mattyowl / RSSMOSPipeline

Pipeline for reducing both longslit and multi-object spectroscopic data from the Robert Stobie Spectrograph on SALT.
https://rssmospipeline.readthedocs.io
GNU General Public License v3.0
4 stars 3 forks source link

Fitting of templates used by rss_mos_visual_inspector #21

Open svw26 opened 1 year ago

svw26 commented 1 year ago

Heya Matt,

Hope you're well. I had a chat with Ilani and others on Friday, and they pointed out that your pipeline has a visual-inspector part, which would allow us to get an idea of the redshift (specifically for the absorption-line objects in the G4Jy follow-up).

Please could you outline what each of the templates are? SDSS-023 to SDSS-028 appear to show a combination of absorption and emission lines, whilst SDSS-029 to SDSS-032 look more quasar-like. And what is 'T03' referring to, for the starburst template? (Sorry if this is a silly question.)

I like that you can select/deselect lots of different lines, and save the output easily :)

Is there also a way to run the redshift-fitting automatically/iteratively? I thought someone mentioned something like "running over the different redshifts takes a few minutes" but maybe they were referring to something else. I.e. please let me know if there's another command (other than the one written in the README.md) that can be run.

Many thanks, Sarah x

mattyowl commented 1 year ago

Hi Sarah, the SDSS-0?? spectra are the SDSS spectral templates, which you can find here:

https://classic.sdss.org/dr5/algorithms/spectemplates/index.html

(I like the 'classic' there in the URL, makes me feel old).

There are regular galaxies, a couple of LRG templates, and some QSO templates (I added the latter when we were using this on MALS target spectra back in the day with an MSc student). T03 is the 'Tremonti et al. starburst template'. I'd need to go digging for the exact reference.

There isn't a way to run the redshift fitting automatically. This is fairly ancient code (it's something I made during my PhD and twiddled with a bit since to keep it working), so it's using the IRAF XCSAO package to do the redshift measurements. I could never get things to work reliably 100% of the time, so I preferred to step through and inspect the results manually (i.e., click a button, see if it finds a sensible looking redshift fit, then move to the next spectrum - if it doesn't work, one can always get a z to 0.00x level accuracy by visual inspection/sliding the template around, which was good enough for my purposes).

Unfortunately, as far as using this on SALT data goes, the cross correlation redshift stuff is hardly working at all - it'd need some tweaking to make sure the XCSAO parameters that are in there are good for using on SALT data. Back in the day this was used on different spectra (Gemini etc.). I haven't had the time/motivation to take a crack at that yet.

Ideally there'd be a nice modern Python-based module for doing cross correlation redshifts, which we could put in, but I'm not aware of such a thing.

Cheers Matt