Galaxy Zoo Talk

mismatched redshift

  • ChrisMolloy by ChrisMolloy

    This galaxy AGZ0004blv has a mismatched redshift between NED and Tools. NED is z=0.009460, Tools is z=0.020227. Is the NED redshift correct as it does look quite close?

    enter image description here

    Posted

  • mlpeck by mlpeck in response to ChrisMolloy's comment.

    NED has a spectroscopic redshift from the 6dF survey, and it is correct: here is the NED link to the spectrum.

    Posted

  • Budgieye by Budgieye moderator

    Those numbers are much the same, when you take the error bars into account.

    SKYSERVER:

    http://skyserver.sdss3.org/dr8/en/tools/explore/obj.asp?id=1237676305095065809

    Photometric z 0.020227
    ..........zErr 0.010023

    or 0.01 to 0.03, and it is an educated guess by a computer program.

    0.01 is close to 0.09,

    NED : the article with 4,700 objects lists 0.009 , and it didn't mention in the abstract how they got the z measurement, but mention the 6dFGS survey. I suppose the z from that would be more correct than a PhotoZ.

    http://articles.adsabs.harvard.edu/cgi-bin/nph-iarticle_query?2005MNRAS.361...34D&data_type=PDF_HIGH&whole_paper=YES&type=PRINTER&filetype=.pdf

    EDIT: I have just seen mipeck's answer, so that is great. There is the spectrum with some emission peaks for accurate measurement. A slightly active nucleus the Hα peaks, and OIII star-forming peaks..

    and I have learned to click on the NED spectrum button.

    Posted

  • ChrisMolloy by ChrisMolloy in response to Budgieye's comment.

    thanks for the responses mlpeck and Budgieye.

    Posted

  • waldo3 by waldo3

    This galaxy appears to be in front of a more distant red galaxy. Would that affect the reading for red-shift? Yes.?

    Posted

  • JeanTate by JeanTate in response to waldo3's comment.

    That's a good question, waldo3.

    In this particular case, the spectroscopic redshift comes from the 6dF survey, which used light from a circle 6.7" in diameter to feed the spectrograph (in the main SDSS surveys the aperture was 3"; in the latest - ongoing - one, 2"). As you can see from this scaled image, very little light from the background galaxy (galaxies?) would have entered the spectrograph, even if the seeing was poor:

    enter image description here

    For the SDSS photometric redshift, it's possible the DR8 method may have incorrectly identified 'nearest neighbors' with higher redshifts, due to there being some light from other sources, but as this galaxy is already shredded into more than two pieces, it would be not at all easy to work out how much of an effect there was (each blue circle is a separate photometric object); besides the estimated uncertainty seems pretty typical:

    enter image description here

    In general, contamination in a spectrum (due to an overlap) results in a lower confidence of the estimated spectroscopic redshift (whether SDSS or not), not a change. I do not know what effect contamination has on SDSS photometric redshifts.

    Posted