Galaxy Zoo Starburst Talk

Bad spectrum: clearly not a z=0.327 galaxy!

  • JeanTate by JeanTate

    My guess: gross failure to stitch the red and blue spectra together, incomplete removal of major telluric ('sky') lines, and more. Most likely a local star (I'm not familiar enough with stellar spectra to hazard a guess), possibly a double/binary; it may be a rather odd quasar (but most quasars are odd, aren't they?)

    ETA: the object is AGS00002ds

    Posted

  • trouille by trouille scientist, moderator, admin

    Yes, if you look at the redshift error on the source, it's 3.724. Huge error. Clearly problematic.

    This source is actually one of our control galaxies. Your find prompted me to look back at my control sample selection process. I had not thought to make sure to first remove sources with large redshift errors before making a random selection from possible galaxies with the same mass and redshift for a given post-quenched galaxy. Luckily, this is the only source in our sample of 3002 control galaxies with a redshift error problematically large. All others have spectroscopic redshift errors of ~10^-5 (small). I'm going to sub in a good control sample galaxy to replace this source (in the classifications) and we'll remove this bad control sample galaxy results before we do any analysis.

    Thank you for this catch!!!

    Posted

  • JeanTate by JeanTate

    What about this one?

    enter image description here

    Johanson69 pointed us to this; it's AGS00001w5. While the redshift is not at all problematic, the continuum is waaaay off, so whether it's a control or a candidate, analyses will need to correct for the break artifact.

    Posted

  • trouille by trouille scientist, moderator, admin

    Nice find. Yes, that one we'll be able to deal with, as you suggest.

    Posted

  • JeanTate by JeanTate in response to trouille's comment.

    If there's one, there's very likely at least one more. Other than by eyeballing the spectra, is there any way to find which are affected by the break artifact?

    As the break artifact is fixed in the observer frame (us) but not the rest frame (galaxy we're observing), its effect on stellar mass models will be variable both in the discontinuity of the continuum and the rest wavelength where it occurs. Ouch!

    Posted

  • JeanTate by JeanTate

    Maybe a mod could split all the 'bad spectrum' cases into a single, new thread?

    Here's another one, AGS000022s (a.k.a. SDSS DR7 ObjId 587741726574444657):

    enter image description here

    Posted

  • JeanTate by JeanTate

    Another one; this time a star, not even a galaxy AGS00002ak (DR7 ObjId 758877274402460510):

    enter image description here

    Posted

  • JeanTate by JeanTate in response to trouille's comment.

    mlpeck posted an update on this object (AGS00002ds), here. It has a BOSS spectrum, and seems to be a z=4.164 QSO.

    Posted