Galaxy Zoo Starburst Talk

What is a "BPT diagram"? Why do astronomers like to use them?

  • JeanTate by JeanTate

    Reader's digest answer: a BPT diagram can tell you if the light from a galaxy is dominated by recent star-formation or by an active galactic nucleus (AGN), which contains a supermassive black hole (SMBH).

    In somewhat more detail: the name comes from J.A. Baldwin, M.M. Phillips, and R. Terlevich, who are the authors of a 1981 paper, Classification parameters for the emission-line spectra of extragalactic objects1. They came up with a relatively simple way to tell if the light from a galaxy with strong emission lines (in the 'optical' part of the electromagnetic spectrum; basically between ~380nm and ~1µ) comes mainly from recently formed stars or from gas (plasma actually) lit up by light from the accretion disk around the SMBH in an AGN.

    The idea is that by taking ratios of the strengths of pairs of emission lines that are close to each other in wavelength, the confounding effects of dust (which causes 'reddening') and different mixes of bright stars (which causes the 'continuum' - the part of the spectrum away from lines - to change) can be minimized. Fortunately, by some accidents of atomic physics, two such pairs are handy: H-alpha (the strongest/main Balmer line, in the hydrogen spectrum) and an [NII] line (see below), and H-beta (the second Balmer hydrogen line, corresponding to an atomic transition from n=4 to n=2) and an [OIII] line (also see below).

    The emission lines in the spectra of galaxies come from GGoGGs - great globs of glowing gas (yep, I just made that up 😉); volumes of gas many thousand (or even million!) cubic light-years in size shining in just a few colors (very narrow ranges of wavelengths, 'lines'). Strangely - such is the wonder of astronomy - this gas is, in terms of what we can make in our terrestrial labs, a very hard vacuum! 😮 In fact, if it weren't, we wouldn't see those [NII] or [OIII] lines; the nitrogen and oxygen ions which produce this light cannot do so if the density is much above that of a vacuum.

    Where does the energy come from, to make the GGoGGs glow? It comes from light from stars or from the SMBH's accretion disk (or both)2. And the light from an AGN accretion disk is much 'harder' than light from the hottest of young, massive, bright blue stars (it contains, proportionally, more UV and x-ray light). Which means that the relative strengths of four emission lines will be different (if anyone's interested, I'll have a go at explaining why). Which in turn means that the ratios of the line strengths can be used to tell if the light that's making the GGoGGs glow comes from bright blue stars or from an AGN accretion disk (or some combo of both).

    And that's what a BPT diagram does: by plotting one ratio against the other, you can tell - from the position in the diagram/plot - if it's bright blue stars or an AGN. Cool, eh? 😃

    One last link: bright blue stars are massive; massive stars which are bright are young (these stars live life to the max; they die young, going out in a blaze of glory ... what, can you say?). Young massive bright stars are always blue. If there are lots of bright blue stars, they must have formed recently, in a 'star-forming region', perhaps even in huge numbers over a very short time, in a 'starburst'.

    Here's a BPT diagram I made, using the QS catalog (details in this thread):

    enter image description here

    For more, here's a PDF by Carl Ferkinhof, Distinguishing Starburst from AGN: Application of the BPT diagram

    (About '[NII]' and '[OIII]': the 'N' and 'O' are straight-forward and well-known, the symbols for nitrogen and oxygen. The 'II' and 'III' likewise, if perhaps less well-known to most zooites: subtract one, and you have the number of 'missing' electrons; 'NII' is nitrogen with one missing electron, a singly-ionized nitrogen atom. So 'OIII' must be ... I'll leave that to you, my reader, to answer. The square brackets are trickier; they refer to 'forbidden' transitions. An atomic 'transition' is when an electron 'jumps' from one energy level in an atom (or ion) to another; if it 'jumps up' (i.e. goes from a lower to a higher energy level) it absorbs light, if it 'drops down' it emits light, a photon with an energy equal to the difference in the two energy levels. The rules of quantum mechanics tell us how quickly (or easily) an electron will drop down; for most transitions it's extremely fast. However, there are some atomic energy levels which are 'meta-stable'; the quantum mechanical conditions under which the electron can lose energy mean that it takes a long time for the transition to happen. And in anything other than the hardest of hard vacuums (vacua?), an ion in such a meta-stable state will collide before it 'decays' (if it collides, it loses its energy without emitting a photon). Transitions from such meta-stable states are called 'forbidden' (don't ask). One [NII] line is close, in wavelength, to H-alpha, 6583.46Å vs 6562.819Å3, and one [OIII] line is close to H-beta, 5006.843Å vs 4861.333Å3.)

