[LEAPSECS] the big artillery

[Tony Finch]

So when I was googling around for more information about what paper
timescales the BIPM publishes, I found this:

http://iag.dgfi.badw.de/fileadmin/IAG-docs/Travaux2013/08_BIPM.pdf

which says:

> The algorithm used for the calculation of time scales is an iterative
> process that starts by producing a free atomic scale (Échelle atomique
> libre or EAL) from which TAI and UTC are derived.

> EAL is optimized in frequency stability, but nothing is done for
> matching its unit interval to the second of the International System of
> Units (SI second). In a second step, the frequency of EAL is compared to
> that of the primary frequency standards, and frequency accuracy is
> improved by applying whenever necessary a correction to the frequency of
> EAL. The resulting scale is TAI. Finally, UTC is obtained by adding an
> integral number of seconds (leap seconds). Research into time scale
> algorithms is conducted in the Time Department with the aim of improving
> the long-term stability of EAL and the accuracy of TAI /UTC

On the other hand the most frequent BIPM publication on time comparisons
is rUTC (not rTAI).

Tony.
-- 
f.anthony.n.finch  <dot at dotat.at>  http://dotat.at/
Rockall, Malin, Hebrides: Southeast 7 to severe gale 9 veering southwest 5 or
6. Very rough or high. Rain then showers. Moderate, occasionally poor.