Talk:International Atomic Time/Archive 1

32 vs 22 seconds difference
The TAI article says that UTC is 32 seconds away from UTC, while the leap seconds article lists 22 leap seconds that have been inserted (and the one for the end of 2005). Where's the discrepency from?


 * TAI was established on 1 January 1958 to be equal to UT1. UTC was established on 1 January 1972 with an inital offset of 10 seconds from TAI to account for all variations in broadcast time between 1958 and 1972, much of it in the form of artificially lengthened seconds. During that period, the time between any two 'ticks' of broadcast time was slightly longer than one atomic second. Since 1972 broadcast time signals have had exactly one atomic second between any two ticks. In a manner of speaking, the extra length that used to be added to every second is saved up until one whole second is reached. Since 1972 22 of these whole atomic leap seconds have been inserted. This info as well as additional info should be in these articles. &mdash; Joe Kress 07:56, 12 July 2005 (UTC)

and New Zealand have whole hour plus 45 minutes offsets from UTC.
But NZ standard time is +12:00hrs from UTC. There was a half-hour shift during World War 1, similar to daylight saving, but it was permanent, year-round and thus merely offsets the NZ time zone to the longitude. (NZ is not bisected by 180 degrees) - Regards, NickyMcLean (in Wellington, NZ)


 * That paragraph is poorly worded. The New Zealand 45 minute offset refers to the time of Chatham Islands at UTC+12:45, not to New Zealand proper. Australia has a total of four half hour offsets, one in Australia proper and three for dependent islands, but no 45 minute offset. In any case, that paragraph is not germane to the article, so I'm removing it. &mdash; Joe Kress 04:51, 14 September 2005 (UTC)

Union Square NYC
Is there any source to suggest that the clock in Union Square is synced to an atomic time server somewhere? Couldn't it just be a clock?

There doesn't seem to be mention of it in the artists fact sheet.

Blckdmnd99 13:13, 25 March 2007 (UTC)
 * A simple Google search will answer the question for you, but here is one hit: http://alcorn.com/cases/metronome/index.html.  Googling is a great way to find some of these answers.  --David Shankbone 13:50, 25 March 2007 (UTC)


 * Well how about that. I hunted around a good deal but mostly on the artists site. This definitely supports the atomic bit, thanks for pointing it out. Blckdmnd99 02:29, 27 March 2007 (UTC)
 * My pleasure - it was a good question. --David Shankbone 02:40, 27 March 2007 (UTC)


 * The description states "It receives its timing reference from the atomic clock at the US Navel Observatory in Washington DC." This is largely true of most computers in the United States although there is a distinction between the civilian NIST time signal and the USNO time signal.  If this is just a computer that is synchronzing using NTP that it hardly is appropriate to include a picture to illustrate an atomic clock.  You could just as easily include a picture of my PC or a commercial "Atomic Clock".  I would recommend removal of this image from this page in favor of something more explicative value. PaigePhault 04:54, 27 March 2007 (UTC)
 * I'd like you to provide a source that most computers in the United States are synced to atomic time. Like it or not, the metronome is an atomic clock and there are more Google hits that confirm that; I only provided one.  Now you provide one showing it is not.  --David Shankbone 12:56, 27 March 2007 (UTC)
 * Well, the majority (>80%) of computers are running Windows XP and by default Windows XP enables time synchronization with NTP.  Microsoft maintains its own time server synched to NIST's atomic clock.  There were problems with this implementation in the past, but that was 6 years ago.  I am quoting from the very source you provided - it says specifically that it is a Alcorn McBride V4+ Show controller synchronized to NIST's clock in Washington DC.  So like it or not (and I don't like it) the Metronome is not an atomic clock. PaigePhault 20:11, 27 March 2007 (UTC)


 * I tend to agree with PaigePhault. The caption for the photo says "Atomic Clock overlooking Union Square, New York City" but that is incorrect; the Union Square "Metronome" is not itself an atomic clock; it merely receives its time reference from an atomic clock. But aside from that, this is an article on "International Atomic Time" (not atomic clocks). Even if the Metronome were a true atomic clock, it would not be a suitable illustration for this article because what it displays is not TAI but civil time in New York City. My suggestion would be to remove that illustration from this article, but instead insert it in the article North American Eastern Time Zone. --Mathew5000 22:32, 27 March 2007 (UTC)
 * Okey-doke. --David Shankbone 23:07, 27 March 2007 (UTC)

It's a coordinate time, not a proper time
I wish to point out the content of http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=1986CeMec..38..155G The abstract alone makes it clear that TAI is a coordinate time, not a proper time. The article itself immediately cites the CIPM approved statement by the CCDS which asserted that it is a coordinate time. Steven L Allen 07:21, 27 November 2006 (UTC)


 * I was just about to say that myself, so I will change it now. Martin Hogbin (talk) 19:22, 26 October 2008 (UTC)

Geoids and rotation
Is IAT is defined as the rate at which time passes on the geoid? So, if i have a perfect atomic clock sitting precisely on the geoid, it reads IAT? Modulo epoch, that is. Or would that be a proper time, and since IAT is coordinate time, it's different? In which case, since coordinate time is defined relative to an inertial observer, where is he? And does that mean that IAT is corrected (in some sense) for kinematic time dilation (or whatever you call it) but not gravitational time dilation? Why correct for one but not the other?

