[IGSMAIL-5192]: UTC redefinition or change

Tue Aug 23 08:13:23 PDT 2005

******************************************************************************
IGS Electronic Mail      23 Aug 08:14:00 PDT 2005      Message Number 5192
******************************************************************************

Author: Kenneth Seidelmann (forwarded by J. Ray)

-----Original Message-----
From: Kenneth Seidelmann [mailto:pks6n at virginia.edu]
Sent: Thursday, July 21, 2005 1:55 PM
Subject: UTC redefinition or change

Dear Friends and Colleagues,

In July 2005, Daniel Gambis, Head of the Earth Orientation Center of the
IERS, sent out a notice to IERS Bulletins C and D subscribers, citing
November 2005 as possibly the last opportunity for input before a
recommendation to redefine the UTC standard is issued by the ITU-R Working
Party 7A (WP7A). [1]  He urged users to express their opinions to national
representatives of the WP7A (but unfortunately, referenced a URL that did
not list these representatives).  The situation is now confused, and I have
heard complaints from people who know nothing about this, but think they
should be informed and consulted.  Hence, I have drafted a summary of the
situation as I have come to understand it.

For those that do not know, I am a member of the US Working Party 7A
(USWP7A) that was responsible for the recent UTC leap-hour proposal.  This
proposal was drafted before my involvement with the group, and while I have
not yet been able to attend meetings in person, I have been attempting to
participate by email.  Since my inclusion I have been seeking to understand
the situation concerning the possibility of redefining UTC, the alternative
choices available beyond the status quo, the justifications for these
alternatives, the procedures to be followed for any change, their timetable,
and the status of the overall process.  These efforts have led me to
conclude that alternative proposals to the status-quo UTC lack documented
justification.

The ITU-R Study Group 7 established a Special Rapporteur Group (SRG) in 2000
to study a question for redefining UTC.  A 2003 chairman's report concluded,
"The result of [the SRG's] efforts had not produced sufficient information
for analysis of alternatives nor produced a clearly defined course of
action." [2] In preparation for the report, a colloquium was held in Torino,
Italy in May, 2003.  While public, the colloquium announcement had very
limited distribution compared to the very large population of precision UTC
users, and the opportunity for outside, contributed papers was announced
weeks before the meeting.  In contrast to the aforementioned chairman's
report, the pre-colloquium announcement noted the SRG's "work to date has
produced a consensual opinion that the SRG wants to present and discuss" in
support of "drafting a recommendation to the ITU-R." [3]  Colloquium
organizers advanced a prepared conclusion which was not wholly accepted by
the attendees; instead, attendees elected to draft a single-page summary of
finding, essentially noting:
a) No recommendation to change UTC was reached;
b) if a change were proposed, no alternative adjustments to leap seconds
(particularly, leap hours) should be advanced "understanding that sudden or
dramatic change to UTC was generally agreed as undesirable"; [4]
c) a broadcast time standard not tied to Earth rotation would be
fundamentally different from existing and historical practice, and thus
should omit reference to "Universal time" by title ("International Time" was
suggested); and
d) recommended changes, if any, should happen in the distant future (A.D.
2022 was suggested). [5]

Speculation about the danger of leap seconds to modern electronic navigation
was a prime motivator for these discussions.  However, as I understand it,
the GPS program office is not promoting a change, as GPS works with the
current UTC and a change may require programming changes for existing
receivers. Operators of the GLONASS system provided the SRG a letter (in
Russian) opposing any change to UTC. [6] The Galileo system will reference
TAI, but operators prefer a decision regarding UTC by 2007 for design
purposes.

The Director of the BIPM expressed opposition to the creation of alternative
atomic time scales that parallel TAI, noting a prior ITU Recommendation
suggesting TAI be used instead. [7]  Nevertheless, at the ITU-R Study Group
7 meeting in October 2004, the USWG7A submitted a proposal to maintain a
broadcast time standard called UTC that suppresses leap-second adjustments
(in favor of leap-hour adjustments) starting 21 December 2007.  The Torino
recommendations were not reflected in their proposal.  The USWP7A proposal
met with objection from the delegation of Great Britain and was not
accepted.

Just before the Torino colloquium, there was a telecon from the US Naval
Observatory among interested parties in the US Department of Defense (DoD).
It was concluded that participants - including some members of the SRG and
the USWP7A - could not support a change in the definition of UTC, and that
there should be a "data call" across the DoD to get the opinions and impacts
of the various commands. [8]  Dennis McCarthy, an advisor to the SRG and the
USWP7A, when recently asked about the status of the data call, wrote to me
the following: "An official DoD data call on this subject would certainly be
in order if there really were an ITU recommendation that were returned to
national agencies for response.  Currently there is no official ITU proposal
for discussion.  Without a real ITU proposal for comment it would be
unproductive to ask for a significant response on something that might
occur.  Of course if something official does come along, then we would have
to go out with another data call.  We certainly don't want to do that twice.
We have a hard enough time getting something now." [9]  Hence, the current
US proposal does not appear to represent any official position of the US
DoD.

