[LEAPSECS] The relation between calendars and leap seconds.

[M. Warner Losh]

In message: <F2A54D51-65ED-42DB-8912-51D8C949F619 at noao.edu>
            Rob Seaman <seaman at noao.edu> writes:

: There is  

: absolutely nothing in the current definition of UTC to stop us from  

: announcing a schedule years in advance.  This is by far the easiest  

: change to make to UTC.


Yes.  But the times that I've proposed loosening DUT1 from +/- .9s (or
is it +/- 1.0s? today) to something like 10s so that you could account
for the long term trends of DUT1 and be able to predict out for 50
years with reasonable degrees of certainty, the details get shot down
by you, Mr. Seaman.  Or at least did in the early days of this debate.
The main practical effect of this would be that there'd be a 10x
increase in the error term for the uncorrected time, worst case, over
methods used today.  Already +/- 1s isn't anywhere near good enough to
aim larger telescopes with the precision necessary to do good science.
You need at least 1000x better numbers, which the better software
grabs off the net from published sources.  If DUT were to float, the
effect would be small.  Many of the DUT1 tables are published with
fixed columns, so the data format there would need to change...

Basically, your suggesting that the mean solar time remaining in sync
to earth's orientation be true, but with a longer averaging time than
we have today.

In reality, an extreme form of this is the leap hour.  It sets DUT1 to
be ~3000x looser than it is today, and extends the mean solar time to
be true on time scales measured in centuries rather than in terms of
decades.  There's really no difference between this and the other
proposals, except the size of the corrections that folks interested in
celestial positions will need to apply to their calculations.

None of these proposals, however, make UTC a mechanical calendar, or
UTC have a fixed radix.  It would still be an observational one, even
if the practical aspects of adjustment were made less burdensome.

At least the Gregorian calendar made the counting of the days totally
mechanical for the next 10k-100k years or so.  This is a far cry from
the counting of the seconds, which is mechanical for only on the order
of a year or two (worst case 6 months).

Warner