    (About 'accretion disk': contrary to what you may read in popsci articles, a SMBH is not like a giant vacuum cleaner, sucking matter into its insatiable maw. For matter to end up in an SMBH, it has to lose energy and angular momentum, much as a returning space shuttle has to 'lose speed' to return to Earth. In a nutshell, it does this by 'spiraling down', much as a satellite in low Earth orbit undergoes 'orbital decay'. Only around an SMBH there's vastly more energy to lose, and losing angular momentum means the 'orbitally decaying' matter gets concentrated into a disk, in much the same way a spiral galaxy or a solar system forms4. Such a disk is called an 'accretion disk'. Around an SMBH, the accretion disk will be tiny, no more than a few light-years across; however, the energy matter must lose before it finally disappears across the SMBH's event horizon is so stupendous that such disks can shine more brightly than a billion stars (most of the energy 'lost' is in the form of light; some goes into 'blowing bubbles' and some into AGN jets). And such disks emit - proportionally - far more light in the form of UV and x-rays than even the hottest of massive bright blue young stars.)

    1 and an Erratum - Classification Parameters for the Emission-Line Spectra of Extragalactic Objects

    2 I'm ignoring 'shock heating', or 'collisional excitation'; this is important in supernova remnants, and AGN jets

    3 "air wavelengths" (source)

    4 note to professional astronomers (and especially astrophysicists): yes, I am aware that I am simplifying, enormously. However, I hope what I have written is not actually erroneous; if so, I welcome any and all corrections! 😄

    Posted

  • jules by jules moderator

    This really is useful Jean! However, I'm still not great at reading BPT diagrams and I'm also none too sure what the differences between the coloured lines are beyond that they seem to be some kind of standard lines based on temperature (I'm still reading about these!) With that in mind I thought it worth repeating the "key" you provide in the original thread:

    "The orange line is Kewley et al. (2013) for z=0.1 (the mean redshift of the 2779 is 0.108)

    the green is from Kewley et al. (2001)

    the brown Kaufmann et al. (2003).

    To the left of the orange/green line, the objects are dominated by star-formation.

    To the right of brown line, the objects are dominated by AGNs.

    In between? "Composite"! 😃"

    Posted

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

    Thanks jules.

    It's important to keep in mind:

    • only 2779 of the 3002 objects in the QS catalog have non-negative fluxes for the four lines
    • the catalog does not include errors on the line fluxes, so objects for which at least one line is no different from zero (once errors are considered) cannot even be identified, much less removed (just one example).

    At this stage, constructing a BPT diagram is more about being able to do so - whether using Tools or not - than about what it tells us, in terms of post-quenched galaxies.

    Posted

  • Steve_Maney by Steve_Maney

    What exactly does the acronym BPT represent?

    Beta, Balmer, Blackhole?
    Plasma, parameters, Plot ?
    Transmission?

    Posted

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

    Welcome to the Quench project, Steve Maney! 😃

    BPT diagram? "Baldwin, Phillips & Terlevich" (Google is your friend), named for the three authors of a 1981 paper ("Classification parameters for the emission-line spectra of extragalactic objects"; link to ADS entry here) that is heavily cited by astronomers.

    It is widely used as a q&d way of distinguishing an AGN (active galactic nucleus) from an SFR (star-forming region) - with the possibility that the part of a galaxy within the spectroscope's entrance slit/fiber is a composite, or neither.

    Would you like to join the small team of enthusiastic zooites who are now well into Stage 2 of this exciting - and ground-breaking - Galaxy Zoo project?

    Posted