From jimr@maia.usno.navy.mil Thu Jul 9 03:28:08 EDT 1998
Received: from relay5.eunet.fr (relay5.eunet.fr [193.107.193.102]) by maia.usno.navy.mil with SMTP (8.7.5/8.7.3) id DAA28075 for ; Thu, 9 Jul 1998 03:26:48 -0400 (EDT)
Received: from relay1.fnet.fr by relay5.eunet.fr (5.65c8d/96.05.03)
via EUnet-France id AA25140; Thu, 9 Jul 1998 09:26:32 +0200 (MET)
Received: from si.bipm.fr (si.bipm.fr [193.104.126.10])
by relay1.fnet.fr (8.8.5/8.8.5) with SMTP id JAA02749
for ; Thu, 9 Jul 1998 09:26:25 +0200 (MET DST)
Received: from tai1 by si.bipm.fr; Thu, 9 Jul 1998 09:24:06 +0200
Date: Thu, 9 Jul 1998 08:55:36 +0100
From: Claudine Thomas
Subject: IGS/BIPM Pilot Project
To: gpst@maia.usno.navy.mil
X-Mailer: Z-Mail Pro 6.1 (Win32 - 021297), NetManage Inc.
X-Priority: 3 (Normal)
Message-Id:
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=ISO-8859-1
Content-Transfer-Encoding: 8BIT
Status: RO
From: C. Thomas and J. Ray, Co-chairs
To: Participants of the IGS/BIPM PILOT PROJECT
TO STUDY ACCURATE TIME AND FREQUENCY COMPARISONS
USING GPS PHASE AND CODE MEASUREMENTS
Circular 5 (09 July 1998)
Subject: Report of the 1st meeting
Dear Colleagues,
Please, find enclosed the report of the 1st meeting of the
IGS/BIPM Pilot Project, that was held on 22-23 June 1998
at the BIPM, Sèvres, France.
We also would like to bring your attention to the session
on GPS phase measurements and timing applications at next
PTTI meeting (deadline for abstracts 15 July).
We intend to organize a short meeting of the IGS/BIPM
Pilot Project during or just before the PTTI meeeting. More
information on this will be available later.
With our best regards,
Jim Ray
Claudine Thomas
09 July 1998
IGS/BIPM Pilot Project
to study accurate time and frequency comparisons
using GPS phase and code measurements
Report of the 1st Meeting
held at the BIPM, Sèvres, France
on 22 and 23 June 1998
Attendees to the 1st meeting of the IGS/BIPM Pilot Project were representatives
of time laboratories (BNM-LPTF, CH, IEN, IFAG, LHA, NPL, ORB, SP, USNO),
representatives of the IGS (IGS Central Bureau, IGS coordinator), and members of
the BIPM. Two IGS analysis centres (AIUB and JPL) were also represented, and was
also present the Chairman of the Sub-group on phase measurements of the CCTF
Working Group on Two-Way Satellite Time Transfer. It should be noted that some
of the time laboratories represented are also IGS stations (IFAG, ORB, USNO) or
have a close relationship with an IGS station (SP with Onsala Observatory, for
example). The meeting was co-chaired by J. Ray (USNO) and C. Thomas (BIPM).
The objective of the meeting was to initiate a dialogue between the geodetic and
timing communities in order to arrange a collaboration between both on the
subject of accurate time and frequency comparisons using GPS phase and code
measurements.
The meeting was organized as a series of short and informal presentations,
followed by questions, comments and discussions. The topics approached were:
- Deployment of GPS receivers: K. Jaldehag (SP) ;
- GPS data analysis: J. Kouba (IGS Coordinator), J. Zumberge (JPL) ;
- GPS receiver design: T. Schildknecht (AIUB) ;
- Analysis of instrumental delays: R. Beard (US NRL) ;
- Site-dependent effects: J. Johansson (SP/Onsala Observatory), G. Petit
(BIPM) ;
- Time and frequency comparisons: L. Prost (CH), J. Levine* (NIST) and K.
Larson* (University of Colorado), G. Petit (BIPM), R. Tjoelker (JPL), J.
Davis (Chairman of the Sub-group of the TWSTT group) ;
*Since J. Levine and K. Larson could not attend, C. Thomas made a summary of
the work they presented at the Frequency Control Symposium, in May 1998.
- Time scale generation: D. McCarthy (USNO).
A detailed description of the presentations is not given here, but a summary of
the products available from the BIPM and the IGS is given in the Appendix.
