Galaxy Zoo Starburst Talk

How reliable are the estimated log_mass and absmag values for "Merging" (and "Both") objects?

  • JeanTate by JeanTate

    Consider AGS000007y, which is classed as "Both":

    enter image description here

    And here's what SDSS DR9 Navigate has on this:

    enter image description here

    The green are 'Outlines'; blue circles, 'Photometric objects', red 'Objects with spectra'. Amazingly, there are four spectra; their redshifts pretty much confirm that all four objects are very likely close (0.086, 0.087, 0.088, and 0.086) in physical space.

    Only one of the four objects is a QS one (the 'central' one). I would guess all four have MPA-JHU DR7 log_mass estimates1 (I haven't checked yet). How reliable are log_mass estimates in objects like this?

    But why does this matter anyway? Because in any analysis we do, in which we're considering hypotheses in which stellar mass plays a key role, we need to at least consider the possibility that log_mass estimates may be quite unreliable in the "Merging" and "Both" subsets of the data. For example, in Mass Dependent Merger Fraction (Control vs Post-quenched Sample).

    AGS00001r8 is another example, a "Merging" (not so lucky with the spectra though):

    enter image description here
    enter image description here

    I'm guessing that the u_absmag, g_absmag, r_absmag, i_absmag, and z_absmag values do not include light from the blob on the right left, because that's a separate photometric object. It's also likely to have a separate log_mass estimate.

    Assuming that we should at least consider how MPA-JHU log_mass and SDSS photometric pipeline-based estimates are biased, in these types of object - in any analysis which involves either - how to go about estimating such bias?

    And I just noticed: what if the orange blob is where the nucleus (or one of them) is hiding? If it is, and if it's an AGN, then shouldn't we consider 'AGN fraction' biases - due to the spectroscopic fiber not including an AGN - too?

    1 This is where the log_mass values in the QS and QC catalogs come from, if I remember correctly.

    Posted