Stephen McKay

DSFGs are the most powerfully star-forming galaxies in the Universe, forming stars hundreds times more rapidly than our own Milky Way. These galaxies contain large quantities of cosmic stardust, small particles that absorb ultraviolet and optical starlight and reemit the energy at far-infrared and submillimeter wavelengths. DSFGs played an especially important role around 10 billion years ago, at the time we call cosmic noon, when the total average rate of star formation in the Universe reached its zenith. They are also intimately tied to the formation of filaments and clusters of galaxies, often helping us to locate gravitationally interacting structures made up of tens to hundreds of galaxies.

Data and Observing

I'm an observational astronomer whose research all depends on multiwavelengh data from multiple observatories, so I'm very grateful to have had the opportunity to observe with, or use data from, some of the most powerful telescopes on (and off) the planet. I rely in particular on observations from the best submillimeter telescopes in existence, including the James Clerk Maxwell Telescope (JCMT), the Atacama Large Millimeter/submillimeter Array (ALMA), and the NOrthern Extended Millimeter Array (NOEMA). I also use data from the James Webb Space Telescope (JWST), which is already revealing populations of galaxies formed when the Universe was less than 500 million years old.

I have had the chance to observe both remotely and in person with the Keck I and II telescopes in Hawaii, especially with the MOSFIRE spectrograph. I use MOSFIRE to measure very accurate distances to galaxies by detecting bright redshifted emission lines via spectroscopy.

ALMA (nrao)

JWST (webbtelescope.org)

KECK (keckobservatory.org)

The Complete Redshift Distribution of Typical DSFGs

Measuring accurate redshifts (a proxy for distance) for DSFGs is extremely difficult for a number of compounding observational reasons. As a result, for "typical" DSFGs (down to submillimeter fluxes of around 1-2 mJy), we don't know the full redshift distribution and we almost always have to rely on dubious photometric redshifts (estimating the redshift based on broadband photometry and SED models) for the large majority of galaxies.

I used a combination of deep ALMA linescans and JWST near-infrared spectroscopy to determine spectroscopic redshifts (the gold standard) for a large and deep sample of DSFGs that is unbiased and flux-complete. The spectroscopic redshift distribution of these galaxies is now about 70% complete in total, but for the first time, above 2.5 mJy it is 97% complete.

These spectroscopic redshifts allowed me to test our best photometric redshift methods (all had outlier fractions above 20%) and to put limits on the fraction of DSFGs at very high redshifts (z > 4), which is critical for understanding the early formation of massive galaxies and the build-up of dust in the very early Universe. I also tested different methods for measuring DSFG spectroscopic redshifts in the near-infrared, optical, and millimeter (e.g., ALMA, JWST, ground-based, etc) and found that JWST is remarkably efficient for confirming DSFGs ... if large enough samples can be identified and targeted..

Identifying DSFGs with Red Colors in JWST imaging

The fact that DSFGs usually look faint and red in optical and near-infrared images can actually be used to identify them in the first place. Our group has used red colors in deep JWST near-infrared imaging to pick out these extreme galaxies from the rest of the galaxies in the sky—like finding a needle in a haystack.

Using direct submillimeter imaging with ALMA, I confirmed that this method can identify the correct near-infrared counterpart to the submillimeter source up to 95% of the time.

I used this red color method to select a large (~200) uniform sample of faint DSFGs and studied their physical properties (How massive are they? How fast are they forming stars? How much dust do they have?) and their shapes (Are they disks or spheroids? Are they merging? Do we see bars or spiral arms?).

I found that, on average, these galaxies have large exponential disks with hints of central bulges in formation. Interestingly, the brightest submillimeter galaxies (the most extreme) seem to be similar in mass and shape to the fainter ones (more common), even though we might expect them to be powered by different processes.

Dust Properties of DSFGs using ALMA