[obs-tap]observable quantities in observations and their UCDs

[Douglas Tody]

While this is technically correct I suggest the solution is to merely
generalize the concept of "observable" a bit, since actual datasets
from telescopes can be delivered with this quantity as the "observed"
value, and the observation is often explicity intended to produce
a velocity.  The same thing goes for other similar cases, e.g., a
normalized spectrum is not directly observed, but we have them in
our archives and need to describe them and deliver them to users.
Instruments are increasingly complicated and may well do some
processing beyond just what can be directly observed physically.
- Doug



On Wed, 23 Jun 2010, Igor Chilingarian wrote:

> Hi Mireille,
>
> As I already mentioned in our discussion here at CDS, I do not think that 
> "velocity" is observable in any sense. I would agree that it can be a 
> spectral coordinate, but we NEVER directly observe velocities. Velocities are 
> results of the [model-dependent] data analysis.
>
> With best regards,
> 						Igor
>
> On Thu, 17 Jun 2010, Mireille Louys wrote:
>
>> Dear DM members,
>> 
>> At the last interop in Victoria, we decided to build
>> up a list of observable quantities that are measured in an observation
>> and to attach to it a proper Ucd. These will be the observables
>> considered for the Obs/Tap specification .
>> 
>> I gathered material from the Polarisation Note and Interferometry note by 
>> Anita.
>> I included items from Francois's observable list.
>> 
>> Please find here a table to start with for this.
>> 
>> Thanks for your inputs, comments and corrections.
>> 
>> Best regards, Mireille
>> 
>> 
>