About the draft itself Re: WD-SIA-2.0

[Patrick Dowler]

Yes, BAND refers to the spectral coordinate and not the redshift axis. 
Nothing in the base use cases calls for querying the redshift axis, and 
it isn't in ObsCore, so support for that will need to wait for DM 
updates and use cases from CSP (for SIA-2.1 and AccessData-1.1).

OK, so BAND values are (vacuum) wavelength in meters with UNKNOWN 
reference position (consistent with ObsCore-1.0).

thanks,



On 06/05/14 08:07 AM, Arnold Rots wrote:
> Hm, TT would be infinitely preferable as a timescale - after all, we are
> doing astronomy.
>
> For the spectral band TOPOCENTER may make the most sense when we are dealing
> with wide bands - and there it does not matter too much.
> High-resolution spectra are a different matter. My sense is that people
> are going to give
> you whatever they have and will not go through the exercise of
> transforming their values
> - basically, because whatever they have is likely to make the most sense
> for the data.
> BARYCENTER might be the preferred one, but I doubt its enforcement will
> be feasible.
> And don't call it barycentric wavelength.
>
> I trust BAND refers to the spectral coordinate, not the
> redshift/Dopplervelocity coordinate.
> There you have the argument that I made above on spectra in spades.
> TOPOCENTER-GEOCENTER-BARYCENTER is one thing, but LSR and Doppler
> definitions
> provide an additional twist and forcing extragalactic conventions on
> Galactic data would be
> unwise.
>
> However, considering that TIME is a much easier case for settling on a
> common reference
> position, and that in that case you accept "UNKNOWN", why not just throw
> up your hands
> and do the same for the spectral coordinate?
>
> Altogether pretty unsatisfactory, that's true. Personally, I think
> that's why the Simple
> protocols are a bit of a curse for the IVOA. As I predicted, when we
> started out with
> them, we will never manage to get beyond the assumptions in the SxAPs.
>
> I have always wondered why meters was chosen if it isn't anyone's
> first,...,nth choice,
> besides wavelength being a quantity that is inferior to Hz or keV,
> physically speaking.
>
> Cheers,
>
>    - Arnold
>
> -------------------------------------------------------------------------------------------------------------
> Arnold H. Rots                                          Chandra X-ray
> Science Center
> Smithsonian Astrophysical Observatory                   tel:  +1 617 496
> 7701
> 60 Garden Street, MS 67                                      fax:  +1
> 617 495 7356
> Cambridge, MA 02138 arots at cfa.harvard.edu <mailto:arots at cfa.harvard.edu>
> USA http://hea-www.harvard.edu/~arots/
> --------------------------------------------------------------------------------------------------------------
>
>
>
> On Mon, May 5, 2014 at 8:07 PM, Patrick Dowler
> <patrick.dowler at nrc-cnrc.gc.ca <mailto:patrick.dowler at nrc-cnrc.gc.ca>>
> wrote:
>
>
>
>     On 14/04/14 01:00 PM, Arnold Rots wrote:
>
>         And what is "barycentric wavelength"? Conversion from frequency to
>         wavelength as applied at the gravitational potential at the
>         barycenter?
>         Or does it mean to reduce the spectral coordinate to the
>         barycenter only
>         dynamically? That's pretty inconvenient for people who are doing
>         Galactic work.
>
>
>     The intent here is that the BAND parameter only supports one unit
>     (m) and one reference position/time/scale/hard-to-__remember thing.
>
>     Since we have to pick one, can you tell me the one most suitable
>     that would apply to BAND and TIME parameters. Note that the draft
>     currently just refers to the same restricted time metadata as DALI
>     uses for literal values (time scale UTC and unknown reference
>     position). What should we use for wavelength values in the BAND
>     parameter, knowing that implementers will actually just query the
>     wavelength(s) they have in files, filter lookup tables, or spectral
>     wcs already.
>
>     PS-Yes, wavelength in meters - no one's first choice :-)
>
>     --
>
>     Patrick Dowler
>     Canadian Astronomy Data Centre
>     National Research Council Canada
>     5071 West Saanich Road
>     Victoria, BC V9E 2E7
>
>     250-363-0044 <tel:250-363-0044> (office) 250-363-0045
>     <tel:250-363-0045> (fax)
>
>

-- 

Patrick Dowler
Canadian Astronomy Data Centre
National Research Council Canada
5071 West Saanich Road
Victoria, BC V9E 2E7

250-363-0044 (office) 250-363-0045 (fax)