[time-nuts] The TAI zero epoch of 1958

[Steve Allen]

On Fri 2019-10-25T05:12:54-0700 jimlux hath writ:
> So why was 1958-01-01 chosen as an epoch time for TAI (or is it, really,
> maybe TAI defines the time scale, and you use yyyy-mm-dd hh:mm:ss.sssss.....
> and it's only some other standard that defines what "zero" means when you
> want to represent it as a single "number"
>
> for example CCSDS 301.0-B-4 "Time Codes"  says:
> "The CCSDS-Recommended epoch is that of 1958 January 1 (TAI) and the
> recommended time unit is the second, using TAI as reference time scale, for
> use as a level 1 time code. This time code is not UTC-based and leap-second
> corrections do not apply. "
>
> https://public.ccsds.org/Pubs/301x0b4e1.pdf

In the full context the exact wording of the 1971 CGPM approval to
create TAI, and other recorded decisions, is important.
https://www.bipm.org/en/CGPM/db/14/1/
and also resolution 2 gloss over recorded discussions that BIH was the
place that could best implement TAI, but that BIH by itself did not
have appropriate international standing to administer TAI, in part
because despite its name 80% to 90% of BIH funding was slush money
from the director of Observatoire de Paris, thus really France.

Among various contributions to the meetings in the late 1960s are some
that express serious concerns that the time scale which became TAI did
not have a rigorous definition.  I have found no recorded response to
these suggestions, just silence.  Everyone in the time services knew
that BIH had re-set all their atomic time scales in 1964 using the
epoch 1961.  They knew the basis of UT2 was shifted as of 1962
https://www.ucolick.org/~sla/leapsecs/seasonal.html
because of the new defintion of UT2 and the change from FK3 to FK4.
They knew that the origin of A3, and thus TAI, was offset from the
current notion of what the value of time should have been.

My impression is that the lack of recorded responses is partly due to
not having any way to fund BIH in an "appropriate" fashion, not having
any way to create a replacement for BIH which had appropriate
foundations, and not wanting to create yet another atomic time scale
effectively the same as TAI but, by basis of a new, rigorous
definition, offset from what everyone had been using in their
communications and publications to that point.

I think also that nobody wanted to clearly document the changes that
happened at 1961 and 1962 because doing that would have emphasized
that the value of atomic time scales is entirely arbitrary, not
connected with any event that is observable, and not connected with
calendar days of earth rotation which were and are the legal and
commonly understood basis of civil time.  So my impression was that
they allowed the origin of TAI to be said to be about 1958 as part of
trying to avoid talking about how exactly that came about.

The switch from old-1960s-rubber-second UTC to new-1972-leap-second
UTC was in part motivated by interpretations of a new law in Germany
that old UTC was no longer legal to brodcast.  So I think the lack of
clear documentation was trying to avoid letting any more national
bureaucrats who controlled the funding of other national time labs
raise questions about whether this new no-longer-related-to-days time
scale could be the basis of legal time in their country.

Looking carefully at what the practitioners of time did shows that the
USNO decided not to use new-UTC in its navigational broadcasts.  In
the USNO time service publications Winkler announced that the USNO
navigational broadcasts would become an offset from TAI (and that is
eventually what GPS, Galileo, and all the other satnav systems also
chose).  Therein lies the basis of the LORAN offset question that
prompted this thread.

Similarly the astronomers did not use UTC in tabulations of any
almanac.  They continued to use just plain UT for things on earth, and
ET (now relativisticly TT, or TCB, or TDB) for things not on earth.

The principal printed appearance of new-UTC was in decrees from
bureaucrats who were protected from seeing that the pracitioners of
time did not agree that new-UTC was the best thing to use for
technical purposes.

--
Steve Allen                    <sla at ucolick.org>              WGS-84 (GPS)
UCO/Lick Observatory--ISB 260  Natural Sciences II, Room 165  Lat  +36.99855
1156 High Street               Voice: +1 831 459 3046         Lng -122.06015
Santa Cruz, CA 95064           https://www.ucolick.org/~sla/  Hgt +250 m