[IGSMAIL-1569] IGS Workshop, GB Meeting 1997 in Pasadena

[G.]

******************************************************************************
IGS Electronic Mail      Wed Mar 26 23:48:37 PST 1997      Message Number 1569
******************************************************************************

Author: G. Beutler
Subject: IGS Workshop, GB Meeting 1997 in Pasadena

Dear colleagues,

three IGS-related events took place at JPL in Pasadena between March 12 and
March 18, 1997, namely the 1997 IGS Analysis Center Workshop (March 12-14),
a Business Meeting of the IGS Governing Board (March 15), and the IGS
Sealevel Workshop (March 17-18).

In the IGS workshop the focus was on Low Earth Orbiters (LEOs), site
specific effects, troposphere modeling aspects, the possible use of the
Russian GLONASS within the IGS and, last but not least, on burning Analysis
Issues.

Action items emerging from the preceeding Analysis Center Workshop were
discussed at the Governing Board Meeting.

The main purpose of the IGS Sealevel Workshop was to set up an interface
with the GLOSS-community for the establishment of a systematic permanent
tide gauge survey using the GPS.

Below, we give a brief summary of the first two events. The sea level
workshop will be dealt with in a separate e-mail by Ruth Neilan (due to
other obligations I was not able to attend the latter workshop).


The 1997 IGS Analysis Center Workshop
*************************************

Low Earth Orbiters
------------------
The first day (March 12) of the workshop was devoted to the requirements
emerging from existing and future Low Earth Orbiters (LEOs) carrying a GPS
receiver on board. Mike Watkins from JPL organized this part of the
workshop. The position paper "IGS White Paper on Low Earth Orbiting
Satellites" (authors B. Melbourne, J. Labrecque, M. Watkins, J. Dow,
and Ch. Reigber) was presented by Bill Melbourne. The paper was an
updated version of the document prepared by the same authors for the 6th
IGS Governing Board Meeting in Paris (October 1996). A first overview of
this paper was given in IGS Mail Message No. 1475. Let me briefly recall
the key issues from that message:

With a letter dated July 2, 1996 John Labrecque from NASA Headquarters (now
at JPL) asked the IGS Governing Board to review and identify its position
in the field "GPS receivers on Low Earth Orbiters".

The request was passed on in July 1996 from the Governing Board to the
(already existing) working group consisting of Bill Melbourne, John Dow and
Chris Reigber asking it to come up with a position paper for the 6th
Governing Board Meeting. This team, together with Mike Watkins and John
Labrecque came up with a document called "IGS White Paper on Low Earth
Orbiting GPS".

The white paper concluded by asking the IGS to consider:

1. Broaden the participation within the IGS and its Governing Board to
   include atmospheric and navigation agencies and institutions ...
2. Encourage the enhancement of the IGS constituent facilities including
   the IGS Global Network, the Analysis Centers and the Data Centers  to
   provide optimum support to spacebased applications.
3. Encourage the participation of the various groups developing satellites
   to carry the appropriate GPS receiver hardware.
4. Encourage the development of GPS occultation science through workshops,
   the development of standards, data exchange formats, data policy, etc.

The position paper was followed by presentations addressing user
requirements. It became clear that all LEO analysts wish to use IGS
products, in particular IGS orbits. Depending on the analysis method
the analysts further need EITHER precise satellite clocks (with a high time
resolution) OR ground-based troposphere estimates for the tracking
stations they use. The first requirement has to be met for analyses using
so-called "single point positioning methods", the second for analyses based
on the "double-difference approach".

It became quite clear that the time of availability for IGS products is a
critical issue for some LEO applications (e.g., of GPS-MET type). Good
orbits should be available in almost real-time. It is obvious that the
refinement of IGS predictions and a further reduction of the deadlines for
the IGS rapid orbits (now available after 24 hours) are vital in this
context.

A high time resolution is required at least from a subset of IGS stations.
It was not clear, on the other hand, what high time resolution actually
means (something between 1 sec and 15 sec to get around SA-induced clock
errors).

An additional issue is the organization of the data flow from the LEO
GPS receivers (data transfer, definition of user-friendly formats,
directory structure) to the IGS data centers.

At the end of the day it was clear that the IGS is very well suited and
willing to supply

 - regular 30s data from its tracking network,
 - data from a subset of this net with high time resolution in
   (almost) real time,
 - IGS orbits and (possibly) consistent satellite clocks,
 - troposphere estimates from ground-based receivers.