The main points of the discussions are summarized below.
1. Needs of the timing community
The timing community unanimously expresses the need for the development of a
frequency transfer method with the capability of comparing ultra-stable
oscillators, such as caesium fountains and ion trap standards, with an
uncertainty of 1 part in 10^15 over averaging times of one day or less. Use of
GPS phase measurements, a technique already largely developed for geodesic
purposes, could fulfil this need. The timing community thus needs the expertise
gained in the geodetic community.
2. Needs of the geodetic community
The IGS products (accurate positioning, precise satellite ephemerides, etc.) are
expressed in the official reference frame ITRF, rather than in the GPS reference
frame WGS84. The IGS is now developing time products, the so-called 'clock
solution', in which clocks on-board GPS satellites and in operation at IGS
stations should be referenced to the official reference time UTC, as delivered
by the BIPM, rather than to GPS time. There is thus a need to establish accurate
links of the clocks maintained in IGS stations to UTC.
3. The geodetic receivers
The IGS stations mainly use Ashtech (model Z12) and Allen-Osborne Associates
(AOA Turbo-Rogue) receivers. These receivers operate with an internal oscillator
or with an input frequency (5, 10 or 20 MHz) delivered by the local clock. Data
output from these receivers are L1 and L2 phase and P-code measurements, and L1
C/A code measurements dated in the time scale GPS time. Generally the receivers
also output a 1 pps signal locked on to GPS time which can be compared, if
necessary, to the 1 pps signal of the local clock through an external time
interval counter.
Some time laboratories operate Ashtech receivers model Z12T, derived from the
Z12. This receiver is entirely driven from outside (no internal oscillator) and
can accept as input the 1 pps signal of the local clock. Only five units of this
type are in operation in the world (two at the OFMET-named GeTT terminals-, one
at the NPL, one at the BIPM, and one at the BNM-LPTF) and were specially
designed by Ashtech to fulfil the need of the timing community. It is not yet
clear if this type of receiver is actually in the Ashtech catalogue.
For time and frequency transfer, there is no absolute need to operate an Ashtech
Z12T receiver; standard geodetic receivers can provide the necessary data if it
is possible to reference the measurements to the local clock with sufficient
accuracy. Time laboratories, however, have always operated receivers, such as
the one-channel C/A code AOA TTR6 model, into which the local 1 pps was entered.
There is thus a preference of the time community for the Ashtech Z12T model.
4. Antennas and cables
Geodetic receivers usually operate with Dorne-Margolin antennas, equipped with a
choke-ring. High-quality geodetic measurements are usually obtained with
antennas placed above (about 1 m) a large grass area. This is preferred to a
roof-top mounting which is too sensitive to multipath effects. At any event, it
is highly recommended to avoid installation of the antenna above a metal plate
or in a metallic environment. Two antennas should at least be separated by a
distance of about 10 wavelengths (2.5 m). At the SP and Onsala Observatory,
there is one main antenna, placed on a concrete pillar (which is kept at a
constant temperature to control dilatation), feeding all main receiver units
placed in the laboratory. Covering the antenna with a dome, or changing the dome
shape is not without consequences: it may induce a change in the phase pattern
of the antenna. Hemispherical domes are sometimes recommended.
These different aspects of antenna mounting are generally not well known by time
laboratories where, however, the temperature sensitivity of the antennas has
been investigated. A sensitivity of about 1ps/°C has been observed on L1 phase
measurements. This is not actually critical and could be reduced by the use of
temperature-stabilized antennas (TSA model from 3S Navigation) or by placing
antennas in ovens, though the effects of such enclosures on the positions of the
L1 and L2 centres of phase should be investigated. Other studies have revealed a
temperature sensitivity of the cables. Measurements indicate an effect of about
1 ps/(m.°C) which can thus induce variations of several hundreds of ps for long
cables placed in full sunshine. Low-loss and low-temperature coefficient cables
are commercially available and already tested at the USNO. They should also be
flexible enough for easy installation. Connectors (between the antenna and the
antenna cable), frequency multipliers (for instance creating a 20 MHz signal
needed by the Ashtech Z12T), and main units of the receivers are also sensitive
to temperature variations (30 ps/°C for the BIPM Ashtech Z12T unit and 60 ps/°C
for one of the GeTT terminals) . At the SP, the antenna cable is inside the
temperature-controlled concrete pillar on which the antenna is placed. An
important point is that the temperature sensitivity of the entire chain (from
the antenna to the main unit) depends on the signal frequency, and thus may have
different impacts on phase and code measurements. Cables and antennas are known
to be insensitive to humidity variations.
