[LEAPSECS] Looking-glass, through

[Rob Seaman]

Apologies for a delayed reply, I'm on travel at a conference.

On Sat, 8 Jan 2011, I wrote:


>> I do not believe the unstated magic timezone notion (if indeed that is an idea motivating the authors of the draft in front of the ITU) can work (or rather, I do not believe that this notion corresponds to a solution of the problem).


...and I remain unconvinced.  Sloshing the timezones around willy-nilly by every regional government on Earth is not a solution to establishing the underlying common timescale.  In fact, the only reason that the willy-nilly timezone system can work now is the existence of a mean solar clock underneath beating at the cadence of the synodic day.

Which is to say that "working" has to be defined.  The goal isn't just to identify a procedure for setting clocks.  There are many (perhaps infinitely many) internally consistent ways to set a clock.  The goal is to match the dominant rhythm of life on Earth.  That rhythm is the synodic day.  The civil timekeeping use cases are diverse, the classes of stakeholders many, the need for clarity and transparency patent.

On Jan 9, 2011, at 12:32 PM, Tony Finch wrote:


> It really depends whether you think easy access to a sub-second accurate realisation of UT1 is part of the problem.


No.  Words like "accuracy" and "precision" have to be defined in context to have any meaning.  The focus on technical use cases and technical timescales (especially ones known only after the fact, like UT1) is a distraction from the real discussion, namely, what are the requirements for civil timekeeping?

Civil timekeeping is based on the natural (and yes, varying) synodic rotation period of the Earth.  This is called a "day".  It isn't fundamentally a question of the mechanism and schedule for occasionally resetting clocks - the fundamental issue is the clock *rate*.  It may be inconvenient that the Earth doesn't rotate at an SI-denumerable rate.  It is also the fact.


> The proposed arrangement of local timezone offsets applied to an atomic timescale only gives you access to local mean solar time accurate to couple of hours or so, which is pretty useless for astronomy or astronavigation but is good enough for civil purposes since that is what the timezone system has given us for the last 50 or 100 years.


No.  The current timezone system provides access to *Greenwich* mean solar time (under the guise of "Universal Time").  Civil timekeeping isn't about setting my particular clock to a local standard - it is about marshalling all those local standards into a single coherent global standard.  Apparent time is a red herring and local time is a red herring.

Also, there is no "proposed arrangement of local timezone offsets".  What there is, is a draft ITU document that addresses no such issues.  Meanwhile, over here in the leap second oubliette there happen to be a few guys promoting a notion that they assert (with no supporting data) fills the gap in the draft document.

   Alice laughed. 'There's no use trying,' she said:
   'one CAN'T believe impossible things.'

   'I daresay you haven't had much practice,' said the Queen.
   'When I was your age, I always did it for half-an-hour a day.
   Why, sometimes I've believed as many as six impossible things
   before breakfast.'

A few lines in an email do not correspond to a proposal.  If you believe it is trivial for the mechanism of "willy-nilly timezone sloshing" to resolve all the issues, everywhere, for everybody, then write a proper proposal laying out how this would work.  For instance, what authority will historians or lawyers consult to learn the applicable timezone offsets that were in force in some location(s) during some epoch(s) in question?

Rob Seaman
National Optical Astronomy Observatory