Galaxy Zoo Starburst Talk

Can we say anything about the quench sample environments?

  • mlpeck by mlpeck

    Maybe.

    One of the papers linked in the series of blogger posts made early in this project was a study by Yan et al. (2011) of the environments of post-starburst galaxies in nearby and distant galaxy samples. A major tool they used was the distribution of a measure of local density for their samples of post-starburst galaxies vs. the general population.

    I wouldn't know how to reproduce their density measure, but I did find this catalog of galaxies in groups by Tempel et al. based on DR8 data: http://www.aai.ee/~elmo/dr8groups/. The catalog is described in Tempel, Tago, & Liivamaegi (2011): it contains density estimates at several distance scales as well as k-corrected magnitudes and absolute magnitudes and some other data for ~576K DR8 galaxies. The catalog contains 2269 objects from the quench sample (and 2320 from the control sample, but I'm just going to use the full catalog as my control).

    For now, I'm just going to post a color magnitude diagram, which has little to do with the following posts. The sloped line is an eyeballed dividing line between the red and blue sequences -- it's a little different from others used in the literature but at this point I don't care. The scatter of colored points are the quench sample (taken directly from the Tempel catalog, not our compilation), color coded by BPT classification.

    enter image description here

    More later.

    Posted

  • trouille by trouille scientist, moderator, admin

    Great to see this post and line of reasoning.

    Another environment data set to consider is the Baldry set, which gives a density value to DR7 galaxies.
    Baldry densities (SDSS DR7):
    http://www.astro.ljmu.ac.uk/~ikb/research/bimodality-paperIV.html
    (Last of the data bullet points.)
    https://vault.it.northwestern.edu/let412/GZQuench/baldry_density.pdf

    Do you see a difference in terms of merger signatures in the Quench sample and their environment/density?

    Tempel et al. (2011) refer to the Yang DR4 group selection. I thought I'd include the link to that here as well.
    Yang groups (SDSS DR4):
    http://www.astro.umass.edu/~xhyang/Group.html
    (scroll down to the DR4 release)

    Posted

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

    Thanks for the links. The Baldry dataset has about 46% of the number of galaxies of Tempel's, probably mostly due to tighter redshift cuts, and it contains only a little over 1000 of the quench sample objects. Just for fun here is a comparison of their density measures for the ~234K objects in both datasets. Tempel et al. provide density estimates at 5 different length scales, from 1 to 16 Mpc. I left out the largest scale estimate:

    enter image description here

    So there's a reasonable correlation at the shorter distance scales, which I think is consistent with what Baldry et al. claimed to be measuring. For now I'm going to use Tempel's "den1" as the density measure of interest.

    Posted

  • mlpeck by mlpeck

    I'm trying to duplicate some of the analyses that Yan et al. (2009) performed in their study of nearby and distant post-starburst galaxies [original link was wrong, sorry - this now goes to the correct ADS abstract]. Specifically they looked at the distribution of densities of their samples and compared them to the general population of blue and red sequence galaxies.

    In my first plot above the equation of the straight line is

    (g-r) = 0.75 - .035(MR + 20)

    where the colors and magnitudes are as tabulated in the Tempel dataset. I'm going to use that as my semi-arbitrary dividing line between red and blue galaxies. Here is a histogram of the density distribution of the red and blue population, and below it the full quench sample. For now I'm using log(den1) as my density measure.

    enter image description here

    The differences are pretty subtle, right? That's why Yan et al. looked more at cumulative distribution functions.

    Posted

  • mlpeck by mlpeck

    First example: Quench sample "merger signature" (anything other than "Neither" in the merging classification) and no merger signature.

    enter image description here

    Now we see that there's a clear separation between the environments of blue and red sequence galaxies -- blue galaxies are preferentially found in low density environments in the local universe. The overall quench sample lies between the blue and red curves, but the disturbed subset lives in higher density environments than the not disturbed majority.

    Two tailed Kolmogorov-Smirnov (KS) tests indicate that each of these distributions is significantly different from all others at high significance levels (low p values).

    Yan et al.'s sample of local post-starburst galaxies closely followed the density distribution of the general population of blue galaxies, so there's already a significant difference here. The full quench sample favors higher density environments than blue galaxies, but lower densities than red galaxies.

