[LEAPSECS] How you know you're having an effect (was Re: ...)

Warner Losh imp at bsdimp.com
Fri Dec 31 15:37:13 EST 2010

On 12/31/2010 05:37, Clive D.W. Feather wrote:

> Rob Seaman said:

>> FACT: civil timekeeping is ultimately tied to the synodic day (i.e., "mean solar time")

> You make this claim regularly and label it "fact". However, it is a claim,

> not a fact. It may be true at present, but that's just as likely to be

> because the people setting civil timekeeping don't understand the issues.

> After all, given that the time that summer time starts and ends in a large

> part of the world depends on what language you speak, can you really claim

> with a straight face that legislators have thought this through?


> The fact that people happily accept a difference of over two hours between

> legal and mean solar time means they aren't *that* worried about it. And

> what little pressure there is on the topic in the UK is in favour of making

> the discrepency *bigger*.


> (Yes, I know the difference between cyclic and monotonic effects. That's

> not the point.

That was exactly the point I tried to make to Rob for years. He keeps
blasting back that tired refrain, and after the first time of trying to
explain that I do get it, I started ignoring it. Rob likes to confuse
the conflate any analogy of moving away from the sun with a total
support for never having a way to align time to the sun. While
countless ways have been suggested to Rob that are no more burdonsome
than the current time zone mess, he continues to wheel out this old saw.

Frankly, people need to agree on what time it is, and the rest of why
somebody says its that time don't matter to all but a vanishingly small
minority. Almost nobody is going to care if the mean solar time differs
from the time their clocks tell them. Well, nobody except people that
care about time and angle of the earth being the same. In the past they
were, but given that the international definition of a second is what it
is, there necessarily must be a disconnect between the two. Even Rob
argues for a mean solar time, like that's somehow better than the actual
solar time. It is better for him, in his astronomical calculations, but
who cares? It is worse for other people. When confronted with this
reality, Rob runs head long into the solar time argument, without
realizing that he's picking an arbitrary solar time. One could get
along nicely with having the 'mean solar time' be uncorrected, except by
legislative action every 500ish years or so to move the timezone east by
one notch. That may be an acceptable definition of 'mean solar time'
for the purposes of civil time, even if it is a horrible definition for
the purposes of astronomy, astronavigation, and other things.

So yes, even if we never 'resyncrhonize' the UT1 time and the UTC time,
there's adequate mechanisms that can be developed to cope (distribute
DUT1 in a format that can grow large enough: 3276.7 seconds gives us
about 500 years, to harken back to an earlier message). Civil
authorities likely would be happy enough with moving the timezones every
500 years or so to cope with the drift. And in 1500ish years time we'll
have to throw out the current ITU standard anyway because leap seconds
will be more frequent than monthly.

Unlike the rotation of the earth around the sun, which is stable over
the course of 100k years or more, the earth's rotation and the second's
definition are in a fight that only gets worse in time, and at best any
standard we come up with today that keeps things in sync will only last
1500 years or so. What is our technology going to bring us in that
time? There's a very real chance that as our hunger for forces us out
into the solar system a purely terran time system will be viewed as
quaint and antiquated in that time horizon.


More information about the LEAPSECS mailing list