[tz] historic sub-second time offsets

Steve Allen sla at ucolick.org
Wed Nov 5 20:30:16 UTC 2014

On Wed 2014-11-05T11:15:47 -0800, Paul Eggert hath writ:
> Was heure definitive really UT2?

UT2 became the goal for broadcasts and the reports starting at the
beginning of 1956.  The distinction between UT0, UT1, and UT2 were
described, and the use of UT2 was prescribed at the 1956 August
General Assembly of the IAU in Dublin.  In August there were no names
for the three flavors of UT.  I think it is probably Guinot who has
written that he supposes the names of the time scales were worked out
during the next few months in correspondence between IAU Comm 31
president Markowitz and Stoyko at BIH.

> My French is pretty weak, but page
> 253 of your source has a table entitled "Temps Universel 2 - Heure
> Definitive", which makes it appear to be listing the difference
> between UT2 and heure definitive.

In the title it is in fact a hyphen, not a minus sign.  The tabulated
numbers are for TU(i) or TU2(i), where (i) is a particular station.
The difference between that and Heure Demi-Definitive was published
bi-monthly with whatever data had been provided to BIH.  For the final
Heure Definitive the differences were published annually after
gathering data from every possible contributor.

BIH was perennially under-staffed, under-funded, months to a year
arrears in publishing, and during that era all of those calculations
and plots were being done by hand.  When the algorithms changed the
new schemes were published in Bulletin Horaire, but the notations
require very close attention to be sure what they thought they were
doing.  It was a very fiddly process with unique parameters for each
data source, and the algorithms are described with terminology that is
no longer in use.

The bi-monthly series of Bulletin Horaire for 1955 Nov/Dec have tables
with the heading (and I am not making a typo here)
	Temps Uniersvel - Heure Demi-Definitive
That issue gives an indication of their strained proofreading
resources; it arrived only 6 months in arrears.

> Also, I just now looked at a
> seemingly-authoritative paper on this (see citation below), and its
> pages 177-8 seem to say that heure definitive was closer to what is
> now called UT1.

I have not dug through enough volumes to find the incept date for that
particular change, but that was true during roughly the 1940s through
1955 as the systematic offsets between what eventually became called
UT0 and UT1 had become recognized.

> It's funny: just today I was wading through the HTML5 spec, which says:
>    "Times in dates before the formation of UTC in the mid twentieth
> century must be expressed and interpreted in terms of UT1." <http://www.w3.org/TR/html5/infrastructure.html#global-dates-and-times>

As stated that is close enough to correct.  Go farther back and the
times are now better described as UT0, but the measurement errors of
the technologies contributing to the numbers which were available to
the general public get as big as the difference between UT1 and UT0.

The point is that neither the HTML5 spec nor the the tz database
should be burdened with the chaos of the way time and time signals
were handled differently from one location (or one decade) to another
prior to 1972.  There was a common goal, that goal was based on what
we know as Universal Time, and the details only dare be incorporated
into the most specialized purposes.

Steve Allen                 <sla at ucolick.org>                WGS-84 (GPS)
UCO/Lick Observatory--ISB   Natural Sciences II, Room 165    Lat  +36.99855
1156 High Street            Voice: +1 831 459 3046           Lng -122.06015
Santa Cruz, CA 95064        http://www.ucolick.org/~sla/     Hgt +250 m

More information about the tz mailing list