Galaxy Zoo Talk

Elliptical ... really? Some oddballs among the GZ ELLIPTICAL class

  • JeanTate by JeanTate

    The recent GZ blog post (and accompanying GZ Talk thread), Galaxy Zoo Literature Search, reminded me of some fun I had recently, looking at some of the GZ ELLIPTICAL class objects (quick background, holler if you'd like more details: the first Galaxy Zoo project produced a catalog of the morphology of ~900k objects, among other things. For a subset, the morphology became ELLIPTICAL, UNCERTAIN, or SPIRAL). What I was looking for (unsuccessfully, as it turned out) is irrelevant to this post; what some readers may find interesting is the handful of oddballs (a.k.a. outliers) I found. Also the method I used.

    SDSS J145051.77+053244.9, fracDeV_r = 0.29 (left), SDSS J130235.59+263944.0, fracDeV_r = 0.42 (center), and SDSS J082225.92+040947.7, fracDeV_r = 0.53 (right):

    enter image description here enter image description here enter image description here

    SDSS J095418.15+471725.1, fracDeV_r = 1, specclass STARBURST, not AGN (left), SDSS J125954.33+264911.8, fracDeV_r = 0.31, speclass STARFORMING (center), and SDSS J120638.80+312402.8, fracDeV_r = 0.59, speclass STARFORMING (right):

    enter image description here enter image description here enter image description here

    SDSS J101318.43+315603.6, fracDeV_r = 0.60 (left), SDSS J100911.58+242357.9, fracDev_r = 0.30 (center), and SDSS J113923.86+190736.9, fracDeV_r = 0.78 (left):

    enter image description here enter image description here enter image description here

    Some of these may be among the "blue ellipticals", a class of object found in the GZ data. Others may be lenticulars (S0), a class of galaxy which no GZ study has ever tried to identify, or tried to distinguish from ellipticals. And I think at least one is actually a spiral! 😮

    These ten are from a random sample of 333; here's the core part of the CasJobs SQL query:

    SELECT p.fracDeV_r as fDeV, g.elliptical as EL
    FROM zooSpec AS g
    JOIN PhotoObj AS p
    ON p.ObjID = g.ObjID
    WHERE p.petroRad_r between 5.0 and 10.0
    AND g.elliptical = 1 
    AND (p.htmID*37 & 0x000000000000FFFF) < (650 * 1)
    

    The last line is what does the random selection. Anyone care to have a go at explaining how it works, and why it produces a random selection ("random" in what sense)?

    Posted

  • leeskelvin by leeskelvin scientist

    An interesting selection! Certainly, some of these galaxies will be mis-classified ellipticals, which is to be expected in a survey of so many objects. Often, follow-up observations and repeat classifications are crucial in picking out and correcting these false-positives. There are more recent Galaxy Zoo catalogues which you may want to cross-reference to to see if any of your ELLIPTICAL classifications have changed.

    As for the SQL query(!), I took a look at this page for more details: http://skyserver.sdss.org/dr8/en/help/docs/realquery.asp . If I'm not mistaken, the final line of the query should be constrained somehow, e.g.:
    AND (p.htmid*37 & 0x000000000000FFFF < (650 * 1))

    To paraphrase the link I pasted above, the htmID is a spatial index ID covering all SDSS galaxies. This is multiplied by the prime number 37 to remove bias and generate a random number from the lowermost 16 bits. From this list of modified htmIDs, a selection of between 650 and 65000 objects may be requested, between 1 and 100% of the data respectively. These limits are based on SDSS DR2 data, so no doubt the numbers have increased since then.

    I hope that goes some way towards answering your question!

    Posted

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

    Thanks.

    There are more recent Galaxy Zoo catalogues which you may want to cross-reference to to see if any of your ELLIPTICAL classifications have changed.

    Not only GZ ones, and not only more recent; there are an awful lot of galaxy morphology catalogs, of various kinds, that one can download from VizieR.

    Certainly, some of these galaxies will be mis-classified ellipticals, which is to be expected in a survey of so many objects.

    Indeed. What I was particularly struck by, however, is just how many "oddballs" there are; I posted nine out of a random sample of 333; it would have been quite easy to pick another nine (albeit not so odd perhaps). I'm mildly curious about just how many GZ ELLIPTICALs are actually not, using only SDSS DR13 images. Of course, GZ volunteers (myself included) were using DR7-based images; DR8 (DR9?) onwards ones are considerably better. And, as I mentioned, GZ regards lenticulars (S0) galaxies as ELLIPTICAL (or, more precisely, did not even attempt to distinguish S0's from E's or Sa-Sm's). At least in this regard, GZ's morphologies aren't very good if you're interested in studying galaxies with stellar disks but which do not have "spiral arms" (and are not so highly inclined for any arms to be impossible to see).

    If I'm not mistaken, the final line of the query should be constrained somehow, e.g.: AND (p.htmid*37 & 0x000000000000FFFF

    Yes, the crazy behavior of GZ Talk to regard the < and > characters as some sort of control/escape is maddening. The last line is actually:

    AND (p.htmID*37 & 0x000000000000FFFF) < (650 * 1)

    Posted

  • leeskelvin by leeskelvin scientist

    Ah, formatting issues confounded me also! Good spot!

    As an aside, early next year we'll be launching a GAMA-KiDS follow up study specifically looking at distinguishing between elliptical galaxies and lenticulars. We'll have more to announce in the new year, but if you're interested, watch this space until then! 😃

    Posted

  • Ghost_Sheep_SWR by Ghost_Sheep_SWR

    Perhaps the distinction is moot from the onstart, as with many other things through history. Just humans wanting to classify things by putting them into seperate classes.

    Happenend with animal morphology where genetics showed sometimes it was 'correct', other times really wrong with respect to species relations. I think we're just getting on the genetics equivalent with galaxies. Distinction between Darwinian evolution vs Lamarck with Darwin the great victor until epigenetics showed a subtler story. Living things vs non-living things turned out to be a bit more difficult to draw a clear line than it seemed at first glance.

    Planets vs. Dwarf Planets being the more obvious human interpretation of random stuff around stars, just a broad spectrum of clumps of matter from small rocks to giant gas spheres that don't care were we draw a line.

    Stars vs not-stars is getting less clear-cut with brown dwarfs and M dwarfs, likely to get more complicated in the future when we find objects exactly 'in between' being both and neither.

    A whole spectrum of objects from small star clusters to the largest galaxies, with grey areas between globular star clusters - dwarf galaxies and dwarf galaxies - galaxies (except there are no actual grey areas 😉 ).

    Perhaps as always it will turn out there aren't 'objects in between classes' but only the human condition to want to put things in seperate boxes is at the core of the issue. If we realise it's just us giving names and categories to things we're less surprised the next time the real world isn't so clear-cut black and white 😃

    Posted