Also, the article says 'the geoid', and equates this to mean sea level, but on the talk page for Geoid, it's asserted that "there is not actually 'a' geoid, but rather many different variations depending on whether or not one accounts for tides, etc", and that it may even not include rotation. Is there a more formal specification of the geoid to which IAT is reduced?

If it does include rotation, that means that the gravitational potential isn't actually the same everywhere on the geoid, although the sum of gravitational potential and kinetic energy are (right?). Does that mean that the gravitational time dilation is not the same everywhere on the geoid? I assume not, since that would negate the whole point of defining IAT to be on the geoid. Or does the equivalence principle mean that clocks on an equipotential surface of any kind tick at the same rate? Or does that only apply if they're moving inertially, which our clocks aren't, not being in freefall?

If the equivalence principle doesn't come to the rescue, then the geoid used for IAT must surely be a non-rotating one, in which case it doesn't correspond to mean sea level, and that remark should be removed from the article (or qualified, or clarified).

Sorry if these are dumb questions. I'm a cell biologist.

-- tom —Preceding unsigned comment added by 62.56.106.202 (talk) 19:26, 3 November 2008 (UTC)

I can only attempt to answer your questions as I am not an expert on general relativity.

All physical clocks measure proper time along their own worldlines. This is not very useful if you are not near a clock. Coordinate time is a way of assigning a time coordinate to any event in spacetime. To get a time coordinate at a general event you need to have some agree convention for setting the time coordinate at that event based on a measurement take from a clock somewhere else.

I believe that clocks on an equipotential do all tick at the same rate, although how do you compare them? Regarding tides, I think that clocks are not yet accurate enough for that to matter. I have read a paper on this subject somewhere which said that as clocks get more and more accurate the definition of a sensible coordinate time will become very difficult, if not pointless. You might try a www search for this.

If you want a better answer than mine, you could try the newsgroup, sci.physics.research Martin Hogbin (talk) 20:08, 3 November 2008 (UTC)

Refimprove template
There is a Refimprove template. I propose to add citations so this can be removed. I expect many of the inline citations will be to the same sources, so I would like to change the article to use either shortened footnotes or parenthetical referencing. Is there any objection? Is there a preference? --Jc3s5h (talk) 22:53, 4 January 2010 (UTC)


 * Since no objections or preferences have been brought up, I have changed to a short footnote style, and added a number of citations. I also removed the Refimprove template. --Jc3s5h (talk) 05:46, 5 January 2010 (UTC)

Short endnotes
This article uses short endnotes, which is handy for articles that are likely to cite different pages of the same source at several points in the article. In the endnotes, notice the name, date, and page number (if any) of the source. Then, in the References section, find the source and examine the page number that has been cited. Jc3s5h (talk) 16:58, 24 May 2011 (UTC)


 * Of course. Oops. Thanks for your patience. --Kvng (talk) 22:10, 25 May 2011 (UTC)

Proposal to suppress TAI
At the end of the article as it currently stands there is discussion of a proposal to suppress TAI in the event that the leap second in UTC is abolished. Any discussion of such a proposal in this article ought to register the idea that TAI is underused, and ought to be used for precisely those purposes that the current UTC is considered unsuitable for (because of the leap second). For instance, we already have GPS time which is a constant-offset variant of TAI.

TAI ought to have been used for the GPS system in the first place. Abolishing the leap second from UTC would make a third constant-offset variant, which apparently motivates the idea of suppressing TAI (and, I suppose, GPS time as well). If we're going to discuss suppressing TAI in the event of the abolition of leap seconds from UTC in this article, perhaps we also ought to discuss the idea that those who seek to abolish leap seconds from UTC might do better to use TAI instead of UTC in the first place, since TAI is already perfectly adequate to their purposes.

--arkuat (talk) 23:59, 26 July 2012 (UTC)


 * It isn't our job to invent arguments for or against this or that time scale; our job is to report on what noteworthy sources are writing. Jc3s5h (talk) 03:04, 27 July 2012 (UTC)

Coordinate vs. proper time in lead
I believe the mention of coordinate time and proper time in the lead should be followed in the article body by an explanation in the article body that can be understood by someone who who has not studied general relativity. It is especially confusing to people who are new to the concept of multiple time scales because of the term "Coordinated" in the name "Coordinated Universal Time", which has an entirely different meaning. I think I might be able to come up with a simplified explanation, but I don't know where to find references that would support whatever I might come up with. Jc3s5h (talk) 13:25, 29 May 2011 (UTC)
 * I agree. Why not add something anyway, references could be added later. Martin Hogbin (talk) 09:40, 27 July 2012 (UTC)