search

spacetelescope / pysynphot

Synthetic Photometry.
http://pysynphot.readthedocs.io/en/latest/
BSD 3-Clause "New" or "Revised" License
23 stars 21 forks source link

Problem with the `Phoenix` models pulled by `Icat` class #102

Open hover2pi opened 6 years ago

hover2pi commented 6 years ago

Some exoplanet folks noticed that the Phoenix model atmospheres retrieved by the Icat class were strange so I looked into it. It appears that some of the models are incorrect and don't match the flux density of the raw Phoenix models.

In particular, pysynphot.Icat('phoenix', 6000, 0, 5) is an extreme example but pysynphot.Icat('phoenix', 6000, 0, 4) is also too cool by about 400K. I compared the Icat results to the files they are pulled from at /grp/hst/cdbs/grid/phoenix/ and they appear to be identical so the fault isn't with pysynphot. Instead, whatever script was used to process the raw Phoenix models and generate the FITS files on the ST servers appears to have made some errors. Perhaps, when interpolating to align the wavelength arrays?

I'm not sure if this problem extends beyond Teff=6000 but I would strongly suspect it. Just wanted to leave a note here so it can be looked into.

I made a Jupyter notebook to do this analysis if you want to have a look. Thanks!

pllim commented 6 years ago

I think you should report this to RedCat. They use JIRA. Maybe @rizeladiaz can advise. Thanks, Rossy!

hover2pi commented 5 years ago

@rizeladiaz Hi Rossy! Any leads on this issue? Should I talk to someone specific on the RedCat Team about it? We're still not sure how many of the models are wrong or if this is an isolated incident but it is definitely problematic.

Also, is the code that generates this accessible, i.e. is there a RedCat repo I should move this issue to?

Also, this may be related to #74. Thanks!

stscicrawford commented 5 years ago

@hover2pi If you haven't already, please open a ticket in the Jira RedCAT project about this.