[time-nuts] How many seconds in a year?
dforbes at dakotacom.net
Wed Oct 15 00:50:46 UTC 2008
At 9:38 AM +1100 10/15/08, Neville Michie wrote:
>the GPS system obviously counts seconds, and weeks. When the solar
>year develops a phase
>difference of more than half a second the UTC crowd consider a leap
>second, so what is the current
>rotational velocity of the Earth, either WRT the Sun or the stars? In
>I could find many answers on Google but none that seemed good enough
>to set your clock by.
>By clock I just mean a rubidium clock that is harmonised to the GPS
>cheers, Neville Michie
Unfortunately, the Earth's rotation and revolution around are not as
stable as the oscillations of your rubidium clock.
There is no formula for leap seconds; the folks who insert or
subtract the leap seconds do so using observations of the stars as a
guide. You could say that leap seconds represent the noisy
least-significant digit in the Earth's somewhat jerky movement.
(Jerky relative to the smoothness of a rubidium clock, that is.)
It would be convenient for clockmakers if they calculated the best
fit to a long-term trend of the Earth's behavior and allocated leap
seconds based on that fit, but alas they do not.
So there is no way to build a clock today that is guaranteed to count
seconds correctly in future years, short of having it receive leap
second announcements twice a year and adjust its timekeeping
--David Forbes, Tucson, AZ
More information about the time-nuts