Galaxy Zoo Talk

Merger, collision or just weird?

  • metalmania by metalmania

    What's going on here? it looks like the nucleus has been pulled out of this edge-on spiral, but it could be a merger with the SQO/AGN, there's also the debris trail to the left...perhaps it is a collision from the left and the bright AGN has passed through the edge-on?

    Posted

  • KWillett by KWillett scientist, admin, translator

    That is wicked. The middle object is an edge-on disk with a prominent dust lane. The spectrum has a broad H-alpha line, which means it's classified as a quasar.

    The spectrum of the small object to the right is very weird, and missing some data in the middle. There's no good redshift fit to this galaxy from the SDSS pipeline - do you see any clear lines that would tell us whether it's physically associated with the galaxy in the center?

    Posted

  • mlpeck by mlpeck

    @kwillett:

    I think the SDSS pipeline botched the redshift of the object to the right, or maybe it's just possible the spectrum is a star/galaxy overlap. Anyway the broad line at ~7075Å is probably Hα. That would make the other narrow emission lines [O III] and [O II].

    I can check if that works in detail, but I don't have time right now.

    Posted

  • mlpeck by mlpeck

    I found some time. The spectrum of the object poking out of the surface of the edge on galaxy is a star/galaxy overlap. The galaxy has redshift z ≈ 0.078, so within a few hundred km/sec of the edge on galaxy. The star is ∼M as classified by SDSS. Details later, perhaps.

    Posted

  • zutopian by zutopian

    I recognized this galaxy/merger! It had been presented five years ago as an "Object of the Day" in the old GZ Forum ! The Stripe82 and SDSS DR7 images had been posted there.:

    http://www.galaxyzooforum.org/index.php?topic=278416.0

    Please read also the replies (2 pages), which had been posted there! One of the replies is by GZ astronomer NGC3314!

    PS: For new volunteers' information: The old GZ Forum is archived/ replaced by GZ Talk!

    Posted

  • zutopian by zutopian

    SIMBAD information: 2SLAQ J233607.97+000550.3 -- Star

    EDIT: I noticed, that the above information actually belongs to a different object! The coordinates are different.

    Posted

  • metalmania by metalmania

    Thanks for the extra info/link zutopian...so it's a star which just so happens to look like it's interacting - could that be down to the fact there is what looks like a polar ring and this star is sitting right in the middle of that feature?

    EDIT: mipeck also suggested it was a star, but based on the edit above, there might be still some uncertainty about the object. If that's the case I guess we can't say that it's a polar ring either as it might be debris.

    Posted

  • zutopian by zutopian

    @metalmania: You are welcome! Please be informed, that I edited my 2nd post! So you might want to edit your reply!

    Posted

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

    @zutopian:

    Thanks for the link to the old GZ forum thread. Bill Keel had it right -- the spectrum is a star/galaxy overlap, with the galaxy at z≈0.078. Here are the details I promised to think about providing yesterday.

    First, here's the SDSS finder chart image with the locations of the spectroscopic objects marked. There are actually 4 spectra of this galaxy pair with two at the same location. The one I'm talking about is centered on the bright point-like object that appears to be punching through the edge on disk galaxy.

    enter image description here

    One of my hobbies is stellar population modeling of galaxy spectra. The basic inputs to the models are a library of "simple stellar population" model spectra (I'm currently using subsets of the MIUSCAT extension of the MILES libraries), a dust attenuation model, and I include a fake for emission lines. For purposes of these exercises I just do non-negative least squares fits, with no effort to force "reasonable" star formation histories.

    About a year and a half ago on a now (presumably) defunct GZ related project the topic of possible star/galaxy overlaps was raised. In an effort to quantify the putative contribution of stars to the spectra in question I extended my models by building a small library of stellar spectra that was created by downloading ∼5000 SDSS stellar spectra and stacking them into a dozen fairly broad temperature bins.

    Here are results of the fitting exercises. First is a fit with the SSP library only. In these graphs the actual spectrum is shown in orange and the fit in green in the top pane. Wavelengths are in the galaxy restframe, and as I said above I estimated the galaxy redshift to be just slightly under z=0.078. The second pane shows residuals from the fit in units of standard deviations.

    j2336+0004_no_star

    So this doesn't work too well. The continuum redward of the gap in the spectrum is way off and there are some prominent absorption features that aren't fit. On the other hand prominent emission lines are where they should be, so there's definitely some galaxy light in the spectrum.

