[time-nuts] UTC and the speed of light?

Tom Van Baak tvb at LeapSecond.com
Tue Aug 30 16:08:58 UTC 2011

> How is the speed of light accounted for in the definition of UTC?
> In other words, how did they solve the conflict where on one hand we'd all
> expect two "perfect" clocks to "tick" at the same time but wether they do
> depends on the location of the observer?
> -- 
> Chris Albertson
> Redondo Beach, California


Correct. Even perfect atomic clocks will run at different rates if
they are not in the same gravitational potential. Clocks at higher
altitude run a bit faster compared against clocks at lower altitude.
The correction is about 1.1e-16 per meter. This sounds really
small until you realize it's 1.1e-13/km or 1.8e-13/mile which is
huge by today's standards. NIST, for example, is a mile above
sea level.

The conflict you speak of is solved when TAI is created from EAL.

Hundreds of atomic clocks around the world are combined into
a free-running uncalibrated timescale called EAL. The input of
a few dozen primary standards allows the gravitational offsets
inherent in EAL to be corrected. This is where TAI comes from.
And UTC comes from TAI. You can think of it this way: EAL is
stable but TAI is stable and accurate. Does that help?

See a good description and pretty diagram of this process at:

If this doesn't answer your question let me know. I'm not clear
what you meant by the speed of light part. The above deals with
the gravitational part.


More information about the time-nuts mailing list