It will not be trivial to meet even these "modest" requirements. It seemed
advisable to form a working group to look into these problem areas. It is
worthwhile pointing out, however, that a consensus could be reached by the
workshop participants that the IGS should get involved in LEO activities at
least on this level. In return, the IGS participants (in particular the
IGS Analysis Centers) will have access to the GPS observables of the
space-borne recivers.

It was not so clear what policy the IGS should follow in analysing these
data. Obviously, some of the IGS Analysis Centers will, so to speak,
automatically get involved in processing LEO GPS data. Obviouly such
activities may be considered IGS activities, or they may be considered
activities outside the IGS. There are good arguments for both approaches.
The issue certainly will have to be addressed in detail by the IGS
Governing Board at future meetings and it is not difficult to predict that
the issue will be somewhat controversial. Should the former approach be
chosen, recommendations (1) and (4) of the "position paper" would have to
be considered very seriously.

Site Specific Issues
--------------------
Yehuda Bock from Scripps Institution of Oceanography was convening
this part of the workshop. The position paper by Jan Johannsen and Jim
Zumberge (presented by Jan Johansson) made it clear that monumentation,
homogeneity of equipment, and special devices like radoms to protect
antennas in problematic climatic conditions are crucial for the subsequent
analysis. Unannounced changes may lead to erroneous interpretation. It
became clear that the Central Bureau Information System plays a vital role
in this context, provided the information is always updated.

The IGS network is, and probably always will be, a multi-receiver and
multi-antenna network. It was encouraging to hear that through the adoption
of the elevation dependent phase center variations file by the IGS, as
prepared by Markus Rothacher, Gerry Mader, and others the negative effects
could be reduced to the sub-centimeter level. The discrepancy between the
adopted variations (which are based on "in situ" GPS measurements) and the
variations deduced from anechoic chamber measurements is not yet
understood.

If antenna phase center variations (and the mapping function) would be
perfectly known, the elevation cut-off angle used in processing would not
matter. Essentially the same coordinates would be expected for all cut-off
angles, the rms is expected to vary with the cut-off angle (increasing with
increasing cut-off). That we are far from this ideal situation emerged from
Arthur Neill's presentation. This issue is not treated homogeneously in
the IGS:  some analysis centers use 15, other 20 degrees, other 15 degrees
with an elevation dependent weighting. There is no obvious recommendation
to improve the situation.

Stick Ware showed that with the technique of pointed water vapour
radiometers (WVRs) it is possible today to significantly improve the GPS
derived heights. His results suggest to organize a coordinated  WVR
campaign in a part of the IGS network.

We are now looking back to about five years of IGS products. Unnecessary to
say that lots of information may be extracted from these time series. This
is true in particular for station coordinates. It is possible to
distinguish between "good" and "bad" stations. It is also possible to check
whether or not the residuals in the station coordinates (after removing the
station velocities) are random or whether there are correlations in time.
There are clear indications (Hadley Johnson, Simon Williams) that "colored
noise" exists in coordinate time series, at least in some cases, and that
monumentation, multipath, and the receiver environment in particular play
an essential role.

A serious site-related issue was raised by Tim Springer. It is well known
that the IGS Analysis Centers made the decision to use the ITRF coordinates
and velocities of 13 IGS global sites (namely ALGO, FAIR, GOLD, HART, KOKB,
KOSG, MADR, SANT, TIDB, TROM, WTZR, YAR1, YELL) in their daily solutions
to realize the ITRF in the daily products. This led to very consistent
results in the past. Unfortunately more and more hardware-related problems
were encountered at these stations in recent years. Tim Springer
demonstrated that four of these thirteen sites should actually no longer be
used for reference frame definition because of unpredictable behaviour of
the equipment. The issue is alarming because a further development in this
direction would inevitably lead to a deterioration of the IGS realization
of the ITRF reference frame.

Jan Kouba, in his presentation (given on behalf of the EMR AC) conveyed
three main points, namely that for most IGS stations, due to several
hardware (i.e. antenna) and offset changes, we do not have a continuous
five year series, instead we have several, much shorter segments, which
significantly limits any time series analyses. The second related point,
that IGS should develop and strictly enforce guidelines for such
hardware/offset changes (e.g. mandating overlapping observation/test
periods). The last point was that neglected or incorrect ocean loading
modeling likely causes about 2 week and longer periodical effects at about
a 2 mm level for some stations. It was suggested that this could be due to
apparent station coordinate rates induced by non zero 24h means of ocean
loading displacement rates, which can reach up to 20mm/d horizontally and
up to 60mm/d vertically with 14 day and longer beat periods.

