the GPS carrier phase data yield significantly different clock
   results.

   Ching-Haur Chang (National Standard Time & Frequency Laboratory,
   Taiwan) described a system being developed to slave a remote crystal
   oscillator to a master clock using GPS carrier phase observations.
   The stability of the remote oscillator can be improved by at least
   four orders of magnitude.

   Juan Palacio (Real Instituto y Observatorio de la Armada) noted that
   SFER is already an IGS station and plans are underway to install a
   second completely independent geodetic receiver there.

   John Davis (National Physical Lab, UK) described their efforts to
   install an IGS station.  Their group is also considering an involvement
   with GPS data analysis.

   Jim Rohde (Earth Orientation Dept., USNO) noted the clock results now
   available at USNO for a number of IGS stations in two processing modes;
   see http://maia.usno.navy.mil/gpsclocks/index.html.

   Carine Bruyninx (Royal Observatory of Belgium) and her colleagues will
   have a poster with further details at the PTTI meeting.  She showed
   interesting effects in GPS data from the BRUS station due to the
   environmental changes caused by building construction near the receiver.

   Gregor Dudle (Swiss Federal Office of Metrology) and his colleagues will
   present their results from a trans-Atlantic test in two papers in the
   PTTI meeting.

   Demetrios Matsakis (Time Service Dept., USNO) showed the plans of their
   group for instrumental calibration studies.

   Gerard Petit described several efforts underway at BIPM, in close
   collaboration with the BNM-LPTF group at the Paris Observatory.  They
   include characterization of the Ashtech Z12T receiver, temperature
   stabilized antennas, and short baseline tests between BIPM and
   BNM-LPTF.

John Davis then reported on the recent activities of the TWSTFT working
group.  A check has shown that two-way facilities are colocated with 
relatively few IGS stations and that these are all located in Europe or
North America only.

Next, a general discussion on instrumental calibration issues was
opened by Petit.  He noted that TAI/UTC is accessed by UTC(k) maintained
by the various timing labs, together with Circular T values.  By
convention, the physical realization of UTC is 1 pps pulses.  However,
GPS receivers maintain their own separate timing circuits which are
driven from external frequency standards at much higher frequencies,
typically 5 or 20 MHz.  Even if a 1 pps input is used by the receiver,
the ultimate timing is set by the high frequency signal.  There was
considerable discussion of the absolute importance of maintaining
consistency between these various signals within the timing labs.  
Levine noted the major effort at NIST devoted to establishing a 1 pps
reference plane and that all measurements must be careful related to
that plane.

Petit noted that the links between the various UTC(k) are currently
ensured by calibrated "classical GPS time receivers."  A few time
receivers have been calibrated absolutely while the others have been
calibrated differentially.  He estimated that the system has an
overall accuracy at about the 3-ns level.  Matsakis questioned this
accuracy estimate, at least in the past although it may now be
reasonable.  While it was generally conceded that instrumental
calibration is the central issue that must be resolved for time
transfer applications, Petit expressed the view that effects on
instrumental stability are sufficiently well understood already
that GPS carrier phase can be usefully applied for frequency transfer,
at least for systems maintained under metrological conditions.  It
was suggested that perhaps a sub-group could be organized to focus
attention on the issues and approaches surrounding the calibration
problem.

A date for the next meeting of the Pilot Project participants was
briefly considered.  One possibility would be in April 1999 in
association with the joint meeting of the European Frequency and
Time Forum and the IEEE Frequency Control Symposium, followed by
the CCTF meeting the next week.  However, such an intense period
of meetings may already be too busy and a later date might be
better.