search

LSSTDESC / DC2-production

Configuration, production, validation specifications and tools for the DC2 Data Set.
BSD 3-Clause "New" or "Revised" License
11 stars 7 forks source link

Decide how to handle proper motion #29

Closed cwwalter closed 6 years ago

cwwalter commented 7 years ago

Proper motions are currently not handled properly in DM. See:

https://jira.lsstcorp.org/browse/DM-8828

So, I will check on the status of this with the DM team, but if DM does not handle proper motion by the time of DC2 we should consider turning it off. For imSim this would mean setting the proper motion components to zero in the instance catalog. For PhoSim I believe we would pass the ICRS coordinates into the instance catalog instead of the modified ones when we do the CatSim lookup. @danielsf and @johnrpeterson can comment.

jchiang87 commented 6 years ago

sure. will do.

cwwalter commented 6 years ago

@danielsf Any other motion components I should know about now instead of in March? :)

cwwalter commented 6 years ago

Where do things like nutation show up?

jchiang87 commented 6 years ago

The various transformations applied to the ICRS coordinates are described here:

https://github.com/lsst/sims_catUtils/blob/master/python/lsst/sims/catUtils/mixins/AstrometryMixin.py#L147

danielsf commented 6 years ago

Everything due to the motion of the Earth about its axis (precession, nutation, and aberration) is applied by palpy. We can turn these off by taking palpy out of the loop, but we have to turn them all off.

cwwalter commented 6 years ago

There are complicated interplays between them and PhoSim right? You have explained it to me before and I have understood it 5 times for 10 minutes each time.

I think we want refraction on in any case so I think we should leave these on unless people think they are big effects we should remove?

danielsf commented 6 years ago

Yes. It is complicated, but I think we could safely turn them off, if we had to.

cwwalter commented 6 years ago

Well, we want to turn DCR on so I don't think it would make sense to turn the overall refraction off. It's looks like that is turned on this this code.

This will leave us with nutation which is on a 19 year cycle (a 9" circle) and the aberration which has a maximum angle of something like 20" over the year. I assume the corrections would all be in the same DM code that isn't correcting for the other stuff. I'm inclined to leave this on to get the refraction.

@egawiser and others do you agree?

danielsf commented 6 years ago

For the record, PhoSim handles refraction itself. We explicitly turn CatSim's refraction model off when generating PhoSim InstanceCatalogs

https://github.com/lsst/sims_catUtils/blob/master/python/lsst/sims/catUtils/exampleCatalogDefinitions/phoSimCatalogExamples.py#L363

And the CatSim refraction model is not wavelength dependent at this point. I would not expect ImSim to produce realistic DCR, unless we do some work to add it.

cwwalter commented 6 years ago

And the CatSim refraction model is not wavelength dependent at this point. I would not expect ImSim to produce realistic DCR, unless we do some work to add it.

Yes, we will add it.

egawiser commented 6 years ago

I'm not an astrometry expert by any means, but I would think that it's fine to simulate "motion of the Earth about its axis (precession, nutation, and aberration)" since those should affect all stars and galaxies similarly and be taken care of by astrometric calibration, which is already part of DM. We already know from DC1 that this doesn't cause catastrophic problems, at least.

cwwalter commented 6 years ago

I'm not an astrometry expert by any means, but I would think that it's fine to simulate "motion of the Earth about its axis (precession, nutation, and aberration)" since those should affect all stars and galaxies similarly and be taken care of by astrometric calibration, which is already part of DM. We already know from DC1 that this doesn't cause catastrophic problems, at least.

Yes, good point. OK we leave these on.

cwwalter commented 6 years ago

For the record, PhoSim handles refraction itself.

This makes me realize there is possible an issue of using the same instance catalog for parameterized or ray-traced atmospheres in imSim. I am going to leave this closed issue now and move that discussion to the imSim repo.

Thanks for everyone's contributions to this discussion!

Documenting a summary of the decisions for when we look at this in the future:

Due to proper and other motion of MW objects not being handled by DM at this time (see https://jira.lsstcorp.org/browse/DM-8828) we have decided to turn off proper motion and parallax in the DC2 instance catalogs.

We know from our DC1 experience, that leaving them in introduces complicated and non-negligible astrometric biases.

We will leave in "motion of the Earth about its axis (precession, nutation, and aberration)" along with refraction when appropriate. These should all affect stars and galaxies in the same way and includes some effects we want to simulate.