IVOA theory sessions in Kyoto

[Philippe Prugniel]

 >Dear colleagues

<>>
 >I would like to know about the plans of any of you to go to the next 
IVOA meeting in Kyoto and your interest in partaking in one or more 
sessions of
 >the theory interest group. In particular I would like to know whether 
you have anything you might like to contribute there, be it a demo or a
 >presentation.
 >Also, do you have any ideas about specific subjects that you would 
like to see
 >discussed?


Hello Gerard and dear colleagues,

Thank you for this proposition of TVO-IG session in Kyoto, I agree
with the goals you are defining and I will normally attend.

I would like to add comments about a specific field of simulated
data: SED or more generally spectra. In the recent AVO/EURO-VO
science demo (in January in Madrid), two science-cases presented
'fits' of observed SEDs of galaxies against theroretical SEDs
(in this case, models of stellar populations). The need for
(i) having such theoretical data available and (ii) to fit them
to observations was clearly expressed as a priority in the scientific
discussions.

These applications are a first interesting example of introducing
Theory. It is indeed easier to directly fit, or compare, observed
and modelled spectra than comparing images (comparing spectra
can be almost as simple as subtracting them).

Though during this demo it was possible to overplot models to
observations, a number of important issues were left open, for
example:

1) The current observational DM (Observation or SED) make mandatory
informations like the position on the sky (or exposure time...), which
are not relevant to modelled data. For this reason it is not possible to
express a synthetic spectrum using the SED DM (so the demo was not based
on 'pure-standards').

2) To compare models with observations, we have to take into account the
transfert between the source and the observer (which is usually not
considered as part of the model): So we need, for example, to tell if a
spectrum is corrected for Galactic extinction, internal extinction
(within the source)...

It seems that the current SED-DM could almost be used for a large range
of modelled spectra (except the problems of point my (1)). And the
question (2) could concern as well observation or model (we may
like either to correct the observation, or to apply some effect to the
model). However, other components of the DM are specific of the model:
as described in the Provenance object of the excellent paper by
L. Shaw and collaborators (on the TVO site).

I do not know how we should proceed, either include the theory specification
in the more general Observation DM, make a T-DM covering everything from
N-body simulations to spectrosopy or make distinct DM for each scientific
case...

But I can work on the requirements for various applications
of model of spectra, trying to connect them with the propositions of L.
Shaw. And we could discuss it in Kyoto.

Philippe Prugniel.