[LEAPSECS] presentations from AAS Future of Time sessions

[Rob Seaman]

On Jan 14, 2014, at 8:41 AM, Warner Losh <imp at bsdimp.com> wrote:


> On Jan 14, 2014, at 8:33 AM, Rob Seaman wrote:

> 

>> On Jan 14, 2014, at 3:48 AM, Poul-Henning Kamp <phk at phk.freebsd.dk> wrote:

>> 

>>> In message <20140114103334.GV21945 at fysh.org>, Zefram writes:

>>> 

>>>> It's dubious to say that they meant UTC if they weren't aware of

>>>> leap seconds.  As that's the defining feature of UTC [...]

>>> 

>>> No.

>>> 

>>> The defining feature of UTC is the bit they put in the name:  Coordinated.

>> 

>> There's also the other two parts of the name, Universal Time:

>> 

>> 	"The terms Greenwich Civil Time (G.C.T.), Weltzeit (W.Z.) and Universal Time (U.T.)  indicate time computed from Greenwich mean midnight without ambiguity."  (http://iau.org/static/resolutions/IAU1928_French.pdf)

>> 

>> UTC is subclassed from UT.  Leap seconds are necessary for that to be true. Coordination is a specific characteristic, but UT is the general class.

> 

> That's one model. One could also model it as a multiple inheritance that derives from 'coordinate time' and 'universal time.’


Are you making a statement about "coordinate time" versus "proper time”?  Or is this a typo for “coordinated time”?

We’re dangerously close to dragging UML diagrams back in here :-)  The term “universal time” as a synonym for “Greenwich mean time” predates modern coordination activities based on atomic clocks.


> Leaps are necessary to marry the two, but it is possible to have coordinated time without it being a universal time.


Sure.  But then don’t call it universal time.  The 2003 Torino meeting resulted in the suggestion of “TI” for International Time, but this could be “TIC” for Coordinated International Time, if you prefer.


> Different users have different needs and would emphasize one over the other.


As they should, and they can choose from the diverse kinds of timescales that already exist.


>>> To everybody else but the scientists who tickled the atomic clocks,

>>> leap seconds was an academic detail of no consequence.

>> 

>> Technology, standards and protocols are often esoteric and ticklish in their details.  That does not imply they are of no consequence to non-academics.  Leap seconds are a means to an end.

>> 

>> It is a simple fact that "day" means "synodic day" on Earth and a couple of dozen other large terrestrial worlds in the solar system.  This ties UTC to Greenwich mean midnight.  A timescale that omits that connection should not be denoted Universal Time of any kind, coordinated or not.

> 

> Day means what the technical committees define it to mean.


Well, no, but a piece missing from the international standards-making is indeed a recognition that redefining UTC would have implications for other terms, in particular “day”.  Other than perhaps ISO TC 37 there doesn’t appear to be anybody worrying about the contingent impacts.


> "day" to a layman may mean "when it is light out" or "one diurnal cycle." Both of these definitions are fuzzy. Synodic day is one way to address this fuzziness, but isn't the only way.


Day, as with many words, has multiple definitions, e.g., day time versus night time.  Day as in the unit of our calendars is close in meaning to “one diurnal cycle” and is very much tied to synodic timekeeping.  There is one fewer day per year than there are sidereal rotation periods, whatever complexity exists in the metrology of Earth rotation or orbital mechanics.

Whatever else is true, a proposal that would reset every (civil) clock on Earth shouldn’t be fuzzy.

Rob