Galaxy Zoo Talk

Estimating the stellar mass of an SDSS galaxy from its colors - how?

  • JeanTate by JeanTate

    I just posted this in the GZ forum (here). I wonder if my question(s) may be answered faster here?


    Suppose I have a bunch of SDSS galaxies (all spirals), with good - spectroscopic - redshifts and good photometry.

    Suppose I would like to make estimates of the stellar mass in each galaxy, using just the photometry (and the redshift); specifically, not using the spectra. How would I go about doing that?

    From my reading of various papers, it would seem that I would need to do something like this:

    • estimate the z-band absolute magnitude (for which you need a good estimate of the redshift)
    • apply a 'dust correction'
    • apply a 'k correction' (this is where having a robust estimate of redshift is most important)
    • derive an estimate of the stellar mass, from the transformed colors (and 'de-dusted' z-band absolute magnitude).

    Is that right?

    If so, how to work out the 'dust correction'? What 'k correction' to use? What equations to use (in the last step)?

    Also, what would be the typical estimated error, of the derived stellar mass (in dex)?

    Posted

  • vrooje by vrooje admin, scientist

    Hi Jean,

    Just as a clarification, are you looking to do this yourself? It can be done but it has probably already been done for most SDSS sources that are close enough to tell are spirals, so you could use those catalogs if you wanted to.

    But of course it's still a good idea to know how it's done even if you don't crunch all the numbers yourself! Yes, you can use the rest-frame luminosity and dust-corrected color of a galaxy to estimate its stellar mass. (You've probably already read it, but Baldry et al. 2006 is a good paper on why/how this is done.) The uncertainty depends on several factors but I believe it's not uncommon for it to be around 0.3 dex. One of the key things it depends on, however, is the choice of bands you use to compute the stellar mass. In general, the redder the bands, the smaller the error bars. If I recall correctly, a K-band luminosity is enough on its own to get a stellar mass without any other band information because there's essentially no color dependency within the errors.

    There is a Blanton et al. program called kcorrect you can use to do K corrections -- I mention it because it's commonly used with SDSS galaxies, but it's not the only program out there.

    Posted

  • JeanTate by JeanTate

    Thanks! 😃

    I posted this question on four different fora, in Starship Asterisk*, the Cosmoquest forum, Galaxy Zoo forum, and here in Galaxy Zoo Talk. I have copied the reply here to the Galaxy Zoo forum, and will continue the discussion there. Please come and join in!

    Posted