Galaxy Zoo Starburst Talk

Automated creation of BPT diagrams, in GZ Tools?

  • JeanTate by JeanTate

    As I understand it, a BPT diagram (and similar diagrams) can be used to decide if the emission line spectrum of a galaxy (or part of a galaxy) is more consistent with an AGN than a star-burst (or star-formation in general?). It's also true - I think! confirmation please! - that BPT diagrams can be created from the data in the SDSS spectra, both DR7 and DR8/9 versions (provided these are clean enough, and the S/N ratio high enough).

    Is there a way, using GZ Tools, to input an SDSS spectrum (presumably a FITS file) and get a BPT diagram (or similar) as output?

    (If this post/thread is in the wrong section of this forum, would a mod please move it to the right board/section? Thanks)

    Posted

  • mlpeck by mlpeck

    The quench data table has the basic information you need, namely emission line fluxes for Halpha, Hbeta, [N II], and [O III]. I don't see how to do basic transformations of data though!

    To create a BPT diagram you plot log10(oiii/hbeta) against log10(nii/halpha) (note: other sets of lines can be used, but this is the classic one). I just don't immediately see how to transform the data using the available tool set. Perhaps that's coming?

    Posted

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

    Thanks!

    Yes, I saw that the underlying data is there (well, it seems to be; I didn't check on things like the estimated errors, likewise how these data values are derived from the FITS spectra). And yes, I agree that, in principal, you'd expect it would be possible to create something like a BPT diagram from the data.

    But how?

    Posted

  • trouille by trouille scientist, moderator, admin

    Just saw that Quench Tools allows 'Log'. Might be a recent addition.

    So you could use the table and prompt functions to create 2 new columns (1) oiii/hbeta and (2) nii/halpha. You could call (1) O3divHb and (2) N2divHa

    And then create a 3rd and 4th new column using the Log function -- Log (O3divHb) and Log (N2divHa).

    I'd be interested to hear if that works. If you do it, could you share your dashboard here?
    http://quenchtalk.galaxyzoo.org/#/boards/BGS0000006/discussions/DGS000013r explains how to share Quench Tool dashboards.

    Posted

  • JeanTate by JeanTate

    I'm wondering if we could get the DR10 emissionlinesPort data, would it be easier?

    For example, for AGS00001up (DR10 image below)1:

    • EW_OIII_5006 is 25.19393 ± 1.098041
    • EW_Hb_4861 is 4.313535 ± 0.6171916
    • EW_NII_6583 is 15.60685 ± 0.689572
    • EW_Ha_6562 is 16.02848 ± 0.9184773

    I'll have a go at creating a BPT diagram, using Tools; then I'll try to share my dashboard.

    enter image description here

    1 I'm not sure whether to use the fluxes or the equivalent widths (perhaps it doesn't matter?); help!

    Posted

  • JeanTate by JeanTate

    Here's my first attempt, both at creating a BPT diagram and sharing a dashboard.

    http://tools.zooniverse.org/#/dashboards/galaxy_zoo_starburst/51fbc6f9be70a626a1000002

    It would seem there's one LINER - the object at the top-right - and the rest are HII regions (or star-forming/star-burst regions). I don't know how to get the Quench ID for this galaxy, other than by going through each one in the dataset and eyeballing it. I used one of my Collections, Orange bulge as data.

    Posted

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

    Answering my own question: flux, not EW.

    Posted

  • JeanTate by JeanTate

    Aaaand here's my second, using QS:

    http://tools.zooniverse.org/#/dashboards/galaxy_zoo_starburst/51fd0ef4be70a677d2000007

    The scales are logarithmic, and there are no error bars (duh!), but it seems to me that there are LOTS of AGNs in QS (and quite a few LINERs too).

    I had some problems - which I've written up in the uPrompt Functionality thread - and I filtered out outliers first.

    Feedback please! In particular, what do you see when you click on the link?

    Posted

  • JeanTate by JeanTate

    Second attempt:

    http://tools.zooniverse.org/#/dashboards/galaxy_zoo_starburst/520647fe0aab2a3a980001fd

    Posted

  • jules by jules moderator

    I see the plot in your second attempt link Jean. Perseverance paid off!

    Posted