    Next is the fit with the stellar library included.

    j2336+0004_with_star

    This is much better! The third and 4th panes in this graph show the modeled galaxy and star contributions. I work in the galaxy rest frame, so the star spectrum is blue shifted to match the galaxy rest frame wavelengths. The deep absorption line at ≈5400Å is Na D, those are Calcium lines bracketing 8000Å, and I'm not sure about the rest but I think the broad troughs in the red are molecular bands. Notice those nicely fit the wiggles in the observed spectrum where the SSP only fit badly fails. Overall the star contributes anywhere from about half (in the blue) to 75% of the light. I can only guess at the MK class, but early M seems about right.

    Posted

  • Capella05 by Capella05 moderator in response to mlpeck's comment.

    Just getting round to reading this thread - awesome everyone!

    A question,

    I found some time. The spectrum of the object poking out of the surface of the edge on galaxy is a star/galaxy overlap. The galaxy has redshift z ≈ 0.078, so within a few hundred km/sec of the edge on galaxy. The star is ∼M as classified by SDSS. Details later, perhaps.

    Isn't km / sec a velocity rather than a distance? How are you calculating the separation?

    Just curious 😃

    Posted

  • mlpeck by mlpeck in response to Capella05's comment.

    Isn't km / sec a velocity rather than a distance?

    Hi @Cappella05: sure it's a velocity. The formula for the radial velocity difference between two objects with redshifts z and z0 is

    Δv = c*(z - z0)/(1+z0)

    I think that's independent of the assumed cosmology (maybe for small redshift differences only?). I could probably find a reference for that formula if pressed.

    Anyway the spectrum closest to the center of the edge on galaxy has z=0.07717, I estimated the redshift of the one poking out the side to be z=0.07798 and the velocity difference follows from the formula.

    Posted

  • Capella05 by Capella05 moderator in response to mlpeck's comment.

    I am familiar with that formula - the excellent Prof Matilsky covered it in the last MOOC I did.

    But, why not convert it to a physical distance, like arcseconds or parsecs? or is there some advantage to keeping it as a velocity?

    Edit: Not doubting the redshift figures, I just like figuring out the science behind things. No Offence intended.

    Posted

  • JeanTate by JeanTate

    It's difficult, in general, to convert an observed redshift difference to a physical distance, when you have just one pair of objects.

    To be ~close in physical space, the redshift difference should be ~comparable to, or smaller than, the typical relative motion of a pair of galaxies. For small groups, this is fairly small, but for a rich cluster it can be 1,000 km/s, or even more.

    Another, major, complication is projection: an angular separation (on the sky) can give only a projected separation, assuming that the redshifts are close. The universe being as big as it is, there are certainly pairs of galaxies, with similar redshifts, whose projected separation is small, but due to perspective, the physical separation is huge.

    Very cool to see you using the composite spectra again, mlpeck! 😃

    Posted

  • mlpeck by mlpeck

    But, why not convert it to a physical distance, like arcseconds or
    parsecs?

    That's a fair enough question, although all you can really calculate is the transverse separation. The angular separation between the fiber approximately centered on the edge on disk and the one centered on the star masquerading as a galaxy nucleus is about 3.6". NED's cosmology calculator gives the scale at the redshift of the EOD of about 1500 pc/arcsec, with some variation depending on the assumed local standard of rest. So that makes the transverse separation 5.4 kpc, more or less.

    Jean gave a pretty good explanation of why we'd look at the velocity difference. It's mostly just a sanity check on whether they're really physically close and gravitationally interacting.

    Posted

  • KWillett by KWillett scientist, admin, translator in response to mlpeck's comment.

    Nice discussion, everyone (and particular thanks to the excellent stellar spectrum analysis, @mlpeck). The evidence both from now and the earlier SDSS discussion seem to settle the question of the bright star; I still think the tidal debris tail to the left of the image is really interesting. It's rarer to see that parallel to the plane, especially when I can't see evidence of the smaller merging nucleus. The slightly bright white bulge lower in the disk is a candidate, or the star could be obscuring details closer to the center ...

    Posted

  • zutopian by zutopian

    In below paper there is an image of a galaxy merger (page 5), which resembles AGZ000ao5x, doesn't it?:

    Galaxy Disruption in a Halo of Dark Matter
    Authors: Duncan Forbes, Michael Beasley, Kenji Bekki, Jean Brodie, Jay Strader

    (...) We have discovered a dwarf satellite galaxy in the process of being torn apart by gravitational tidal forces as it merges with a larger galaxy's dark matter halo. Our results illustrate the morphological transformation of dwarf galaxies by tidal interaction and the continued build-up of galaxy halos.

    (Submitted on 31 Aug 2003)
    http://arxiv.org/abs/astro-ph/0309020

    Posted