I'm glad to see the basic issues (finally) get some discussion.
I disagree that the civil time issues don't matter. We hold the
standards that underlie civil time in trust for an incredibly large
and diverse set of communities. And yet, for over three years this
ad hoc process of re-evaluating the UTC standard has focused on
nothing but technical issues - rather naive technical issues at that.
The confusion between secular and period effects is embarrassing.
UTC has proven a successful and robust standard. The time and frequency
community have proven to be able stewards of the standard - right up to
the point of trying to abandon the standard by the side of the road.
The current standard already permits issuing leap seconds at a more
frequent cadence that will allow the standard to persist for many
hundreds of years, at which point a graceful update will be possible
(see, e.g.,
ftp://gemini.tuc.noao.edu/pub/seaman/leap/leapsec.txt).
There is no reason this process has to be hurried. The haste and
the opaque nature of the decision making process have been offensive.
A few talking points:
0) Who owns the civil time standards? We are to understand that at
some point in the last few decades that control of this passed from the
astronomical community to some obscure telecommunications commission.
I don't see it that way. Each individual government obviously has
ultimate control of their own civil timescales. For many purposes
it has proven critical - for commerce and safety, for instance - to
rely on an international standard. The astronomical community can
point back to hundreds or thousands of years of careful stewardship
of the world's clocks. Implicit in the UTC standard is the idea that
civil time will continue to track the Earth's rotation. (The standard
actually explicitly states that UTC should be regarded as the general
equivalent of GMT.) There isn't the slightest notion hidden in the
standard that this role could be abandoned by simple committee fiat.
The "users" we serve may, on occasion, be annoyed by leap seconds -
that does not mean that were they to understand the full implications
that they themselves would choose to discard this fundamental
requirement of civil time. Any mandate to reengineer UTC requires
a far higher level of consensus than other standards discussions,
a notoriously excruciating process under the best of circumstances.
1) There is no coherent proposal on the table, just a notion that
the practice of issuing leap seconds should cease - perhaps as soon
as immediately. How are we to comment on a plan that doesn't exist?
2) The Keck Observatory estimates that it would cost $110,000 to
retrofit their systems to support their guess at what the new UTC
standard might be. This estimate seems quite low to me, considering
that Keck observing time is valued at around $50,000 per night. Just
verifying the changes to all affected observing modes could require
several nights. Conservatively multiply this estimate by perhaps a
hundred for the astronomical community as a whole (the costs will scale
closer to the number of telescopes than their sizes - even ignoring all
non-telescope related astronomical applications). It would be much
more expensive than Y2K remediation was for our community - we wouldn't
have to just fix bugs, but rather reimplement algorithms that are
fundamental to many of our activities.
3) While astronomy perhaps requires special consideration, many other
technical fields would likely be affected - in addition to the rather
strange collection of ham-handed projects that have appeared to be
driving this initiative. Assume, for the sake of argument, that all the
effects in fields from navigation to time keeping (surely we must know
how this will affect our current GPS, WWV and NTP hardware and software,
right?) to geography (disconnect UTC from GMT - what happens to the
prime meridian?) to transportation (have we talked to our friends in
air traffic control?) would be minor. But how much will verifying all
these minor changes cost? Y2K remediation has cost hundreds of billions
of $US - are we really to believe that a change to UTC would cost not
much more than pocket change?
4) What about non-technical fields? The awareness of this issue in
astronomy is currently at the few percent level. That's your best
case of informing the affected communities. Has anybody even thought
to ask a lawyer about potential issues? And not Jacoby and Meyers,
more like Lawrence Tribe. We've learned that civil time in the U.K.
is still referenced to GMT (and why wouldn't it be?). Presumably the
relationship of civil time to UTC in many other countries is also
cloudy. Should we change the standard before or after figuring out
this situation? Can the equivalence between UTC and GMT be broken
before *all* the world's countries sign off on the change? Alternately,
would neighboring countries really like to return to the days before
the standard time zones? You would not only have to convince the U.K.
to support a drifting UTC, you would have to convince them to adopt
this emasculated standard in preference to one of the last, very
popular, remnants of the Empire.
5) What about cultural and political issues? The international
climate is incredibly tense. Many religious traditions care deeply
about calendrical issues and the timing of astronomical events. For
instance, observatories get frequent requests for lunar and solar
information from Moslems, Christians and Jews. Is it prudent to
introduce an obscure change that can be of no possible benefit in
these calculations - but that just might appear to be a deadly insult
to a quarter of the world's population? Are politicians from even a
single country on the globe likely to want to deal with issues like
these at this point in time?
6) What about commercial interests? We have a very few projects or
individuals who appear to be pushing for this change. (Although I'm
not convinced any of these individuals has even once sent a message to
this list.) Has anybody asked the manufacturers of commercial telescope,
navigation, surveying, or - oh yes - timekeeping hardware or software
what they think about it? (I'm sure that's not a complete list.)
What about the related service industries such as planetarium or
museum education? Millions of amateur astronomers would be able to
detect the discrepancy between UTC and UT1 after only a few years.
Big deal? Perhaps not, but it's a very annoying small deal
multiplied by a heck of a lot of people.
7) This is a .mil mailing list. I can think of arguments for why the
U.S. Department of Defense might either be backing this change to the
UTC standard from behind the scenes - or why they might be ignorant of
the debate, but would be completely opposed if they learned about it.
I doubt that DoD would have no opinion. Has anybody asked them? More
to the point - would the militaries or governments of other countries
be supportive or unsupportive of the U.S. DoD's desires?
To sum up: the proponents of the notion of abandoning leap seconds
have some obscure agenda of their own. This change to the standard
is a sop to lazy projects who either can't be bothered to use UTC
correctly - or shouldn't be using UTC at all, but rather, TAI or a
related timescale like GPS. In any event, many of these projects will
have died a natural death before any change to the UTC standard could
become a reality. The proponents of this initiative can't even be
bothered to generate a coherent proposal describing their plan - one
has to wonder if there is a plan at all. The change would cost the
astronomical community on order $10,000,000 and would produce zero
benefit. No attempt has even been made to estimate the cost to other
communities. Political, legal, cultural and commercial issues?
There is zero evidence that these real world concerns have been
even entertained.
The approximation that UTC = GMT is a useful one and should be
preserved. The burden for abandoning this rests with the parties
pressing for a change to the UTC standard. If there is an actual
proposal to go with this "no new leap second" notion, let's hear
it - hopefully it will be better conceived than the surveys that
have been so narrowly worded and disseminated.
Folks, this isn't just some obscure technical question.
Rob Seaman
National Optical Astronomy Observatory
Received on Mon Jan 27 2003 - 15:36:08 PST