Home Team Projects Papers PI




RESEARCH PROJECTS

The following are research projects being pursued by the DustyGal team.

Exploiting the SCUBA-2 COSMOS Survey

The SCUBA-2 COSMOS Survey (S2COSMOS) is a collaboration of ~90 scientists across the communities of the East Asian Observatory, as well as the UK and Canada, who run the James Clerk Maxwell Telescope (JCMT). S2COSMOS is an EAO Large Programme to complete and extend the SCUBA-2 coverage of the COSMOS field. This project builds upon the existing S2CLS coverage of COSMOS by completing the coverage of the full area of the field to the depth of the deepest existing S2CLS maps (sigma_850=1.5mJy rms). The survey exploits the immense increase in mapping speed, fidelity and sensitivity of the new SCUBA-2 submillimetre camera on the JCMT. The goal of the survey is to provide the first large samples of extragalactic sources selected in the 850-um waveband. This atmospheric window allows us to access the redshifted far-infrared emission from luminous but highly, high-redshift galaxies and AGN - pin-pointing an intense era of activity in the early Universe associated with the formation of massive galaxies and black holes. This large survey will significantly improve our understanding of submillimetre galaxies, via future exploitation of Atacama Millimeter Array (ALMA), James Webb Space Telescope (JWST) and the Square Kilometer Array (SKA).

 


ALMA follow-up of SCUBA-2 Cosmology Legacy Survey

Building upon the success of the LESS programme, we are now exploiting an even larger area submillimetre survey which was undertaken with the SCUBA-2 submillimetre camera on the James Clerk Maxwell Telescope (JCMT) sited on Mauna Kea, Hawaii. This survey, the SCUBA-2 Cosmology Legacy Survey (S2CLS), is the largest deep submillimetre survey published to date, covering >5 square degrees of sky, 20x the area of ALESS to a similar sensitivity limit. S2CLS was a collaboration of over 100 researchers in the UK, Canada and the Netherlands.

We are exploiting the wide area surveyed by S2CLS to identify a large sample of bright submillimetre sources for study with ALMA to test the conclusion from the ALESS survey. We were awarded ALMA time in Cycles 1, 3 and 4 to obtain maps of the dust continuum in 724 bright submillimetre sources drawn from the S2CLS UDS field. This project, AS2UDS, is on-going but initial results confirm a high level of multiplicity in the bright submillimetre sources.

We anticipate further work with ALMA (as well as the Smithsonian Millimetre Array, SMA, and IRAM's Plateau de Bure Interferometer, PdBI) on S2CLS samples to exploit the large survey area and the rich archival multiwavelength observations which exist for the S2CLS survey fields.

 


ALESS: An ALMA survey of submm galaxies in ECDFS

The ALESS survey is a high-priority Cycle 0 ALMA programme to map the dust emission in a sample of 126 submillimetre galaxies. These galaxies came from an earlier survey used the LABOCA submillimetre camera (on the APEX telescope in Chile) which mapped a moon-sized patch of sky, called the Extended Chandra Deep Field South (ECDFS), to identify all the submillimetre sources within that region. This is a long-term project: planning for the LABOCA survey begin in 2004, with the data published in 2009 and the ALMA observations obtained in 2012. The ALMA observations provide an order-of-magnitude improvement in resolution over the LABOCA discovery map, with an angular resolution of 1.5 arcsec, compared to ~20 arcsec. This improved resolution means that the ALMA maps can precisely pin-point the galaxy emitting the submillimetre flux seen by LABOCA, a process which previously had to use indirect indicators of the likely counterpart (which ALMA has shown were sometimes wrong). The ALMA maps have also demonstrated that the submillimetre sources seen in the low resolution LABOCA map frequently comprise blended emission from multiple submillimetre galaxies. This had important implications for the interpretation of results from low-resolution surveys and their theoretical modelling.

