Dusty star-forming galaxies

Vigorous star formation results in the creation of large amounts of dust, which serves to obscure our view of subsequent star formation. This dust absorbs ultraviolet and visible photons emitted by the stars, the photons warm the dust, which then cools by reradiating this energy in the far-infrared waveband (wavelengths around ~ 100 um). Studies of the distant Universe in the far-infrared and sub-millimetre wavebands suggest that dust obscuration is an increasingly significant factor in the early evolution of galaxies, during the first ~3 billion years of the age of the Universe (corresponding to redshifts, z, of z > 2). Perhaps over half of the stars formed in galaxies during this early phase in their growth were obscured from our sight in the visible waveband by dust, with many of them being formed in very active systems, including so-called "ultraluminous infrared galaxies", which have star-formation rates 100 times higher than our Galaxy at the present day. While these galaxies are very faint in the visible waveband, the energy being reradiated by dust in the far-infrared waveband (redshifted into the sub-millimetre) means this population is most easily detected in the millimetre and sub-millimetre wavebands and so facilities such as the Atacama Large Millimeter Array (ALMA) and the Northern Extended Millimeter Array (NOEMA) are important tools to study them.

Understanding the properties of these dusty and luminous systems is critical if we wish to obtain a complete understanding of galaxy formation. The goal of our work is therefore to tackle top-level questions such as: What is the physics of star formation in these systems? and What biases exist in current surveys of obscured high-redshift galaxies and how can we account for them? To make progress on these problems we aim to address specific issues:

Fueling and triggering: is the enhanced star-formation activity in this obscured population driven by mergers or interactions between galaxies in the early Universe, or does it triggered by secular (internally-driven) instabilities in isolated, but very dense, gas-rich disks, or do different processes drive the population in different luminosity ranges (mergers triggering the most luminous, secular bursts being more commonly found in less luminous systems)? We are undertaking dynamical studies of these systems using ALMA and near-infrared integral field spectrographs (including the KAOSS Large Program using the KMOS instrument on VLT) to elucidate the kinemetry of large samples of these galaxies and so identify how they are triggered and whether this process varies as a function of the far-infrared luminosity of the system.

Stellar masses: the combination of high dust obscuration, strong current star formation and potentially complex star-formation histories, as well as the contribution from active galactic nuclei (AGN) in a small fraction of sources, mean that estimates of stellar masses for dusty star-forming galaxies are highly uncertain. This makes it difficult to assess the specific star-formation rates of these sources and hence their relationship to any so-called ``main sequence'' of star-forming galaxies at high redshifts, for this reason we focus on sub-millimetre-flux-limited samples. Nevertheless, we anticipate that high-resolution near- and mid-infrared imaging from the James Webb Space Telescope (JWST) will make important contributions to this problem in the next few years. JWST will be able to spatially resolve the structured dust obscuration within these galaxies, allowing their underlying stellar emission to be more accurately mapped.

Star-formation rates: similarly, the star-formation rates of these obscured galaxies are uncertain due to the combination of high dust obscuration and potential contamination from AGN in some sources. Again this is an area where JWST will have a significant impact, with its sensitive near- and mid-infrared spectroscopy of near-infrared recombination lines and PAH emission. We note that while our focus is on sub-millimetre-flux-limited samples, these have redshift-dependent dust temperature biases and so one element of our work is to investigate the overlap between such samples and truly far-infrared-luminosity-limited surveys. Far-infrared samples detected by the Herschel Space Observatory would be less biased, but are severely limited by the low spatial resolution of those surveys and hence strong confusion, meaning that only rare lensed examples (which have additional biases) of typical dusty star-forming galaxies can be individually studied, with analysis otherwise relying on statistical stacking.

Gas masses - both ALMA and NOEMA are providing measurements of the molecular and atomic gas (CO, [CI] or [CII]) emission in high redshift dust-obscured galaxies. However, the conversion of these measurements into total gas masses for the galaxies relies on a number of uncertain assumptions which fundamentally limit the precision of the estimates for individual galaxies (at the factor of ~ 2 level). Comparable or larger systematic uncertainties apply to attempts to use the dust-to-gas ratio and the cold dust mass to derive gas masses. Similarly, in-direct methods such as the inversion of the Kenicutt-Schmidt relation rely on a reliavke measure of the star-formation rate and the assumption that this relation is the same in these intense, high-redshift systems as it is in local disk galaxies. Obtaining total masses for the systems from dynamical studies is one area where we intend to push hard with ALMA and NOEMA to place limits on the gas mass conversion factors (modulo the uncertainties in the stellar mass and dark matter contributions).

Our work in this area is underpinned by our work on panoramic surveys with the SCUBA-2 sub-millimetre camera on the James Clerk Maxwell Telescope (JCMT) on Mauna Kea, Hawaii, including the SCUBA-2 Cosmology Legacy Survey (S2CLS) and the more recent S2-COSMOS survey, as well as with the LABOCA camera on APEX, which was used for the LESS survey, and exploitation of far-infrared surveys with the Herschel Space Observatory.

We use these wide-field surveys to identify luminous dusty galaxies for subsequent detailed study with a range of multiwavelength tools. However, before we can apply these tools we have to precisely locate the individual galaxies responsible for the far-infrared/sub-millimetre emission. This step can only be reliably undertaken with high-spatial-resolution sub-millimetre interferometry such as ALMA. We have therefore used ALMA in a series of studies to identify samples of sub-millimetre galaxies selected from our S2CLS and LESS surveys. An example of an ALMA sub-millimetre images of one of these systems is shown in the figure. This image is just 0.7 arcsec across and demonstrates the level of detail which ALMA can provide on the dust distribution in these distant galaxies. These studies are already teaching us a lot about the connections between this population and the formation of massive galaxies at high redshifts (e.g. summary paper) and ALMA promises even more exciting discoveries in the next few years.

Durham is one of the leading groups studying high redshift, dust-obscured star forming galaxies, having led one of the first papers reporting the discovery of this population in the late 1990s and we have published over 200 refereed papers on all aspects of this topic since then.

Staff associated with this project are Ian Smail, Mark Swinbank and Dave Alexander.