There are unaddressed concerns about the legal requirements of broadcast
time standards versus the official time specifications of various countries.
For some nations, including the US, the specification is for "mean solar
time" from a reference meridian such as Greenwich.  According to Dr.
McCarthy, "I am unofficially aware of a UK response that is being
formulated.  It argues that, since GMT is the official time of the UK, the
adoption of any change would likely force UK to adopt UTC and that would
involve action by parliament.  Since this would be expensive the UK ITU WP
7A will not support any change.  It is interesting that some governments
have not yet gotten around to adopting UTC officially." [9]

"In the US, if the ITU did actually adopt a proposal for comment by national
agencies, the State Department would be in charge of formulating the
official US response.  Presumably they would canvas the appropriate
government agencies for comments before they would arrive at the official US
vote" according to Dr. McCarthy. [9] Hence, the current US proposal does not
appear to represent a well-formulated opinion of the US Department of State
or other appropriate government agencies.  It also appears that some
governments and international organizations are actively considering this
matter, while others are / can / will not because official status is
pending.  However, should a change be proposed by the ITU-R WG7A in November
2005 as announced, that would require international commentary, which should
take some time, then the ITU would have to approve the proposal.  Noting the
proposed effective date in 2007, even the most alert consumers of UTC would
have very little time to make necessary system changes upon official
adoption.  This is most unrealistic.

Regarding the status of the IAU Working Group on the Redefinition of UTC,
Dr. McCarthy further noted "The IAU WG is composed of those having a wide
range of opinions, and is in the process of preparing a final report for the
Prague General Assembly [August 2006].  The members are well aware of
surveys already carried out by URSI, IERS, and CRL, among others. If the
membership of the WG would like to carry out an IAU-wide survey, it could be
done. However, no one on the WG has suggested that, and in my opinion it
would produce nothing much more substantive than we already have recorded.
There may also be reluctance on the part of the IAU Executive to sanction a
Union survey on something that is not likely to excite the great majority of
the IAU members." [9] Hence, the current proposal is not based on any formal
recommendation or conclusion of the IAU or astronomers in general. However,
at least three (3) letters to the editor of the American Astronomical
Society Newsletter this past year suggests that there is interest in the
subject among affected astronomers, and the aforementioned surveys support
user satisfaction with status-quo UTC across various fields. Recently, the
WIYN site engineer estimated  their cost of $3M for the software conversion
plus a significant time delay if there is a change in UTC.

Having given this matter attention and after consulting some of those
formally involved, I regret to conclude the following:
a) There is no publicly available documentation that adequately or
consistently justifies the proposed re-definition of UTC.
b) Another atomic time standard parallel to TAI is unneeded, given there are
already parallel scales being broadcast, e.g., GPS system time.
c) The recently proposed change to UTC is not only insensitive to the
most-expressed opinions of UTC users, but also the transition timeline is
not realistic.
d) The existing UTC standard (with leap seconds) remains capable of
time-tagging events completely and with full atomic accuracy for centuries
to come, while also satisfying long-standing requirements for civil clocks
maintaining mean solar time.

Finally, under the current standard, the difference between UTC and UT1 is
just small enough so that many applications can neglect it. Because of this,
they may not rely on leap-second adjustments or Earth orientation
parameters, and for this reason, these users would not necessarily subscribe
to IERS bulletins or similar sources of information. I fear it is largely
this group of users, potentially most affected by the proposal, who are
least aware of its status.

   With best wishes.  Ken Seidelmann

P.S.  A leap second has been announced for 31 December 2005.
P.P.S. I will be out of email contact for the next few weeks while
traveling; however, I thought the substance of this message was sufficiently
important so as not to delay its distribution.

[1] http://www.ucolick.org/~sla/leapsecs/gambis.html
[2] Beard, R., (SRG Chairman), "ITU-R Special Rapporteur Group on the Future
of the UTC Time Scale," 35th Annual Precise Time and Time Interval (PTTI)
Meeting, San Diego, California, December 2-4, 2003.
http://tycho.usno.navy.mil/ptti/ptti2003/paper29b.pdf
[3] IERS Message No. 40: Colloquium on the UTC Timescale
http://www.iers.org/iers/publications/messages/message_040.txt
[4] Colloquium Report Information Paper,
http://www.ien.it/luc/cesio/itu/colloquium_report_info_paper.pdf
[5] Special Rapporteur Group 7A (SRG 7A) on the Future of the UTC Time
Scale, UTC Timescale Colloquium 28-29 May 2003 Report, p. 7,
http://www.ien.it/luc/cesio/itu/annex_a.pdf
[6] Russian Statement, http://www.ien.it/luc/cesio/itu/russian_statement.pdf
[7] Circular letter from Director of Bureau International des Poids et
Mesures (T.J. Quinn), 1999, http://www.ursi.org/A_97-99.htm
[8] McCarthy, D.D., Summary of UTC Telecon of 9 May 2003, USNO.
[9] McCarthy, D.D., personal communication, 23 June 2005.