The discussion towards the end of the day made it clear that the issue
raised by Tim Springer was considered of vital importance for the IGS. It
was recommended to set up a working group to look into that problem and
to come up with a proposal to improve the situation.

Troposphere Issues
------------------
In the report of the IGS Analysis Center Workshop 1996 (IGS Mail, Message
1266) two action items were recommended:

(1) MET stations of a defined high quality should be deployed -- at least
in a part of the IGS network. Met information already available at the
stations or becoming available in the near future shall be sent routinely
in Met RINEX files to the IGS data centers, where they will be available
for scientific purpose.

(2) IGS tropospheric delay estimates should be studied and combined by
special associate analysis centers. GFZ promised to build up such a
center.

RINEX-MET files for a limited number of stations are avaliable today. It is
hoped that this number will increase considerably in the near future. Met
equipment of sufficient quality is also available today and should be
deployed in a significant portion of the IGS network.

Gerd Gendt showed troposphere combinations based on the majority of IGS
Analysis Centers. Troposphere combinations are now regularly performed by
GFZ. It is expected that an official product (documented by a weekly
report) will become available through GFZ starting mid 1997.

Azimuth dependent troposphere models were studied by Yoaz Bar-Sever from
JPL. The results are of preliminary nature and seem to indicate a slight
improvement in station coordinate repeatabilities.

GLONASS
-------
Yehuda Bock was convening this session, too. Ruth Neilan gave an overview
of the Russian GLONASS. At present 19 satellites are available. They are
orbiting the Earth in 3 orbital planes, which are separated by 120 degrees
in the equator and inclined by 64.8 degrees. Very few receivers of
interest for IGS-type analyses are available today, a few more will become
available in the near future.

Gerhard Beutler pointed out that GLONASS is of interest to the IGS (better
coverage in high latitudes, better sampling of the atmosphere). The user
community would undoubtedly benefit from IGS orbits for the GLONASS
satellites -- which would remove all reference frame inconsistencies.
The GLONASS phase observable is available in L1 and L2 (which differ from
satellite to satellite and are well separated from the corresponding GPS
carriers). Theoretical considerations concerning ambiguity resolution for
the GLONASS phase seem to work in practice.

First GLONASS campaigns (organized by DLR, IfAG, IfEN in Germany) were
organized in 1996. Results based on the phase observable prove that
GPS-like results should be achievable in the near future.

Werner Gurtner developed the necessary generalization of the RINEX format
to accomodate GLONASS observations. So, in principle, we are ready. In
summary it may be said that the "writing is on the wall":  GLONASS will be
an issue for the IGS as soon as good dual band equipment is available in a
subset of the IGS network.

Analysis Center Issues
----------------------
This session was held, so to speak, "on public demand". Jan Kouba was the
convenor of this session. Yves Mireault from NRCan reviewed the IGS
orbit/clock/eop combination as they are in place since 30 June 1996. He
then presented his approach to come up with a predicted combined IGS orbit.
He could show that the predicted orbits are more reliable, and even
somewhat more accurate. The submission deadlines for the rapid products
were discussed. It seemed that a reduction from 23 to 21 hours would be
much more convenient for the combination.

The performance of the three GNAAC (G-SNX) station combination solutions
was reviewed by J. Kouba as well a new summary report, prepared for
IGSREPORT distribution, was introduced. The report intercompares weekly the
three G-SNX combined solutions, the station rms and geocentre offset /scale
agreement with ITRF94, as well as any SITE information conflicts amongst
the three SINEX solutions. All G-SNX combinations appear to be of equally
high quality and are consistent at about 2-3mm horizontally and less than
5mm vertically. A relatively large number of SITE information conflicts was
noticed and the IGS CB was urged to help to resolve these and to establish
an authoritative electronicly readable station log with all the pertinent
SINEX SITE information.

The workshop was concluded with a few words by the IGS chairman thanking
the IGS Central Bureau for perfectly organizing and sponsoring the 1997 IGS
Analysis Center Workshop.

