communique from the leap second wars

Summary: We should make plans now about how FITS would support UTC
should the issuance of leap seconds be halted.

Straw proposal: Replace all instances of "UTC" with "UT1" in the FITS
standard. Begin working on a more extensive proposal for generalized
time handling in keywords and tables.

Discussion: For the past five years (beginning before Y2K), a group of
representatives of the precision timing community have been attempting
to halt the issuance of leap seconds. The effect of this would be to
completely disconnect civil time, which would remain UTC in name only,
from the rotation of the Earth. I won't belabor the implications of
this except to note that a wide range of astronomical software and
systems would have to be rewritten simply to preserve current
functionality, just as with Y2K. The latest attempt is to cast this as
a replacement of leap seconds issued roughly every 18 months, by leap
hours issued roughly every 600 years (the effect is quadratic). One
has a hard time imagining how such a standard would be enforced -
Universal Time would simply cease to exist as we know it.

I've appended the retroactive announcement that such a vote occurred
(and the implication that they'll try again next year). If you think
the IAU governs this decision, think again. There is no indication
that the draft standard will ever be released for comment to the (many)
interested parties, but a bootleg version is available from
http://www.fcc.gov/ib/sand/irb/werit...1893wp7a/1.doc.

Steve Allen of Lick Observatory provides an excellent web resource on
the issues:

http://www.ucolick.org/~sla/leapsecs

My own take on the subject is available from:
http://iraf.noao.edu/~seaman/leap

Rob Seaman
National Optical Astronomy Observatory
---

From:
Subject: [LEAPSECS] ITU Meeting last year
Date: January 19, 2005 1:19:42 PM MST
To:

This is a very brief description of what happened at last October's ITU
meeting in Geneva.

A resolution was proposed to redefine UTC by replacing leap seconds by
leap hours, effective at a specific date which I believe was something
like 2020.

This proposal was not passed, but remains under active consideration.
Presumably something like it will be considered next year. My quick
computation indicates that, should this proposal be adopted, it would
take about a century for UT1-UTC to diverge by one minute, and many
centuries before a leap-hour would be called for.

I did not attend the meeting, and this is all I know. I was told the
ITU web pages had essentially this same information in them, but could
not find anything there with their search engine.

On Fri 2005-01-21T08:11:52 -0800, hath writ:
Summary: We should make plans now about how FITS would support UTC
should the issuance of leap seconds be halted.

Rob Seaman suggests preparing to redo the FITS documents to replace
UTC with UT1. The original motivation for having UTC be the default
time scale in FITS was that UTC is the most practically available time
scale. For practical purposes it is my impression that most things
writing FITS files will not actually have access to a suitably
accurate value of UT1.

It has been opined that FITS and VO are destined to merge. In some
cases the VO has been following FITS, but in this case I think that
Arnold Rots has taken the VO to an appropriate solution first with his
Space-Time Coordinate Metadata for the Virtual Observatory document as
seen at
http://www.ivoa.net/Documents/latest/STC.html
When it comes time to have a FITS WCS Paper V (VI? VII?) with a title
like "Representations of temporal coordinates in FITS" I see no
reason to expect that it will not follow the lead of the Virtual
Observatory document.

I expect that FITS will continue to specify UTC as the default,
because (despite many valid objections) that will probably remain the
name of the time scale most practically available to most writers of
FITS files.
For unambigous (but not precise) historical purposes I expect that
UT will remain the best time scale.
For unambiguous precision purposes of observations near earth
I expect that Arnold Rots is right to have TT be the default.

--
Steve Allen UCO/Lick Observatory Santa Cruz, CA 95064
Voice: +1 831 459 3046 http://www.ucolick.org/~sla
PGP: 1024/E46978C5 F6 78 D1 10 62 94 8F 2E 49 89 0E FE 26 B4 14 93