Galaxy Zoo Talk

Alice and the Ellipticals, Galaxy Zoo to the Rescue! (Wed 26 Jun '13)

  • JeanTate by JeanTate

    This is an experiment, to find out how easy (or not) it is to copy an Object of the Day from the Galaxy Zoo forum. This is one I posted just yesterday.

    Alice has a way of writing quotable words: "But first please now pass me the smelling salts because after so many blobby ellipticals and no no no no no no nos I almost passed out! 😄"

    Today's Object of the Day is about many of those blobby ellipticals, and how zooites came to the rescue of a damsel in distress help a team of astronomers produce a new cosmic yardstick. But first, let us remind ourselves of what we are talking about here ... [6]

    How do you know that M87 (a.k.a. NGC 4486) - an elliptical galaxy - is 16.4 ± 0.50 Mpc (megaparsecs) from home here? Because astronomers have, over a century or more, painstakingly established and calibrated a cosmic (beyond-our-galaxy) distance ladderWP [1]:

    In a recent astro-ph preprint [2], a team headed by Christoph Saulder reports a new method of estimating extragalactic distance, one that relies on zooites' classifications in the original Galaxy Zoo 8) 8) 8); it makes use of something with the esoteric, and rather odd, name The Fundamental Plane.

    Ellipticals may look blobby, and more cookie-cutter copies of each other (with a bit of melting in the oven, so some are a bit less perfectly round than others) than individuals each with their own special character, but astronomers found that there is a rather simple pattern in their properties. From an analysis of the data in photometric images - such as those we zooites were served up to classify - two key properties are highly correlated, the size (as measured quantitatively by the effective radius) and the central brightness (as measured by the mean surface brightness within the effective radius). When you add a key property derived from analysis of the ellipticals' spectra - the (central) velocity dispersion [3] - you find all three are correlated, and a single, simple, equation connects them:

    log10 (R0) = a · log10 (σ0) + b · log10 (I0) + c [4]

    It's not immediately obvious how you could use this as a giant celestial ruler an extragalactic yardstick, but as you can get σ0 from (good) spectra, and I0 from (good) photometry, you can work out (or estimate) R0 provided you know a, b, and c. And as *R0 is the physical effective radius (measured in kpc, kiloparsecs), comparing this with the apparent effective radius - measured in arcsecs (which you obtain by analyzing (good) image photometry) - gives you the distance to the elliptical! 😛 8)

    Of course, this kind of ruler isn't much use if real elliptical galaxies don't all lie on the Fundamental Plane [5]; what Saudler et al. found is that if you use the Fundamental Plane relationship in five (SDSS) photometric bands, you get a distance estimate that is likely to be within 15% of the 'true' distance (on average).

    How did your clicks help then? This figure, from the preprint, illustrates how [6]:

    The bottom right panel is all 852,173 SDSS DR8 galaxies with spectroscopic redshifts (x-axis is absolute magnitude, fainter galaxies to the right; y-axis is color, redder galaxies to the top) ... the 'red sequence' (slanted line with red/yellow etc in the center; think of it as the head and body of a bird) and 'blue cloud' (the bird's 'wing') are obvious. The bottom left panel is due to us ... it's all 170,234 ellipticals, galaxies zooites classified as ELLIPTICAL. To get a uniform sample to analyze, the team imposed a number of 'cuts', which are reflected in the two top panels (details in the paper).

    Here's a plot (chart? graph?) of how good the resulting Fundamental Plane fit is: [6]

    Oh, and in case you're wondering, the 'Fundamental Plane' distance yardstick is a generalization of the "Dn - σ" one, in the blue square at the top left of the graphic above (the "D" is the elliptical's apparent angular diameter); it is the only method to use a 'standard ruler' rather than a 'standard candle'.

