QS objects whose spectra are not nuclear
-
by
JeanTate
These QS catalog objects are not necessarily outliers, but any analyses using BPT diagrams - to estimate fractions of QS galaxies which contain AGN, for example - need to correct for these: the galaxies may indeed contain AGNs (or not), but if the fiber doesn't cover (or include) the nucleus, we can't tell.
I'm slowly working my way through the catalog; so far ~5% of the objects have non-nuclear spectra. Well, sorta; in some cases it's hard to say where the nucleus is (or even if there is one!); in others there's obviously more than one nucleus; and then there are the nuclei that are swamped by the foreground star in the fiber; ...
Here are DR10 cutouts of 12. The scale is such that the fiber covers approximately the space between the (inner) ends of the cross-hairs1: anything in that space will end up in the spectrum, plus a little bit outside that (how much depends on the seeing at the time). The uids are included in the HTML image filename.
About the last three:
- AGS00000hm - that's an M3 star in the foreground (no one had noticed this before, as far as I can tell)
- AGS00001ka - another foreground star (already noted; nice that I found it quite independently)
- AGS0000272 - this may be the nucleus of one of the galaxies in this merger; the bright white blob is a spectroscopic object (more later)
Oh, and once again a BIG THANK YOU to mlpeck! 😄 The coordinates in the QS catalog - especially the RA values - often have too few significant figures to work out where the center of the fiber is! But from the QS table he created (available in the CasJobs group he set up), I was able to get much better coordinates for all (but three he omitted).
1 except for the first one; I zoomed way out
Posted
-
by
mlpeck
in response to JeanTate's comment.
There are two spectra for AGS00000hm. the one that as of DR8 is no longer the science primary is correctly identified as a z=0.058 galaxy. Note that [O II] 3727, the 4000 Å break, absorption lines in the blue, Na D, Hα+[N II] and other prominent lines are all correctly located. It's possible that an M star spectrum is blended in there, but the galaxy spectrum dominates in the blue.
Posted
-
by
JeanTate
12 more, same deal (the first is zoomed out to show how far from the nucleus the fiber is):
Middle row, last on the right is AGS00000mt : it's hard to say where the nucleus is, but I think it's off to the NW, just beyond the edge of the fiber
Last row, left to right:
- AGS00001jk - looks like two SFRs, and the nucleus may be within the fiber
- AGS00000mt - zoomed waaay out; likewise, the nucleus may be within the fiber
- AGS00002ak - an 'old friend' outlier (star+high-z QSO?); good to know that I could 'recover' it independently
- AGS00001z3 - another 'old friend' outlier (a star)
I will continue checking, but I doubt I'll find even five more, which means these - while quite interesting - comprise only ~1-2% of all QS objects.
Posted
-
by
JeanTate
in response to mlpeck's comment.
If you look at the DR5 templates (also used for DR7, I think), it's easy to see features in the two AGS00000hm spectra which match those of an M3 star, e.g. CaH and TiO. Likewise, as you say, distinctive absorption and emission lines of a z=0.058 galaxy. Curious how the SDSS spectroscopic pipeline gives no warnings, for either spectrum!
Posted
-
by
mlpeck
AGS0000272 is a great find by the way. There's an obvious late stage merger with an obvious post-starburst spectrum and an equally obvious starburst object that just happened to be captured spectroscopically in an overlapping plate even though it's well within the 55" limit for fiber collisions.
This would be a great subject for follow-on observations if a catalog ever manages to get published. If not, it will probably be forgotten.
Posted
-
by
JeanTate
in response to mlpeck's comment.
I noticed that too. And there are a few others that may be as interesting (not exactly like this one of course). I plan to post them later, and not necessarily in this thread. Once I've finished the 'non-nuclear' spectra, I'll be posting some (more) of the marginal/hard-to-say objects I have found. And I also seem to have found a fairly clear, distinct subset of the QS objects, ones for which both fiber covering fraction and Petro_R50 estimated values are likely to be quite misleading.
The two spectra for AGS0000272 - closer than 57" - had led me to wonder how to find which of the QS objects have 'objects with spectra' within 30" (say)? I'm sure it's a fairly straight-forward CasJobs Query, but haven't worked it out yet ...
Posted