keyword units

[Peter Bunclark]

One of the heated internal debates we had recently at the RGO was about
whether
keyword units should be SI units; we ended up putting the telescope
focus in
metres, whereas mm is far more natural.

A general point is, I think referring to the IAU style-guide is a `cop
out', 
having read the FITS Users Guide I did just that, and the style guide is
targeted
at something completely different, ie published papers.  A
machine-readable 
standard has to be far tighter.

In the example I just mentioned of telescope focus, SI units are pretty
silly in
fact, so I would welcome an official blessing on being able to use
`engineering units'.

William Pence wrote:
> 
> 1234567890123456789012345678901234567890123456789012345678901234567890123456789
> EXPOSURE=               1800.0 / [s] elapsed exposure time
> V_HELIO =                16.23 / [km s**(-1)] heliocentric velocity
> LAMBDA  =                5400. / [angstrom] central wavelength
> FLUX    = 4.9033487787637465E-30 / [J/cm**2/s] average flux
> 

There should be a table of combination units, eg [km s**(-1)], which
have to be strictly adhered to to make parsing easy.  So if your
velocity is standard, then your FLUX would have to be in units
of [J cm**(-2) s**(-1)]

And finally, I leave you with this thought:

DATE-OBS=          1997-07-15   / [UTC] today's date.


Peter.