EroAGN Meeting October 6th 2021
Attending: Mara Salvato, Marcella Brusa, Kirpal Nandra, Julien Wolf, Antonis Georgakakis, Thomas Boller, Andrea Gokus, Adam Malyali, Zsofi Igo, Teng Liu, David Homan, Mirko Krumpe, Zhu Liu, David Bogensberger, Riccardo Arcodia, Kohei Ichikawa, Georg Lamer, Tom Dwelly, Sophia Waddell, Johan Comparat, Esra Bulbul, Andrea Meloni, Steven Haemmerich, Tanya Urrutia
Minutes taken by Tanya
Begin: 13:33. End: 14:41
All sesion only has one topic: Prepare ourselves for eRASS1!
Mara reports on the meeting on ero-FollowUp, update on how to organize themselves, the different groups (stars, AGN, Galaxies, Cluster, ...) to remain self-consistent, especially in the counterparts.
Legacy-DR8 is complemented by CATWISE (may be biased, but covers entire sky). Do now have Legacy-DR9, solved some issues in source-detection, Legacy-DR10 adds a new i-band and deeper WISE (30% more sources to select counterpart and better photo-z).
Antonis presented an update to the maximum likelihood method in the meeting. Meetings are every 2 weeks on Thursdays, anybody welcome to join. Need to make sure to have shared spaces! All papers should reflect the joint information, e.g. flags for different counterparts.
Julien presents the 100 ero-days sample. Comparison of c946 with c947 (includes the eFEDS area).
Simple positional match, ERO_DET_LIKE_0 > 6, ERO_EXT_LIKE = 0. Approx, 60k / 80k are recovered, many of the 20k are artefacts from c946 (known issue). Also, unfortunately pattern in c947, remnant uncleaned sources. Need to clean the ovelap regions better.
c946 have higher positional errors. Use eFEDS to quantify this, 2266 matches to the eFEDS main catalog. Conclusion is that the higher errors have no effect on the counterpart, ~2080 / 2266 coincide astrometrically for both c946/c947 between LS8 and CW2. Difficult to use eFEDS as "truth table". Comparison in X-ray flux looks fine, until <10E-13 fluxes, worried about the "thinnening" below that. Need to rework the flux-flux plot. Teng points out that the "thinnening" may be due to the SEP sources, which will go deeper in flux.
Should present all of this to EroCat with somewhat reworked plots.
Could there be a systematic offset, plot Delta(Dec) / Delta(RA) numbers, instead of total separation, for example. Might have something to do with boresight correction? Probably no systematic offset, because Georg notes that afterwards corrected on real WISE sources, so only larger errors, which is what we see. Tom points out that they are comparing two flawed catalogs to each other, might there be a "truthier" comparison, e.g. SDSS bright quasars? TD notes - for some tests, it would be even better to use a well-selected reference X-ray sample from XMM, Chandra etc (see e.g. EroFollowup/TrainingSamples) - whilst the number of available objects is much smaller, the reduction in the number of ambiguous cases is huge.
Going back to discussing EroFollowUp...
What is the scope between the work of EroFollowUp and the various science groups. AGN WG takes the point sources, but of course not every point source will be an AGN. In EroFollowUp it is supposed to do the reconciliation of that (not always easy, many disagreements).
NWAY and MLR is actually quite democratic in comparison to other counterpart methods. The Hamburg group HAMSTAR actually *only* checks for stars and ignores the possibility that the star they see is just there "by chance" with an actual AGN counterpart (or convincing stellar people that not every star is a GAIA source). This is often where the difficulties in reconciliation come in, since in the end we only want one catalog. Probably not everybody should come with their own counterparts and argue for them, but the work should be done together.
Things to clear up:
If you have in mind a eRASS1 catalog + counterparts, will then the science groups only focus on that counterpart? Probably not, but that should be the ideal case. This should be communicated to the ero-FollowUp. But the MAIN catalog should be always be compared with, so that you then explicitly mention, when you do not agree with that! The whole process needs to be communicated from eroFollowUp to the different science groups, also!