Re: [LEAPSECS] Upgrade, don't degrade

From: Ken Pizzini <ken_at_halcyon.com>
Date: Mon, 09 Apr 2001 21:20:06 -0700

Rob Seaman <seaman_at_NOAO.EDU> wrote:
> SYNCHRONIZING CLOCKS
>
> What is it that we are really trying to do? A lot of technical jargon is
> disguising a very simple need - the need to keep two clocks synchronized.

Where does this need come from? To me this is a basic "requirement"
which needs to be questioned. While Mr. Seaman's analysis of
improving UTC's agreement with UT1 by increasing the frequency of
leap second samplings up to the current standard's permitted
once-per-month interval is quite compelling if such agreement is the
primary goal, I argue that you can get even better agreement by
using UT1 directly.

I will sheepishly admit that I am ignorant of the details of the
problem(s) that leap seconds were designed to fix; I look forward to
reading a good technical history of how UTC came to be. But for now
I remain perplexed as to why UTC is considered to be useful: if one
wants a precision, uniform, unsegmented time scale, then TAI suits
this need, and has been defined even longer than UTC has been; if
instead one wants a timescale based on the relation of a fixed
position on earth's surface to the mean sun, then UT1 is a useful
time base, and the noise of (UTC-UT1) is just an additional error
that has to be taken into account if one uses UTC for precision
work that depends on the earth's orientation in space.

Leap seconds are just the technical mechanism chosen to implement
UTC. If there were agreement to let UT1 and TAI each track their
own time bases (the one varying, the other our best attempt at
uniform), and if there was a willingness to widely communicate
accurate values of each to the world, then is there any residual
need for UTC and leap seconds?


Economically, Rob Seaman's suggestion surely makes more sense than
killing UTC outright, since it is an enhancement of the status quo
rather than a drastic change, but throughout this discussion I have
still yet to read an explanation of why UTC itself is of value.
(Furthermore, as I brought up once before, UT1 as defined against
mean solar noon seems less useful to me than some form of "sidereal
UT" for the majority of purposes that a high-precision UT would be
useful for.)

                --Ken Pizzini
Received on Mon Apr 09 2001 - 21:34:13 PDT