5. Receiver calibration
Geodetic receivers provide code and phase measurements. The phase measurements
are intrinsically ambiguous at the level of integer cycles but are very precise
and less sensitive to multipath errors than code measurements. Code measurements
may be accurate if the absolute values of all delays involved are known. The
accuracy of frequency comparisons obtained from phase measurements is limited by
the stability of the equipment over the averaging time of interest; there is
therefore no need to know the absolute delays. In contrast, these are necessary
for accurate time transfer between remote clocks operated in IGS stations and in
timing laboratories. It may thus be necessary to calibrate the geodetic
receivers involved.
Geodetic receivers could be calibrated in a differential way, through the
transportation of a portable receiver and side-by-side comparison of P-code
pseudo-range measurements. An absolute calibration is then necessary for one
single receiver, which could be done by comparison with a one-channel C/A code
GPS receiver of the usual type in operation in timing laboratories. The absolute
delay of the geodetic receiver would not be known very accurately, however, and
no calibration would be provided for the L2 frequency (essential for estimation
of ionospheric delays). Another method consists of using a GPS signal simulator.
Such a measurement has already been performed in a French military laboratory
(LRBA) on the two Ashtech Z12T units belonging to the BIPM and the BNM-LPTF.
This, however, concerns only the main units and not the antennas and associated
equipment. The US NRL has also at its disposal a GPS signal simulator with the
capability of emitting the signals in a temperature-stabilized chamber which can
house the entire equipment. It was suggested that one of the Swiss GeTT
terminals could be absolutely calibrated at the NRL before the end of 1998
(taking advantage of an experimental study where this terminal will be placed at
the USNO). It should be noted that the uncertainty on this absolute calibration
may not be smaller than a few nanoseconds.
6. Time scale generation
When accurately compared and linked to the time laboratories network, the clocks
in operation in IGS stations could take part in TAI. This would increase by
about 40 the number of clocks contributing to TAI (23 hydrogen masers and 16 HP
5071A) and would provide a regular link between the IGS clocks and UTC. It could
help improve the short-term and long-term stability of TAI.
The IGS could also produce a near real-time time scale from a number of clocks
in the IGS stations and time laboratories. This time scale could be available
daily in a provisional form and within a few days in its final version.
7. Conclusions
It is highly desirable that some institutes in the world be at once timing
laboratories contributing to TAI and IGS stations, with a strong link between
the timing and the geodesic part of the institute. This is already the case for
national timing centres such as ORB, IFAG, USNO, but should be reinforced. This
would allow:
- to establish a link between the clocks in IGS stations and UTC, with the
immediate consequence that IGS products would benefit from the full accuracy
and stability of an internationally agreed reference time scale ;
- to carry out accurate time and frequency comparisons between time
laboratories, as a benefit from the data processing provided by IGS analysis
centres (ultra-accurate coordinates, estimation of ionospheric delays, precise
satellite ephemerides, etc.).
The actions decided at the meeting can be listed as follows:
- Provide information to both communities, so the advantages of the
collaboration may be understood.
- Encourage timing laboratories to become IGS stations (information on the web
site http://www.maia.usno.navy.mil/gpst.html).
- Continue studies on accurate frequency comparisons using GPS code and phase
measurements in order to demonstrate whether the goal of 1 part in 10^15 over
averaging times of 1 day or less can be reached.
- Continue studies on the comparisons with different time and frequency transfer
methods, such as GPS/GLONASS code measurements and Two-Way satellite
time transfer.
- Investigate the absolute and differential calibration of geodetic GPS
receivers.
- Use the web site of the IGS/BIPM Pilot Project
(http://www.maia.usno.navy.mil/gpst.html, also available from
http://www.bipm.fr) for exchange of information and for organisation of the
above studies.
A session on GPS phase measurements is scheduled at the next PTTI meeting
(Reston, USA, 1-3 December 1998); the IGS/BIPM Pilot Project will take this
opportunity to organize a short meeting (probably on 30 November 1998). An IGS
Workshop on IGS stations, network, and sub-network for timing is also scheduled
(Annapolis, MD, USA, 2-5 November 1998). In February 1999 a report on the
activities of the IGS/BIPM Pilot Project will be written for presentation at the
CCTF meeting in April 1999.
