jwst
jwst copied to clipboard
spacetelescope
Use GAIA DR3 in tweakreg
Currently the Gaia alignment for absolute astrometry is turned off by default in tweakreg, but if turned on, it queries MAST using the requests module and uses DR2. DR3 is now out, it is better, and it should be used instead.
Also, astroquery.gaia is already implemented as an interface to the ESA GAIA databases, and could easily be used to access DR3, either as primary or secondary.
https://astroquery.readthedocs.io/en/latest/gaia/gaia.html#examples
So in order of priority:
- Use DR3
- Allow user to select which source it comes from, if needed
- For ESA, use
astroquery.gaia - For MAST, use
astroquery.mast
Whether to add astroquery as a dependency is something to think about, but most astronomers will already have/need it for getting their JWST data.
I can! You agree that this is desirable? Or have other thoughts?
Is there a reason that GAIA alignment is currently turned off in the default pipeline? Tagging @hbushouse and @nden here.
I suppose it can easily be turned on per-instrument as needed via pars-tweakreg ASDF reffiles, but perhaps now that real data is routinely being run through the pipeline, GAIA alignment would be desirable.
The align_to_gaia was set to False by default when the routine was first developed and we didn't have any real JWST data on which to thoroughly test it. That was seen to be the safe thing to do, especially since it wasn't clear how well entries in GAIA would compare with what JWST sees at much longer wavelengths (especially MIRI). Now that we have real data experience, that could be reconsidered, but should probably be decided by the Cal WG, based on real life experiences of the various imaging teams.
That makes sense. Then maybe as a first step it would be good to just update the current code to use DR3 and then defer decisions to turn it on until later. Can still write unit and regression tests to make sure it is working.
This would allow scientist to flip the ON switch and have it work well with the latest GAIA sources and positions
I can! You agree that this is desirable? Or have other thoughts?
WRT this, my thoughts are irrelevant. @hbushouse and @stsci-hack thoughts are more important. Especially, I suspect @stsci-hack might have something in the works already.
Your thoughts are always relevant Mihai. 😄
I don't see an issue for adding GAIA DR3 outside of this one. Related, it seems the sync between Jira and Github issues is no longer working? 😞 As an outside contributor who would like to help, it makes it hard to see if there's duplicate effort already happening inside at ST on a particular issue.
I think that at least adding DR3 as an option is fine and we should go ahead and implement that. Whether we want to make use of astroquery should get some more thought and discussion within the group.
Regarding the github<->Jira bot, it was intentionally turned off early this year to avoid the possibility of someone "in the know" accidentally adding embargoed commissioning-related information to a Jira ticket and then having it become public by being copied to github. We asked ITSD to turn it back on again about a month ago, which they claim they did, but it's obviously still not working. Thanks for the reminder that it's time for me to ping them again about that.
Created https://jira.stsci.edu/browse/JP-2916 for this issue, for discussion and priority ranking by INS.