[time-nuts] Simple question I am sure

Tom Van Baak tvb at leapsecond.com
Tue Jan 2 00:01:18 EST 2007


> How do we know which is the more accurate timekeeper?
> 
> Keep in mind we have no other units to compare with. Only these two.

If there is only one clock -- that is the exact time
and you are sure of it (and strictly speaking, in
this case, the word "accuracy" doesn't apply).

If you have two clocks, only two clock, and no
other reference point, then you cannot tell which
is the more accurate.

But, you can still tell how well they agree. And
in many cases having an upper bound on their
error is itself important information. You can get
this simply by comparing the two clocks. One
can serve as a check of the other, etc.

To get a better idea which one clock is the best,
you need to add another clock to the mix. If you
do pair-wise comparisons among them it's usually
quite easy to tell which of the three is the worst
or which of the three is best.

If they all appear to be equally accurate, then they
probably are all equally accurate, in which case
the mean of all three (an ensemble clock) is better
than any one of the three.

Now you know why UTC is a sort of mean of some
150 atomic clocks around the world.

> The reason for my question? How do we know when
> someone has invented a more accurate device to
> measure time?

Usually you compare it to your previous best clock
or clocks. Or you build two or three identical new
clocks and intercompare them.

/tvb




More information about the time-nuts mailing list