[LEAPSECS] Nit-pick: SI second

[Warner Losh]

On 02/07/2011 10:58, Tom Van Baak wrote:

> Now to make a time-scale out of these identical but differing SI

> seconds does require some statement of common elevation. So

> this is why reported clock data is all adjusted to sea level in the

> computation of TAI (and hence UTC).


It is also why TAI's rate was adjusted in the 1990's to compensate for 
the red-shifted data that had been collected at NIST in Boulder, since 
it sits at about 5400' (1700m) above sea level (as well as other 
facilities not at sea level).

Warner