Splinter on Radioastronomy in the VO

 11th October 2019; 5 - 6 pm; DOT and then Student Hotel

 At the meeting: 

 François Bonarel (CDS, Former DAL WG chair 2014-18)
 Mark Allen (CDS, IVOA chair)
 Janet Evans (CXC, TCG vice char)
 Harold Verkouter (JIVE)
 James Dempsey (CSIRO)
 Zheng Meyer Zhao (ASTRON)
 Mattia Macini (ASTRON)
 Yan Grange (ASTRON) 
 Marco Iacobelli (ASTRON) 
 JJ Kavelaars (CADC, IVOA exec)
 Mark Kettenis (JIVE) 
 Mark Lacy (NRAO, IVOA CSP member) 
 Andreas Wicenec (ICRAR, IVOA exec)
 Raffaele D’Abrusco (CXC, IVOACSP) 
 Bruno Merín (ESA, IVOA CSP chair) 

 All the presentations from the “Radioastronomy in the VO” session on Friday 11/10/2019 morning can be found at:

 https://wiki.ivoa.net/twiki/bin/view/IVOA/InterOpOct2019RadioVO

 Proposals for small things that we could do to improve the discoverability of Radioastronomy data in the VO (low hanging fruit):

     Mark Lacy: add Largest Anglular Scale in ObsCore (already part of the CAOM model, JJ recommends using the same language as in CAOM not to create confusion)
     JJ Kavelaars: adding the number of antenae in ObsCore for interferometry
     François Bonnarel: options to partially achieve that
        add  a Radioastronomy extension to ObsCore
        define a provenance extension
     Harold Verkouter : A point spread function of UV plane would contain useful information for someone to explore whether data would be useful. The minimum and maximum resolution is a crude estimate and might be useful as metadata.
         François Bonarel:  extension of Characterization data model
     James Dempsey: what about UV coverage ?
     Raffaele D’Abrusco : we need to either generalize some of the mandatory fields to cover your use case or add new things to reflect. E.g. target is very well defined and e.g. it cannot be a number of celestial sources.
     François Bonarel: Francesca from the Italian VO said that they wanted to add CAOM extending ObsCore. This is something to be checked in the DM and DAL group.
         JJ Kavelaar: we express our ObsCore as a view in our service, there is no translation between CAOM and ObsCore. That makes things easier.
     Zheng Meyer Zhao: How are we going to organize work done in this working group ?
         Bruno Merín: I propose we organize a telecon once in a while with everyone in copy to try to identify the low-hanging fruit actions and try to implement them.
     François Bonarel: changing ObsCore might imply problems in other domains so there are something that we talk often which is the concept of an extension, which has never been done so far but it could be useful
     Andreas Wicenec: we should add to ObsCore the little strings with the configuration, which is very specific to each instrument. This is very difficult to standardize.
     Marco Iacobelli : any representation we do should be based on physical units (bandwidth, frequency ranges, spectral and time resolution). In the interferometry, some of these things depend on the UV coverage. It would be very interesting to map all the instrument features/properties that can be translated for physical parameters for the user to query.
     Mattia Macini : Would it be an idea to map the visibility in the sky with tessellation and then allow searched based on that tessellation?
     Mark Lacy: Faraday Rotation and Rotation Measure are slightly harder problems than those discussed here so we might leave them for a later iteration.

 Further discussion on implementations :

     Mark Kettenis: what is the best practice for finding sources in an ObsCore response?
         Bruno Merín: the query just responds how many observations there are on a particular region of the sky.
         François Bonarel: the intended target is part of the metadata
         We had then a discussion about the fact that mapping the observation “target” field to the actual central positions of the different observing modes in radio is not straightforward since there are off target points for background, or complex dithers that constitute part of a single (or several) observations. In spite of all that complexity, it was believed that the notion of an observation “footprint” was a valid one that could be developed and used to allow for searches in archives. there was a proposal to use "MoCS" to implement these footprints

 Next steps:

     We agreed that we need now two to three data search use cases from the different observatories to then be able to identify the highest impact change to either ObsCore or an extension to ObsCore.
     This prioritized list of use cases (by impact for users) will be discussed then with the IVOA TCG (Technical Coordination Group) to identify the fastest possible implementations.
     We should all meet in a telecon in a month time together with the TCG chairs to review the use cases and identify possible implementations and/or updates of standards. In order the find the best date and time, please give me your availabilities at the following doodle poll:


     https://doodle.com/poll/2mct7p9a2bw4eeek


 Actions:

     Action on François Bonarel: provide 3 data search use cases for radioastronomy data in CDS
     Action on Mark Kettenis and Harold Verkouter : provide 3 data search use cases for radioastronomy data in JIVE
     Action on James Dempsey: provide 3 data search use cases for radioastronomy data in CSIRO
     Action on Marco Iacobelli: provide 3 data search use cases for radioastronomy data in ASTRON
     Action on Mark Lacy : provide 3 data search use cases for radioastronomy data in NRAO
     Action on Jun Han : provide 3 data search use cases for radioastronomy data in FAST
     Action on Andreas Wicenec: provide 3 data search use cases for radioastronomy data in ICRAR and SKA