[LEAPSECS] the big artillery

Zefram zefram at fysh.org
Thu Nov 6 07:09:46 EST 2014


michael.deckers via LEAPSECS wrote:
>The IERS certainly won't fudge on their units.

I'm afraid they do.  Everyone does in this area.  Even the IAU resolutions
fudge the units.

However, Warner is *also* fudging units, in a different manner, and I
think that's causing you trouble.  Warner has said things like "the UT1
second is 1e-9 different from the SI second".  That statement implies
that the UT1 second is a physical quantity, with dimensionality of
proper time, which can thus be measured using the SI second as a unit.
That's an unusual and problematic view.  Its logical conclusion is that
it's meaningless to denominate the UT1 time scale in UT1 seconds.

My view is that the UT1 second is a unit, not a variable quantity.
It's a different unit from the SI second, and can't be described in
terms of the SI second.  If anything it's a unit of angle, and so can be
described in radians or an equivalent, but it's philosophically valid
to treat it as distinct even from the angle units.  The UT1-related
quantity measurable in SI seconds, which Warner has referred to as "the
UT1 second", is more accurately described as "the duration of the time
period in which UT1 increments by one UT1 second", or more succinctly
"the duration of the UT1 second".

There's a lot of philosophical choice about dimensionality.  Aliasing
units previously treated as distinct always leaves correct equations
still correct, and distinguishing units previously treated as two uses
of one unit can often increase clarity.  So let me acknowledge here that
it *is* also valid to alias the physical (SI) second, the angular (UT1)
second, and the pure angle equivalent, as our traditional equations do.
Aliasing the SI second to angle (as we do indirectly) isn't compatible
with SI, but then neither are Planck units.

-zefram


More information about the LEAPSECS mailing list