IGS Governing Board Meeting
***************************

The meeting took place on Saturday, March 15, 1996 in the San Martino room
of the Doubletree hotel. The meeting started at 9:15 a.m. and lasted till
1:15 p.m. We will only summarize the highlights of this meeting.

In the Central Bureau (CB) report Ruth Neilan, director of the Central
Bureau, explained some changes in the Central Bureau organization that are
about to take place in the near future. She pointed out in particular,
that UNAVCO will be responsible for many network-related CB activities
(quality control and monitoring) as a contractor for the Central Bureau in
future. These changes will be announced prior to implementation in more
detail in an IGS mail issued by the Central Bureau.

Ruth Neilan could also announce that the new issue of the IGS Colleague
Dictionary will be mailed in the near future.

The IGS project "densification of the ITRF through regional GPS Analyses"
was reviewed. It was decided that a small group (Jan Kouba, Geoff Blewitt,
Claude Boucher) should come up with a report, to be presented at the next
Governing Board Meeting in Rio de Janeiro, in order to formally terminate
the Pilot phase and to come up with an official product. This report of
course has to be prepared in close cooperation with the (associate)
analysis centers active in the pilot project.

It was then decided to formally establish a working group "LEO Missions and
the IGS" chaired by Mike Watkins (JPL) to come up with an action plan for
the IGS. Mike Watkins (chair), John Dow(ESA), Bob Schutz (UTX), Chris
Rocken (UCAR), a representative from GFZ/DLR (t.b.d.), one from CDDIS
(t.b.d.), and Ruth Neilan (IGS CB) were appointed as members of the working
group. A few more names will be added to the working group later on.
The working group will come up with an action plan at the next Governing
Board meeting (Rio de Janeiro, September 1996).

It was acknowledged that there are quite a few site specific problems
which affect the quality of the Global IGS network at present (see
summary of the workshop). It was agreed that a working group should study
the issue of maintaining and/or improving the technical standard
(equipment, data transfer, monumentation, etc.) of the Global IGS Network.
It was not so clear what the actual task of the working group should be.
There were arguments that the working group should review all components of
the IGS (Central Bureau, Analysis Centers, AC Coordinator, Data Centers,
Network, Governing Board). Whereas it was quite clear that such a general
review will be necessary at some point in future, it was also clear that
a working group with such far-reaching goals could not be established "so
to speak" in an ad hoc way without preceeding broad discussion within the
Governing Board. It was then clear that a small working group with the
limited charter of establishing a report on "improving the performance of
the Global IGS Network (receiver/data transfer/station maintenance)"
containing a list of recommendations for the next Governing Board Meeting
would make sense. No consensus concerning the composition of the Working
Group could be reached at the Governing Board Meeting. The working group
thus will be set up by the chairman through e-mail in the near future.

In IGS-message No. 1525 Yuki Hatanaka announced in January 1997 that his
new data compression algorithm is available for extensive tests. The
algorithm should considerably reduce the data volume (about a factor of 2
compared to the algorithms used now within the IGS). Werner Gurtner and
Yuki Hatanaka were given the task to organize the necessary tests leading
to the use of the new compression algorithms within the IGS in the near
future. It is hoped that this team will give a short report concerning the
successful implementation of the new algorithms within the IGS in the near
future.

Two proposals for the next Analysis Center Workshop (spring 1998) were
discussed: Chris Reigber sent in a written proposal to host a workshop
focusing on LEO aspects at the DLR facilities near Munich, John Dow
repeated his offer to host an Analysis Center Workshop in Darmstadt.
In view of the fact that both potential hosts were absent at the time the
issue was discussed, it was decided that Ruth Neilan should try to find a
solution in direct negociations with the two mentioned potential hosts.

The GLONASS issue was only briefly discussed at the meeting. It seemed
clear, however, that the IGS should become active as soon as GLONASS
receivers will be available at a significant number of IGS stations.
In this context Werner Gurtner was asked to prepare the data flow within
the IGS for GLONASS data gathered within the IGS network.

The meeting was closed by thanking Ruth Neilan and her co-workers from the
Central Bureau for organizing the meeting.

With the best wishes for the Easter-Holiday I remain

                                                    Yours sincerely

                                                    Gerhard Beutler
                                               Chair, IGS Governing Board




[Mailed From: Gerhard Beutler <BEUTLER at aiub.unibe.ch>]