Galaxy Zoo Starburst Talk

Reddening from Balmer decrement

  • mlpeck by mlpeck

    JeanTate was onto something months ago, and I'm going to finally follow up on posts she made. The quench sample is significantly dustier than the control sample. This holds for the full samples, the proposed subsets, and starforming subsets (per the BPT diagnostic). Below is a histogram of E(B-V) calculated from the Balmer decrement (Hα/Hβ) for subset 2, and more informative cumulative distributions for the full subset 2 and starforming galaxies within the subset. KS tests say the differences in the CDFs are significant at arbitrary significance levels. In the first histogram the solid red bars are counts for the controls, the open black bordered bars are quench objects.

    enter image description here

    enter image description here

    enter image description here

    As a quick review of how these are calculated I use the formula

    E(B-V) = k * 2.5*log10(Hα/Hβ/2.86)

    where k is a multiplier to convert from E(Hβ-Hα) to E(B-V), Hα and Hβ are the measured emission line fluxes corrected for stellar absorption, and 2.86 is the assumed intrinsic ratio of line strengths. One would cite a paper or text by Osterbrock to justify 2.86. The value of k depends on the assumed attenuation law. I adopted 0.794 from Fitzgerald's (1999?) galactic extinction law; Calzetti's relation would produce about the same coefficient.

    For this calculation I required estimated S/N ≥ 3 in both lines and also discarded any objects with Hα/Hβ < 2.86, which would imply a negative color excess. 90% of the QS sample in subset 2 have color excess estimates, versus 67% in the control sample.

    Obviously these estimates apply to emission line regions sampled by the fiber. Attenuation may be different where the stars live.

    Posted

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

    Interesting!

    Whether the proposal about restarting the main Quench project succeeds or not, would you be interested in continuing to work on this? If so, would you be willing to have me - and any other zooite - as a collaborator?

    Posted

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

    Sure, although I'm going to be somewhat time-constrained through mid-February.

    Are you referring specifically to reddening? Where do you see taking this? If it's a research publication I think we would need a co-author with an institutional affiliation even for an arxiv.org submission. And there might be intellectual property rights issues. Who owns the rights to the quench sample? Yan-mei Chen? Laura Trouille? The University of Wisconsin? SDSS (if so, no problem. Just include the boilerplate and a data release reference)?

    Posted

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

    Worth exploring, I think. But let's put it on the back-burner for now ... (though maybe we could chat about the sorts of not-data-analysis/not-paper-writing challenges in our spare time?)

    Posted