Acronyms of institutions
AIUB Astronomical Institute, University of Bern, Bern, Switzerland
BIPM Bureau International des Poids et Mesures, Sèvres, France
BNM-LPTF Bureau National de Métrologie - Laboratoire Primaire du Temps et
des Fréquences, Paris, France
CCTF Comité Consultatif du Temps et des Fréquences
CH Consortium of laboratories in Switzerland
IEN Istituto Elettrotecnico Nazionale Galileo Ferraris, Turin, Italy
IFAG Institut für Angewandte Geodäsie, Frankfurt am Main, Germany
IGS International GPS Service
JPL Jet Propulsion Laboratory, Pasadena, CA, USA
LHA Laboratoire de l'Horloge Atomique, Orsay, France
LRBA Laboratoire de Recherches Balistiques et Aérodynamiques, Vernon,
France
NIST National Institute of Standards and Technology, Boulder, CO, USA
NPL National Physical Laboratory, Teddington, United Kingdom
NRL Naval Research Laboratory, Washington D.C., USA
OFMET Office Fédéral de Métrologie, Bern, Switzerland
ORB Observatoire Royal de Belgique, Brussels, Belgium
SP Swedish National Testing and Research Institute, Boras, Sweden
USNO US Naval Observatory, Washington D.C., USA
Appendix: Some information on the IGS and the BIPM
IGS, International GPS Service, http://igscb.jpl.nasa.gov
IGS Global Data Centres:
- Institut Géographique National, Paris, France ;
- NASA Goddard Space Flight Center, Greenbelt, MA, USA ;
- Scripps Institution of Oceanography, University of California, San Diego, CA,
USA.
IGS Analysis Centres:
- Astronomical Institute, University of Bern, Bern, Switzerland ;
- European Space Operations Center, European Space Agency, Darmstadt,
Germany ;
- GeoForschungsZentrum, Postdam, Germany ;
- Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA,
USA ;
- National Oceanic and Atmospheric Administration, Silver Spring, MA, USA ;
- Natural Resources Canada, Ottawa, Ontario, Canada ;
- Scripps Institution of Oceanography, University of California, San Diego, CA,
USA.
IGS products (product, availability, accuracy)
- GPS satellite ephemerides
* Predicted: real-time, 50 cm ;
* Rapid: 22 hours, 10 cm ;
* Final: 2 weeks, 5 cm.
- GPS clocks
* Predicted: real-time, 150 ns ;
* Rapid: 22 hours, 0.5 ns ;
* Final: 2 weeks, 0.3 ns.
- IGS station locations
* Weekly Solutions: 4 weeks, 3-5 mm.
- Earth orientation - Pole
* Rapid: 22 hours, 0.2 mas ;
* Final: 2 weeks, 0.1 mas.
- Earth orientation - Pole Rates
* Rapid: 22 hours, 0.4 mas/d ;
* Final: 2 weeks, 0.2 mas/d.
- Earth orientation - UT1-UTC
* Rapid: 22 hours, 300 microsec ;
* Final: 2 weeks, 50 microsec .
- Earth orientation - Length of Day
* Rapid: 22 hours, 60 mics/d ;
* Final: 2 weeks, 30 mics/d.
BIPM, Bureau International des Poids et Mesures, http://www.bipm.fr
Time Section, Pavillon de Breteuil, Sèvres, France
BIPM Time Section products (availability, accuracy), published in the monthly
Circular T
- UTC - UTC(k) [for MJDs ending in 4 and 9]
Final: 6 weeks, a few ns (depends on the laboratory k).
- TAI - TA(k) [for MJDs ending in 4 and 9]
Final: 6 weeks, a few ns (depends on the laboratory k).
- UTC - GPS time [daily]
Final: 6 weeks, 10 ns.
- UTC - GLONASS time [daily]
Final: 6 weeks, a few hundreds of ns.
- Rates relative to TAI of the clocks contributing to TAI [monthly]
Final: 6 weeks, 0.1 ns/d.
- Weights of the clocks contributing to TAI [monthly]
Final: 6 weeks.
- A certain amount of information, such as data from primary frequency
standards, GPS and GLONASS tracking schedules, equipment of contributing
laboratories, frequency steering applied to TAI, GPS receiver calibration
reports..., is available in the Annual Report of the BIPM Time Section (annual
publication, at the end of February for the previous year) or on request from
the BIPM.