PDA

View Full Version : Update on UTC change



PraedSt
2011-Dec-02, 07:14 AM
There was a conference in October to discuss whether or not to decouple UTC from UT1 by abolishing leap seconds. The final vote will take place in January.

Review of proceedings:

http://arxiv.org/abs/1111.7007

PraedSt
2011-Dec-02, 07:21 AM
This is the conclusion from the above paper:.

John H. Seago, Robert L. Seaman, and Steven L. Allen:

"While numerous points of view were well expressed by the contributors of this colloquium, a few points stood out in the opinion of the chairmen. Primarily, the motivations for decoupling civil timekeeping from Earth rotation are not entirely apparent, and the supposed advantages to making a change have not been shown to outweigh the supposed disadvantages both now and in the future. A coarser coupling between civil timekeeping and Earth rotation would result in adjustments of less frequency and larger magnitude, which would be more noticeable and less practical to maintain than the current system having predictable insertion points (at end of the month), constrained values (-1, 0, or 1), and a prescription for tagging events occurring during the adjustment (23:59:60). Because the existing UTC system with leap seconds is already implemented, it should be strongly preferred over alternative proposals or protocols that lack obvious advantages.

The historical tendency of timekeeping practices has been to move from empirically observed adjustments to predictable ones based on calculation. Perhaps the creation of a predictably accurate time-of-day adjustment algorithm, valid for at least a decade or more, might provide a viable alternative, but this has not been seriously studied up to the present. It appears that Earth rotation may now be predictably accurate to within one second out to three years with 99% confidence. If so, perhaps we should continue to seek advances that would improve the prediction of Earth rotation with high confidence out to one decade or better, with the goal of eventually replacing empirically predicted leap seconds with an algorithmic correction to time-of-day. Such an approach acknowledges that the coupling of civil timekeeping and Earth rotation is fundamentally an issue of calendar maintenance, and thereby satisfies public expectations of preserving astronomical relationships. Yet, because this approach may not be able to provide a sufficiently accurate realization of Earth rotation from clocks, it seems best to leave the current system in place until such time that Earth rotation models are improved, and for systems that prefer a uniform internal time scale to use IEEE 1588-2008 (Precision Time Protocol) with GNSS time or other time scale closely tied to TAI.

Finally, an outstanding question is whether “UTC without leap seconds” really represents a permanent scenario for decoupling. A persuasive argument can be made that the expectation of civil-timekeeping’s astronomical basis is so deeply ingrained in our society that a departure from it is unlikely to last forever. If so, an attempted decoupling now may be the worst of all possible options, because it creates issues for both present users reliant upon the existing system and it causes unforeseeable complications for generations to come. The want of the current decoupling proposal to address the most fundamental issues erodes its favorability within many technical communities and the public, resulting in a lack of consensus."

grapes
2011-Dec-02, 10:36 AM
So, probably no. But then a faction of non-traditionalists voted against Pluto. The romance of creating a non-Earth-based "Star Date" may be too irresistable for some.

The idea of predicting Earth rotation out beyond a second every three years has already been shown to be impossible, by the records of the last twenty years, I'd think.

PraedSt
2011-Dec-02, 09:35 PM
I hope no. I think it's a daft proposal.

NEOWatcher
2012-Jan-19, 03:20 PM
Today's the day, but I don't know what time.

Is time running out for the leap second? (http://us.cnn.com/2012/01/19/world/europe/leap-seconds/index.html?hpt=hp_t2)

I wonder if they will start the meeting at the exact second.

grapes
2012-Jan-19, 04:52 PM
Today's the day, but I don't know what time.

Is time running out for the leap second? (http://us.cnn.com/2012/01/19/world/europe/leap-seconds/index.html?hpt=hp_t2)

I wonder if they will start the meeting at the exact second.
From the article:


Without the leap seconds adjustment, the precise time measured by atomic clocks and time according to the Earth's rotation will start to deviate.

Although the tiniest inaccuracies might play havoc with our smart phones most of us would not notice the changes in our body clock -- it is estimated it will take more than 200 years to register an hour's difference.Wouldn't that be more like a minute's difference, rather than an hour's?