-- XavierHaubois - 22 Jan 2014

Objectifs de la réunion 16/01/2014

• Rassembler les utilisateurs intéressées pour discuter des fonctionnalités de l'OIDB

• Connaître les plans de l'ESO concernant une base de données interférométriques et établir un dialogue avec les personnes en charge du projet

Compte rendu de la réunion

ESO-JMMC databases meeting

Special guests: Isabelle Percheron (ESO), Claudia Paladini (Université Libre de Bruxelles), Pierre Fedou (Observatoire de Paris)

JMMC oidb group: Patrick Bernaud, Gilles Duvert, Xavier Haubois, Guillaume Mella, Johan Olofsson

- Introduction by Isabelle about ESO phase 3 for VLTI data. They’ll start working on this in 2015. They’ll produce internal (for PIs et coPIs of ESO programs) and external products (other ESO users identified on the portal). Basically, they will store L1 data and maybe L2 but they haven’t defined exactly how. Their ressources are limited.

- Phase 3 is only for science-graded data. AMBER and MIDI reduction don’t produce science graded data (see http://www.eso.org/sci/data-processing/sdp.html). It means that AMBER and MIDI data won’t be in the ESO L2 database in a first instance.

- ESO doesn't store L3 (important aspect of the JMMC oidb and the links towards publications through bibcode).

- Isabelle and Christian Hummel told the ESO IT engineers (who will build the eso database) that they’ll have to handle OIFITS format. They have no specific data model for the moment. We suggested they could have a look at the oidb definition document and at the ObsCore model.

—> Follow-up on the collaboration with ESO in ~6 weeks —> early March 2014.

- Claudia expressed the worry of several other users about L2 data. They usually don’t trust a calibration made by somebody else for several reasons. One of them being that the calibration scheme differs with the scientific goal you want to achieve with the data. For these interferometry specialists, the database would be good for a first look at the data before they check the calibration process themselves. A way to make our oidb L2 data more trustworthy would be to indicate what calibration scheme that has been used. May be we could come up with a classification of calibration schemes ?

- Claudia : Need to have the VINCI data on the database

- Some old L2 data are in ascii, need to write a script (may be in yorick, IDL and in python) to transform ascii files in oifits that are accepted by the oidb.

- In the future, OIFITS 2 ?

- Vincent Coudé du Foresto engaged with Xavier to put FLUOR data in the oidb.