An earlier thread in LEAPSECS raised the possibility of correcting the
eventual full day discrepancy between TI and UT1 by modifying the
Gregorian calendar. I have checked with the local expert on the
stability of the solar system and he also handed me a relevant book.
The bottom line on the stability of the solar system is that we can
believe the ephemerides that are calculated 4000 years into the
future to an accuracy of 1/2 degree in the longitude of the earth.
Indeed, almost anything imaginable that could cause a secular drift
that large in 10000 years should already be detectable.
This does not, however, automatically extend to knowing the longitude
of the equinox, for the value of the precession constant is more
uncertain than the secular terms in the longitude.
The relevant book is Marking Time by Duncan Steel (John Wiley & Sons,
2000), and it is about the development of the calendar. Regarding the
calendar, Steel is as pedantic as many contributors to LEAPSECS. He
delves into many of the social and political ramifications that have
accompanied calendrical, and time keeping, reforms. He posits that a
33-year cycle of leaps invented during the reign of Elizabeth I, and
which matches the length of the year even better than the Gregorian
calendar, was used as a Protestant ploy to refute the Pope and
motivate the settlement of Virginia (and partially explain the mystery
of the Lost Colony of Roanoke).
But back to to the point of LEAPSECS, Steel also points out that
Herschel's modification of the Gregorian scheme, which omits a leap
year every 4000 years, is not reasonable. Part of the
unreasonableness is due to the uncertainty of earth rotation. On page
185 Steel writes emphatically:
Trying to design calendars for use on timescales of more than a
millennium makes no sense astronomically, logically, or
sociologically.
In Chapter 18 Steel covers the adoption of standard time, and resistance
to it, along with the International Meridian Conference.
Chapter 22 considers the tidal deceleration of earth rotation in
combination with the secular behavior of the earth's orbit.
Appendix B delves even more deeply into the meaning of the length of
the year and the design of calendars. Steel goes so far as to
demand that textbooks of the future not identify the tropical year
as the mean time between vernal equinoxes, for those are different.
He amplifies his earlier warning:
Of one thing you can be sure: anyone trying to design calendars
for use beyond the next millennium has no understanding of the
cycles (and variability) of the heavens. And the same applies to
anyone making comments about the accuracy (or otherwise) of the
Gregorian or any other calendar, unless they acknowledge
_precisely_ which year is appropriate for that calendar, and do
not confuse the tropical year with the VE [vernal equinox] year.
Steel's treatment of history covers many instances of human reaction
to authorities messing around with time. It almost seems that the
only thing about responses to changes of time that he has missed is
the subject of this LEAPSECS list.
So, I again plead that the LEAPSECS list (and the ITU) avoid bringing
the calendar into the question. It is sufficient that any change to
civil time provide some mechanism which remedies the inevitable
deviation between atomic time and universal time. Anything more than
that is an issue best left to posterity.
--
Steve Allen UCO/Lick Observatory Santa Cruz, CA 95064
sla_at_ucolick.org Voice: +1 831 459 3046 http://www.ucolick.org/~sla
PGP: 1024/E46978C5 F6 78 D1 10 62 94 8F 2E 49 89 0E FE 26 B4 14 93
Received on Wed Jul 30 2003 - 00:54:37 PDT