List of Observables for Observation Core components DM
skoda at sunstel.asu.cas.cz
Wed Sep 29 06:44:47 PDT 2010
>> AS I understand the radio examples - they have some function representing
>> flux out of the source, some complicated function of averaged intensity in
>> given beam as a reference (e.g. antenna temperature).
>>> You're missing "scaled counts".... You could argue that they are the same
>>> [as ADU] from a philosophic point of view, but to an astronomer, they mean
>>> something different.
>> Scaled count again is something "normalized" by given function.
I think that amount of such a variables as "scaled something" is quite
high and it is probably not wise just to add every value someone will
suggest in the table - the table could be quite long and probably not
complete for for every specialist.
>> Example of Francoise - the Halpha filter imaging of solar chromosphere is
>> spectral axis - this imaging is in fact shifting whole plane along spectral
>> axis - but in every point of a prominence you can imagine frequency
>> shifting (if you tune the filter by changing its temperature - you are
>> scanning the spectral axis).
> Don't you have a very small range on this spectarla axis anyway ?
I think it is not the point - the range on given axis is very relative -
depends o what kind of information you want to get - you may have not only
Halfa but other lines as well e.g. Ca H, K line, Ba II line and the system
may have multiple cameras with different filters working at the same time.
But I did not want to go into technical details - I just wanted to propose
the idea of decomposing the problem of observables into two parts
requiring proper vocabulary - the list of real observable (wavelength
(energy), flux, and the scaling or normalizing function or more abstract
concept - like solid angle.
> Can we compare these normalization functions to some kind of "flat field" or
> sensitivity function ?
_ I am afraid there is a little misunderstanding - perhaps introduced by
my (wrong) example of reduction of spectra.
In fact the Sensitivity function or Flta field are the technical aids
helping to get unbiased estimetes of a "real value" of observables.
They want to remove the instrumental signatures. - e.g. proper -
intrinsic idea of flat field is to be really flat - all values near the
unity - just to estimate the relative response of every pixel - this is
sometimes called super-flat and people are "misusing" the
different type of flat fields for other purposes using another properties
- e.g. in spectroscopy the FF removes the fringing - colour gradients etc
But the handling of this technical details belongs to the "black magic of
flat fielding" and to the know-how of instrument specialists.
But in the very sense the FF is just a correction matrix - like e.g. X-ray
response matrices etc ...
But What I wanted to introduce is a ideal observable (free of instrument
effect) as the physical concept. If you go to textbooks the structure of
physics (and perhaps other sciences) is built on the concept of several
observables (mass, length, current, flux, time ....) and the rest is just
combination of them (e.g. density is mass/length^3 etc....
If you would like to ask ObsTAP for finding the table with densities of
different asteroids or planets or table of strength of magnetic fields of
Ap stars there is no easier way then to use exactly the concept as it is
introduced in classes of physics.
In that case Characterzation
> contains a pointer attribute to this. This does not solve fiding a name or
> ucd for the "normalized" quantity.
As I understand the purpose of characterisation - it describes WHERE part
of the obesrvation (range of wavelength, time, filters etc ..)
The Observable describes WHAT (i.e. observable=function(characterisation)
in every element of characterisation domain (or more precisely in the
support of characterisation in sense of distributions - e.g. for delta
functions in case of points - e.g. single frequency - having the natural
So for such a fundamental concept of the very scientific content of the
act of observation (the purpose of all the effort of building large
telescopes and complicated instruments is just to know this observable
- but it must be used complicated reasoning to obtain its real value)
the short table with several values of "randomly" selected variables very
I understand the reason to make all quick and simple to allow the
mining in current VO archives - where is 95% of images and catalogues,
some spectra (I want to prepare some critical review of spectra in VO in the future)
and other tables thanks to Vizier electronic attachments from papers.
But as the Obs model is a fundamental base for further development of
other services and protocols I think it should be designed with wider
scope. Otherwise we would not be able describe the particle observations
(e.g cosmic rays, neutrinos ) (only emg radiation currently in VO?) ,
despite the fact the VO could acquire a lot of support from technology of
people working with neutrino or gravity wave detectors - all the so called
I would like to hear the comments from people working on vacabularies and
unit standards (probably the same logic for developing standards for units
may be used for defining observables .....
BTW - we did not solve the problem of defining WHAT even in the SSA -
in fact all the VO now supports is the selection of product (spectrum
image ...) but not WHAT is in the spectrum or image .
Just to remember what implication for the whole VO science may have the
simple definition of flux calibration in SSA - the relative vs normalized.
and all following discussion with topic "relative fluxes"
But it was not yet elaborated further, however....
* Petr Skoda Phone : +420-323-649201, ext. 361 *
* Stellar Department +420-323-620361 *
* Astronomical Institute AS CR Fax : +420-323-620250 *
* 251 65 Ondrejov e-mail: skoda at sunstel.asu.cas.cz *
* Czech Republic *
More information about the dm