    Without all your clicks, the team would have had to spend many 'Kevin weeks' going through the images of all the objects in the red sequence, to remove all the non-ellipticals. And even if they'd done that, their selection would certainly have been much messier than using GZ DR1, because there'd have been only one classification, rather than >ten independent ones! 8) [7]

    Zooites, take a bow! ;D

    Of course, I can't leave it at that; you must be hankering for more boring, blobby ellipticals, mustn't you? OK, here you are [6]

    [1] click on link for more info, including source of graphic

    [2] Calibrating the fundamental plane with SDSS DR8 data:

    In this paper, we present a calibration of the fundamental plane using
    SDSS Data Release 8. We analyse about 93000 elliptical galaxies up to
    z<0.2, the largest sample used for the calibration of the fundamental
    plane so far. We incorporate up-to-date K-corrections and use
    GalaxyZoo data for the classification of the galaxies in our sample.
    We derive independent fundamental plane fits in all five Sloan filters
    u, g, r, i, z. A direct fit using a volume weighted least squares
    method is applied to obtain the coefficients of the fundamental plane,
    which implicitly corrects for the Malmquist bias. We achieve an
    accuracy of 15% for the fundamental plane as a distance indicator. We
    provide a detailed discussion on the calibrations and their influence
    on the resulting fits. These re-calibrated fundamental plane relations
    form a well-suited anchor for large-scale peculiar velocity studies in
    the nearby universe. Aside from the fundamental plane, we discuss the
    redshift distribution of the elliptical galaxies, and their global
    parameters.

    [3] "how wide the absorption lines are, in the SDSS spectra" is a neat (if somewhat oversimplified) description; in more detail (but still somewhat inaccurate): the atoms and ions which produce the lines we see in the spectra are moving at various speeds, along our line of sight, which produces relative blue- and red-shifts, and the distribution of those shifts is more or less a Gaussian whose width (standard deviation) is the 'velocity dispersion'

    [4] R0 is the (physical) effective radius, σ0 the central velocity dispersion, I0 the 'renormalized' (central) surface brightness, and a, b, and c are constants.

    [5] why is it called this? because if you make a 3D plot you get a sheet, or plane, which all the data points lie on (or close to). if you project this plane onto any of the (three) pairs of two axes (R0 and σ0; σ0 and I0; and R0 and I0 - each of these has its own name), you get a straight line.

    [6] taken from the preprint; Figure 1, Figure 5, Figure 16, and Figure 3 respectively

    [7] our clicks helped in another, less direct way too; the lower bound for the velocity dispersion was determined by checking a sample of CLEAN ELLIPTICAL galaxies (see paper for details)

    Posted

  • fatha731 by fatha731

    And what's the outcome of the experiment? And is that a date in the title?

    Posted

  • JeanTate by JeanTate in response to fatha731's comment.

    Great questions, thanks for asking!

    It's too soon to say (what the outcome is), but some preliminary points:

    • it would be better to copy the original to a word processor (not a super-duper one), and do global Find/Replace, to change the formatting (e.g. GZ forum uses [i][/i] as tags for italicizing, for example, Talk - in its current version - uses a pair of *)
    • smiles look different here; it's probably worth playing around with various options and replacing some
    • the very pale grey background here creates a quite different look (GZ has white text on a black background); I haven't decided which I prefer yet. What's your preference?
    • I like cyan-colored links better than the orange ones in the forum
    • Talk's main panel (the one in which this is appearing) is narrower than the forum's; for an article longer than a few dozen lines, this is means a lot of wasted space
    • Talk's Preview is much preferable to the forum's; instant vs slow and 'on demand'
    • maximum image widths are about the same; however, I already know Talk doesn't allow you to re-scale overly wide ones (I think the outcome of vrooje's experiment was clear in that regard [1])
    • up one level, I can't see how the structure of Talk is conducive to 'regular columns' like Object of the Day is (or was); the structure is too flat (just six "Discussion boards")
    • lack of a counter is a bit of a nuisance; I used that as a rough guide to how popular my articles were.

    Yes, it is a date. Traditionally, the forum's OotD have titles in the form "Wednesday, 26 July, 2013: Alice and ..." Obviously that uses up a lot of characters, so if your title is long, you have to get creative. I changed the format this time, also as an experiment.

    Having now seen both, which do you prefer?

    [1] strange; I can't find it! Not by using Search, not by reviewing the discussions in Help that I thought might contain it ...

    Posted

  • fatha731 by fatha731

    Well first of all thanks for responding. To answer your question I prefer this one to the forums because it's, I don't know, "brighter". But as the backgroung in the forum is black, the orange links do blend in well in my opinion. In all I think I'd prefer this format with slight changes. But what is the discussion you're looking for about?

    Posted