Spectral quantity in the object's rest frame?

[Baptiste Cecconi]

Hi Markus, 

Ideally such information should indeed be in a full STC description, rather than in a UCD (which should be seen as a "hint" metadata). 

The "pos.obs / pos.rest" extra UCD could be discussed anyway. I would be reluctant to put them under "pos", unless it could be applicable to other entities than spectral quantities. That is to be discussed in a VEP. 

About the list of lines, there had been a discussion 10 years ago or so, when we tried to prepare new solar system UCD. We decided that we would not change the em.line UCDs, and certainly not add more of those, but rather deprecate them eventually. FIY, there were about 20 new em.line UCD proposed, but rejected because we didn't want to open up a path for an infinity of lines within the list of UCD. So I'd be reluctant to change the UCD em.line in overall.

Baptiste


> Le 11 sept. 2024 à 12:04, Markus Demleitner via semantics <semantics at ivoa.net> a écrit :
> 
> Dear Semantics folks,
> 
> I have recently had to UCD-annotate a column containing a spectral
> quantity (actually, a wavelength, but that is beside the point here)
> that was given in the object's rest frame.  Now, the pertinent UCDs
> (em.wl, em.wavenumber, em.freq, em.energy) do not actually state the
> frame, and undoubtedly for a scientifically complete characterisation
> of a spectral column we need a full STC description.
> 
> On the other hand, telling "observed/barycentric/similarly
> observer-related" from "rest frame/object-related" sounds like
> something we can straightforwardly do and that might be useful in
> discovery, where we won't have full STC for a long time (and perhaps
> shouldn't even have it in the first place).
> 
> So... what about having modifiers "pos.obs" (say) and "pos.rest" that
> would let people say in roughly which frame a spectral quantity (and
> perhaps others; this *might* be relevant for periods, for instance)
> is given?  Or can we already do that and I just fail to see how?
> 
> 
> Incidentally, while grepping for "rest" in the UCD list, I
> encountered this:
> 
> S | em.line.OIII                 	|  [OIII] line whose rest wl is 500.7 nm
> 
> I like the spectral characterisation here, although "rest wl" is
> somewhat too terse for my taste.  So, I think I'd prefer "The [OIII]
> line at 500.7 nm".  More to the point, though, I think it would be
> nice if we annotated the remaining em.line UCDs with the wavelengths,
> too, even though I give you that there is less ambiguity for most of
> them.  So, for instance, I would like to have:
> 
> S | em.line.Lyalpha              	|  The Lyman alpha line at 121.5 nm
> 
> Should I prepare a UCD VEP that would do that?
> 
> Also, there is em.line.CO, which is different from all the others in
> that it refers not to a specific line but a whole set of them.  The
> current description is:
> 
>  CO radio line,  e.g 12CO(1-0) at 115GHz
> 
> I think we should stress that this is rather different from the
> others and say:
> 
>  Any radio line of CO
> 
> or something like that.
> 
> I also think we should at least briefly think about whether this
> is a pattern we want to avoid in the future (i.e., have em.line.X
> refer to exactly one line, modulo fine structure and all that), or
> whether "some line of element/molecule X" is actually something we
> want to have UCD atoms for.  I see arguments in favour and against
> both stances, so perhaps we should just leave it open.  Still, it
> does not feel quite right to me that both kinds of things sit next to
> each other.  I a way, em.line*s*.CO would feel less arbitrary.
> 
> Thanks,
> 
>            Markus
>