What do we learn from UV spectroscopy of galaxies? (Part 2, Lyman α )

The UV spectra of galaxies also contain features in emission.  The strongest is the Lyα  line, which  turns out to be quite tricky.  I’ve spent a lot of time on it, so I guess I’m attracted to complicated problems!  Let’s go over the basics:

  • Lyman α  (Lyα  for short) is a recombination line of hydrogen, just like the Hα and Hβ lines seen in the optical spectra of galaxies, and discussed in the context of the Green Peas in a previous post. When the electron and proton in ionized hydrogen recombine, the electron makes its way to the ground state by emitting recombination lines.  The intrinsic line ratios (i.e. emission probabilities) are set by atomic physics.  For normal conditions seen in galaxies, each ionization is followed by a recombination, and 2/3 of all ionizing photons then lead to a Lyα photon.
  • But, Lyα, unlike the optical lines, is strongly resonant.  This means that when a Lyα photon is emitted, it has a high probability of being absorbed again in neutral hydrogen.  And then emitted again.   This scattering leads to a random-walk process, where Lyα  photons tend work their way outward from the ionized regions where they were created, potentially to large distances. An example is shown below, where Hubble Space Telescope imaging is used to isolate Lyα (in blue).   The stars (green), and the Hα (red) are much more compact.

    Image courtesy of M. Hayes and G. Ostlin. This galaxy, known as LARS 14, shows Lyα (blue) that is much more extended than its young stars (green) and ionized gas (red).
  • If Lyα photons are random-walking their way out of galaxies with lots of scattering, they have an increased probability of running into a dust grain and being absorbed.   While we’ve known that UV photons are highly susceptible to dust absorption, Lyα photons are affected even more.
  • if Lyα photons scatter in the outflowing gas that is Doppler shifted, they can be re-emitted out of resonance with the bulk of the scattering neutral hydrogen.  Then their probability of scattering and being destroyed by dust is lower. So we think that outflows might help the Lyα to get out of galaxies.
  • When we observe Lyα from galaxies, it is all over the map!  Sometimes it shows up in absorption, sometimes in emission, and sometimes we even see P-Cygni profiles, where there is blueshifted absorption and redshifted emission. The bottom image shows just some of the variety of Lyα profiles that can be seen in galaxies.  Keep in mind, the Lyα photons are originally injected into the interstellar medium of the galaxies in symmetric profiles that can usually be represented by a Gaussian.  The scattering process modifies the profiles to create the shape that we observe. Below, on the left side, the top two panels are good examples of typical P-cygni profiles, whereas the others are pretty complicated.   The blue emission that emerges in some profiles is probably an indication that there is not as much neutral hydrogen to scatter the Lyα.
Lyα emission profiles from nearby galaxies, observed with the Cosmic Origins Spectrograph (COS) on the Hubble Space Telescope and published in Heckman et al. (2011).  These galaxies are part of a sample called Lyman Break Analogs, which are similar in some ways to the Green Peas. The emergent Lyα emission shows a variety of shapes, even though the photons are originally emitted as symmetric (Gaussian) profiles.
  • The variety in Lyα profiles is actually pretty neat, because it is telling us something about the gas that is scattering the photons.  In most  galaxies, except for very nearby systems, this gas is hard to observe.  At the same time, however, it is important, because it is the accumulation of ejected material from supernova-driven outflows that halt star-formation, and also the fuel that is coming back into the galaxy to form new stars.   So we hope that studying the Lyα might be able to help us figure this out.
    Radiative transfer models take an intrinsic Lyα profile, and show how scattering changes the line when different densities, velocities, and gas distributions are adopted. This figure is taken from Verhamme et al. (2015).

    One way to do this is to use radiative transfer models, where computer simulations model the scattering in gas with different densities, distributions, and velocities. Some examples  of  these models are shown at the left.  The dotted line shows the assumed intrinsic profile of the Lyα line, and then the spectrum is simulated having several different HI column densities.  (Column density is the surface density– atoms per square centimeter– that you would see if you took a pencil beam from us to the source, and summed it along the line of sight.)   In these simulations, we assume that the HI gas lies in a spherical shell that is expanding at some velocity– 50 km/s in the figure.

  • Finally, we’re pretty sure that Lyα is telling us about the reionization of the intergalactic medium, or IGM, at very high-redshifts.      After the Universe was about 1 billion years old (z<6), the IGM–or gas between galaxies– contained fully ionized hydrogen. But, sometime before that it was completely neutral, and underwent a phase change.  We think this happened because the first stars and galaxies turned on, and leaked ionizing photons out into the IGM.  But exactly when it happened and how long it took are open questions, which we might be able to address using Lyα.  If the IGM is neutral, it will be much harder for Lyα photons to be detected.   We now think that this effect is observed, but a lot more work is needed to make sure we understand all of the radiative transfer effects discussed here.

Next up– what does Lyα from the low-redshift, Green Pea galaxies look like?