What makes galactic hot DOGs so hot?
Starbursts, active galactic nuclei, and more — these strange transitional galaxies have it all. What’s going on?
On galactic scales, the universe approximately eleven billion years ago was an interesting place. With the oldest galaxies already a few billion years old, many were now going through a dramatic period of collisions and mergers. In the process, large quantities of gas were accreted by supermassive black holes, leading to the rise of quasars and active galactic nuclei (AGNs), while other interstellar clouds spawned a wave of star formation.
Speaking spectroscopically, this period in the universe corresponds to redshifts in the range z ~ 2–3. It turns out that many of the most dramatic galaxies with these redshifts fall into one of two main classes:
- Sub-millimeter galaxies (SMGs), which turn hundreds or thousands of solar masses of gas into stars each year. They have cold dust, and show strong emission around 850 microns.
- Dust-obscured galaxies (DOGs), which are in a transition phase from having their emission dominated by stars to having their emission dominated by an AGN.
Learning how these classes relate to one another is really important for our understanding of how the most massive elliptical galaxies grow and evolve. This meant that astronomers were excited when the Wide-Field Infrared Survey Explorer (WISE) turned up an even rarer and more powerful type of galaxy at these redshifts. Known at first only as W1W2-dropouts, they’ve since been given a new name — hot DOGs — and have proven to be some of the most luminous galaxies in the universe.
How do we find ULIRGs?
WISE was looking for more examples of ultra luminous infrared galaxies (ULIRGs), which shine at trillions of solar luminosities. Both SMGs and DOGs are specific sub-populations of ULIRGs that exist at z ~ 2–3, and the WISE survey was expected to turn up examples of both. The satellite studied targets in four wavelength bands, denoted W1 through W4, which corresponded to wavelengths of 3.4, 4.6, 12, and 22 μm, respectively. ULIRGs should be clearly visible in the W3 and W4 windows, but very faint in W1 and W2.
Sure enough, a number of galaxies turned up that matched these criteria, and were categorized as W1W2-dropouts — W12drops for short. Many had redshifts of z ~ 2–3. Intrigued, astronomers performed follow-up infrared observations at the Caltech Sub-millimeter Observatory (CSO), as well as optical spectroscopy on the Keck I telescope (Wu et al. 2012). The Spitzer Space Telescope was also able to observe candidates in the W1 and W2 windows.
The team then tried to fit a number of different galaxy templates to the confirmed candidates to try to understand what was going on. They modeled something called the spectral energy distribution (SED), which shows how energetic an object is at different wavelengths. Their list of possibilities included quasars, dust torii around AGNs, starburst galaxies like Arp 220, DOGs, and the hybrid AGN-starburst galaxy Markarian 231.
The starburst models were terrible, and while the AGN fits were slightly better, even the dust torus template couldn’t fit all of the key characteristics: relatively weak sub-millimeter emission, a power-law SED at mid-infrared wavelengths, and then a small bump, also in the mid-infrared. There was also a peak in the SED at shorter wavelengths than in other galaxies — indicative of hot dust of 60–120 K. The SED models also showed that W12drops were about 10–20 times more luminous than DOGs, SMGs, or any previously-known type of galaxy at z ~ 2–3. At least one candidate was in excess of an astounding 100 trillion solar luminosities.
How powerful can a hot DOG be?
Before we talk about how W12drops fit into our models of galaxy evolution, let’s talk about that extreme candidate (Eisenhardt et al. 2012). Known as WISE 1814+3412 for short, it was first studied in detail by the same group that analyzed the W12drop galaxies as a whole. Analyzing the candidate turned out to be tricky, because there were actually four main components near each other, labeled A through D. Component B is an interloper — a red dwarf inside the Milky Way — while the other three components do appear to have redshifts of z ~ 2.45. The spectra of A and D are indicative of so-called Lyman break galaxies, a particular type of high-redshift star-forming galaxy, while C appears to be a traditional quasar.
Further analysis indicates that component A is responsible for most of the emission, and therefore truly corresponds to WISE 1814+3412. With no evidence of gravitational lensing, its extreme brightness is intrinsic. About 10% of its luminosity comes from stars, while most of the rest appears to be from an AGN. The trouble with this galaxy is that some typical AGN spectral features are absent, although radio emission adds to the probability that this explanation is correct.
Our best model of the galaxy includes a few hundred billion solar masses of stars that formed over about one billion years. The AGN is powered by a one billion solar mass-black hole — much larger than the one at the center of the Milky Way, but certainly not the most massive one known. What’s interesting is that the ratio of luminosity to black hole mass doesn’t fit known relations. A possible solution is that AGN activity is dominant because star formation has yet to fully pick up, but this poses problems for ULIRG evolutionary models. Regardless, though, WISE 1814+3412 seems to be a hot DOG — that much is agreed upon.
Where do they fit into ULIRG evolution?
DOGs themselves can be classified according to their SEDs. One major sub-population shows a bump in their SEDs from 3–10 μm, while another shows a clear power-law in the mid-IR. The two are thought to be distinct stages of DOG evolution; bump DOGs still have significant star-forming activity, while power-law DOGs have AGNs that are growing in power. In this framework, W12drop galaxies would be so-called “hot DOGs”, each with an extremely powerful AGN surrounded by large quantities of dust. However, the SEDs still require a contribution from cold dust, indicating that star formation is still going strong.
What’s truly exciting is that this means that hot DOGs could lie in a key point on the LIRG evolutionary sequence. Some groups have theorized that many of these galaxies become SMGs with impressive starbursts before becoming DOGs when AGN activity increases. Hot DOGs could lie on the far end of the sequence, with WISE 1814+3412 being a strange exception. Hopefully, we can get a better idea of their place in ULIRG evolution by understanding the trends behind SMGs and DOGs in more detail. Perhaps radio observations can shed even more light on AGN activity.