    I'll look at other subsets of the sample as time permits (suggestions welcome). Separating by BPT classifications should come next.

    Posted

  • zutopian by zutopian

    Sorry, but I can't say anything (due to my ignorance) about the environments of the QS galaxies, but I know someone, who is familiar with the environments of quenched galaxies.:

    Towards a physical picture of star-formation quenching: the photometric properties of recently-quenched galaxies in the Sloan Digital Sky Survey

    In a forthcoming paper (Mendel et al. in prep.) we will discuss more directly the relationship between quenched galaxies and their host environments.

    J. Trevor Mendel (MPE), Luc Simard (HIA), Sara L. Ellison (UVic), David R. Patton (Trent)
    (Submitted on 26 Nov 2012)
    http://arxiv.org/abs/1211.6115

    PS: The above paper was submitted one year ago. I couldn't find the forthcoming paper, but would like to know, when it will be submitted?

    Posted

  • mlpeck by mlpeck

    Here is a breakdown by BPT classification in 2 separate graphs. First the starforming and unclassified subsets:

    enter image description here

    Then the other 3 regions - transitional/composite, AGN, and LINERs:

    enter image description here

    So, the quench starforming objects favor lower density environments than even the general blue population, and so far this is the only subset of the quench sample that does so. It seems to me that some argument still needs to be made that the starforming objects in the quench sample actually belong in the sample at all. That is, in what sense are they quenched?

    The 3 groups with some level of nuclear activity are, statistically, indistinguishable in their density distributions.

    Posted

  • mlpeck by mlpeck

    I had noticed that the density distributions for most subsets of the quench sample were somewhat at odds with the results of Yan et al., who found that local K+A galaxies live in the same range of environments as local blue sequence galaxies. It turns out that with redshift cuts the tension goes away. First, here are the density distributions for z < 0.1 and 0.1 ≤ z <0.2 (the upper limit is set by the upper redshift limit of the Tempel data set):

    enter image description here

    For a more direct comparison to Yan here are the results for the same redshift range they used, 0.07 < z < 0.12. Slightly over 1/3 of the quench sample is in that redshift range and just over 900 of them are in the Tempel data set.

    enter image description here

    I suppose these results are due to selection effects rather than actual evolution in recently quenched galaxy properties.

    Posted

  • mlpeck by mlpeck

    One final graph. I divided the Baldry et al. data into blue and red subsets using the same color-magnitude cuts as the Tempel data set and also using the same k-corrected magnitudes. The plot shows the distribution of Baldry density estimates and the portion of the quench sample in the Baldry data. Their data set has redshift range 0.01 ≤ z ≤ 0.085.

    enter image description here

    The curve for the quench sample lies almost exactly on top of the blue sequence distribution.

    Posted

  • JeanTate by JeanTate

    Very cool results, mlpeck! 😃

    Posted

  • jules by jules moderator

    Very cool indeed!

    Posted

  • mlpeck by mlpeck

    Thanks. Unfortunately the Tempel data may need to be used with a little more care than I realized. They made corrections for Malmquist bias and other selection effects, but there is still a systematic trend in the minimum density with redshift:

    enter image description here

    The systematics are reduced but not eliminated with a larger smoothing scale:

    enter image description here

    Ignoring the residual trend for now, here is a comparison of the density distribution of the entire quench sample with red and blue galaxies from the entire sample using the same color cut as before:

    enter image description here

    This time the quench sample curve lies right on top of the blue galaxy distribution.

    A possible alternative measure is group size, which is tabulated for each galaxy in the Tempel data set. Here is a comparison of distributions:

    enter image description here

    62% of the quench sample objects are in groups of 1 (that is, isolated) compared to 52% of the Tempel red galaxies and 56% of the blue ones.

    On the other hand 13 quench sample objects are in groups of 100 or more. That's about 0.6% of the quench sample vs. 1.2% of the entire Tempel data set in groups that large.

    Posted

  • JeanTate by JeanTate

    Wow! Are the umbrellas (trees, whatever) at the top in the second plot galaxy clusters? If so, a very interesting depiction of the FOG ("Fingers of God") effect.

    Posted