Wednesday, 10th July, 2013: Where Does This Lead To?

by JeanTate

That's how it began. In the SDSS DR8 Navigate tool, this is a GALAXY.

It's not, of course; it's a small nebula, which goes by the name of SDSS J230311.47+614203.0, and is green because the H-alpha emission line is in the SDSS camera's r-band, and the r-band is mapped to the G in the RGB JPEG images we get served up.

AlexandredOr's splendid Dark Nebulae thread has the broader context [1]:

source: author=AlexandredOr, date=April 11, 2012, 09:00:11 pm

1237663234974156836

Not likely to be any background galaxies visible in SDSS images of such dark nebulae, surely! But what about something not so obviously black and opaque? Like this, from The Nebulae Collection (formed from many threads) thread perhaps?

source: author=mitch, date=February 21, 2013, 06:38:40 pm

http://skyserver.sdss3.org/dr9/en/tools/explore/obj.asp?id=1237663786889642346

http://cas.sdss.org/astro/en/tools/explore/obj.asp?id=587738065664737602

Hi Liz & Peter

Another one that hopefully seems to be new, if my number checking is correct 😃

cheers

mitch

For example [1]:

DR8 ObjId 1237663786889773151, DR7 ObjId 587738065664868442.

Maybe this looks a bit like it's behind a veil of dust, maybe not. However, if you click on PhotoObj in the left-hand panel, you get a huge long table of photometric parameters and their estimated values. Among them are "modelMag_r" (15.803908) and "dered_r" (15.243956), which differ by 0.559952 magnitudes, or 0.56 for short. That means what, exactly?

Well, it means that the automated DR8 SDSS photometric pipeline estimates the r-band magnitude of this galaxy as 15.80, and that its 'de-redened' r-band magnitude is 0.56 mags brighter. This dimming by dust - which astronomers call 'extinction' - is estimated from observations made more than a decade ago now by COBE (the COsmic Background Explorer) and IRAS (InfraRed Astronomical Satellite) by D.J. Schlegel, D.P. Finkbeiner, & M. Davis in a 1998 paper that has over 8,000 citations (you can read all about it on their Dust Maps website).

OK, but so what? It's obviously important to take account of extinction (redening) by dust, if you want to get an unbiased sample of galaxies. For example, "to combine or compare galaxies across a range of redshifts", which is what the Galaxy Zoo Science Team set out to do, as described in one of the main, early ~~Galaxy Zoo~~ Zooniverse Published Papers ("Galaxy Zoo: the dependence of morphology and colour on environment, Bamford+ 2009.", click on the link to get a copy of the actual published paper):

[quote=Bamford+ 2009]

In order to combine or compare galaxies across a range of redshifts we must account for the redshift dependent selection biases. To remove selection bias from the analyses in this paper we restrict the galaxies considered to those that would meet our apparent magnitude, size and surface brightness criteria if they were located at the upper limit of the redshift range considered. As measures of galaxy size and surface brightness we use the radius containing 50 per cent of the Petrosian flux, R₅₀ , and the average surface brightness within this radius, μ₅₀ , all from the r-band imaging. Given the upper redshift-limit we adopt, z LESSTHAN 0.085, the Main Galaxy Sample limits (r LESSTHAN 17.77 mag, R₅₀ GREATERTHAN ~1 arcsec and μ₅₀ LESSTHAN ~23.0 mag arcsec⁻² ), and our assumed cosmology (see below), we thus limit to the subsample of magnitude-limited sample galaxies with M_r LESSTHAN −20.17 mag, R₅₀ GREATERTHAN 1.6 kpc and absolute surface brightness μ₅₀ LESSTHAN −13.93 mag kpc⁻² . We refer to this sample, which contains 125923 objects, as our luminosity-limited sample.

[...]

Throughout this paper we assume a Friedmann-Lemaître-Robertson-Walker cosmology with Ω_m = 0.3, Ω_Λ = 0.7, H₀ = 70 km s⁻¹ Mpc⁻¹.

OK, maybe that's a bit too technically dense to easily grasp; the key point - for my OotD question ("Where Does This Lead To?") - is "magnitude-limited sample galaxies with M_r LESSTHAN −20.17 mag".

And even more simply, as long as the redshift is between 0.03 and 0.085, and the r-band magnitude brighter than 17.77, it's in play [2]

So the galaxy above (DR8 1237663786889773151) is in play, because its r-band (apparent) magnitude is brighter than 17.77.

What about this (DR8 1237663786889838753), also from the same part of the sky [1]:

Its apparent r-band magnitude is 18.10, and its deredened ("corrected for Galactic extinction") one is 17.72, an extinction of 0.38 mags. Its r-band absolute magnitude is certainly brighter than -20.17, but because its apparent magnitude is fainter than 17.77, it would not have been selected! [2] [3]

How the GZ Science Team addressed this is - I think - pretty important, if only because their key paper's conclusions (and findings) seem to rely heavily on their sample being "luminosity-limited" [4] Putting this another way, if it could be convincingly shown that the sample was not, in fact, "luminosity-limited", none of the conclusions made assuming it was would be valid. 😮 [5]

But how many galaxies are we talking about here? Ones with absolute magnitudes brighter than -20.17, but apparent magnitudes fainter than 17.77 (in the redshift range 0.03 to 0.085)? Time to dust off CasJobs, wouldn't you say?

I'm going to close with a question: how did the GZ Science Team address this question?

You never know where spotting a small green blob might lead you.

[1] I'll post an image with a pointer to this green blob/edge-on spiral/blobby elliptical in a later post, if you can't find it/them in the zoomed out image

[2] This is, as you might expect, an oversimplification. Even ignoring the size and surface brightness criteria, what follows in this OotD does not consider the effects of the necessary K correction

[3] This galaxy would not have been selected anyway, as its redshift is is too high (0.176), but we know that only because a spectrum was taken in a later phase of the SDSS.

[4] This can be seen in an earlier para in Section 2.1: "the luminosity range over which we are volume-limited (M_r LESSTHAN −20.17 mag)"

[5] Or, rather, their validity would be indeterminate.

Posted July 10, 2013 12:40 AM
by JeanTate

Why GREATERTHAN and LESSTHAN? Because the standard characters/symbols cannot be used in GZ Talk!! 😮 Check out this bug report for more details.

Yes, this is today's Object of the Day, over in the Galaxy Zoo forum. And yes, fatha731, I'm still experimenting 😉

Posted July 10, 2013 12:50 AM