Dear all,
I just provide some feedback to the just posted document by Laurie.
My point of view is more focussed in the data model that refers to:
So, my suggestions are oriented to include these kind of theoretical data in the proposed working draft.
My suggestions concern to the next points in the posted document:
4.2. Characterisation: I would let the characterisation items in the
simulations the most similar to the
ones in observations as possible. And include some additional "objects"
for simulations (like "physical parameters" and something like "Model
Resolution" instead reuse "Resolution").
Here comes my reasons: In the case of synthesis models, several times
the results have been obtained by the use of observed objects (e.j.
observational stellar libraries as commonly used in some synthesis
models like STEBLIB in Bruzual & Charlot, HST data in Starburst 99 or
ELODIE in PEGASE-HR).
In this situation, the results of the models, are not purely
theoretical, but based in observations: results depends on:
Summing up: I would propose that the theoretical data model "includes"
the observational one, PLUS
some extensions (e.g.: physical parameters or model resolution),
instead to re-use the objects used in observations with a different
meaning.
4.3. Provenance:
I have an additional items to be added to provenance, and an extension
of 4.3.5. Resources
Here are some possible text to be added:
In 4.3.5:Resources:
Additionally to the technical details in the computation, Resources would include the list of steps and hyperlinks needed to recover the simulation in a VO environment if possible, in such a way that, in the case that the model would be:
Such a object in the data model would allow the intrinsic pipe-line characteristic of some theoretical models.
As an example, a stellar library (that can be obtained by a VOtool by a
selection of stellar spectrum either observational and/or theoretical),
which would be a final product by itself, would be used by a
synthesis model to obtain the SED of a galaxy. In that case it is not
needed to obtain first the whole
library, but to know "how to obtain some elements" in the library and
what is the parameter space
(log g, Teff, metallicity and so on) covered by the library.
4.5.7.: Credits and Executive Summary?
It is worth to note that no lost of the corresponding credits will occur. In opposite to observational data, managed by institutions, most of theoretical model are managed and developed by small groups, and sometimes by individual persons (examples would be the extensive use of Kurucz atmosphere models or the photoionization code Cloudy, by G. Ferland, by the community). If the data model is not able to recover the corresponding credits, it would produce that some groups will not join the theoretical VO iniciative, since they would considered that their work will receive no credit.
From a metadata perspective it would be required some additional text fieds that sumarize in a univocally way the model by its ingredients, and the cautions for its use. It would be quite useful (from a final-user point of view) if such fields are writen in a "latex-ready" format including bibitem references and the corresponding fields with the \bibitem references in ADS/CDS format, in such a way that it can be used by "cut and paste" the corresponding fileds of the VOTable.
Well, sorry for this long e-mail. I hope that it would be useful
Miguel
El 24/05/2005, a las 9:09, Laurie Shaw escribió:
>
> Hi everyone,
>
> I've now added the latest draft of the simulation data model to the
> IVOA
> theory twiki page [Simulation.v.0.2.doc]. Comments and feedback would
> be
> great!
>
> Thanks,
>
> Laurie
>
>
Received on 2005-05-24Z11:16:44