ALESS produced a significant revision in our understanding of submillimetre galaxies (summary paper), with the robust identifications for the multiwavelength counterparts allowing the morphologies, far-infrared luminosities and radio properties, AGN content and redshift distribution. These have shown that typical submillimetre galaxies are less extreme in some of their properties than previous studies would have implied. However, they remain a significant component of the star-forming population at high redshifts and one that is particularly challenging to study due to the combination of their great distance and their high dust content, which makes them very faint in optical wavebands.


High-resolution imaging of starburst galaxies

To understand what is driving the intense starburst activity seen in the brightest submillimetre galaxies we must resolve the processes occuring within these galaxies. This is challenging as the angular sizes of these galaxies at cosmological redshifts is only of order 1 arcsec (corresponding to ~10 kiloparsecs, or 30,000 light years in diameter), thus to resolve the internal structure of the galaxies we need angular resolution of at least ~0.1 arcsec (~1 kpc). This is the angular resolution achievable by Hubble Space Telescope at optical wavelengths, but optical imaging of these very dusty and very faint galaxies is particularly difficult to interpret, so to obtain a clearer view of what is going on within these galaxies we need to image them at ~0.1 arcsec in wavebands where the light is not effected by the dust obscuration: the submillimetre or radio wavebands.

We are therefore pursuing two projects with the UK-based eMERLIN radio interferometer to study samples of submillimetre galaxies at radio wavelengths with angular resolutions of ~0.1 arcsec. The first of these involves a very deep eMERLIN observation of the GOODS-North field taken as part of the eMERGE legacy programme. These observations will yield sensitive high-resolution radio morphologies for a few 10's of submillimetre galaxies detected in this region using SCUBA-2. The second, longer-term, project is to integrate a radio dish at Goonhilly in Cornwall, into eMERLIN to allow higher fidelity imaging of southern fields from S2CLS, in particular the COSMOS field. With this expanded eMERLIN capability we would then undertake a survey of the radio morphologies of bright submillimetre sources from the wide-area COSMOS region, to investigate the variation in radio continuum morphology (and hence gain insights into both the processes responsible for triggering the starburst events, as well as the AGN activity which potentially terminates them).


The faintest high-redshift starbursts

The most massive collapsed structures in the Universe, rich clusters of galaxies, have sufficient mass concentrations in their central regions that it can bend the path of the light from distant, background galaxies seen through the cluster core, as predicted by Einstein's General Theory of Relativity. This bending can act to focus the light onto an observer on Earth, much as a magnifying glass does, providing a much brighter and higher resolution image of the background source. We are using this effect in conjunction with sensitive submillimetre cameras, in particular SCUBA-2 on the James Clerk Maxwell Telescope, to search for highly-magnified examples of the faint submillimetre galaxy population seen through the centres of massive clusters. These rare, highly-magnified sources are intrinsically far fainter than the population detected in current wide-field submillimetre surveys. As such they provide a population which should lie between the most extreme, high-luminosity submillimetre sources and the ``normal'' high-redshift galaxy population, allowing us to trace the variation in star-formation properties and triggering mechanisms between these various classes of galaxy.

Gravitational lensing is a powerful tool to study high-redshift sources. However, its main drawback is that it is very difficult to construct well-defined samples of lensed galaxies for statistical studies, due to the variations induced by the lens amplification in different cluster lenses and for different background sources. For that reason it is still essential to undertake statistical surveys in ``unlensed'' fields (those not containing rich clusters, although of course all galaxies can potentially lens background sources), while employing the lens studies of individual sources to understand the internal phyiscal processes operating in these galaxies on scales and sensitivities which are otherwise unobtainable.


Starburst activity in high-redshift clusters

The majority of galaxies in the densest regions of the local Universe, the central cores of clusters of galaxies, appear to have ceased to form stars a considerable time ago, around 8-10 billion years. This in contrast with the galaxies in the surrounding lower-density ``field'' volumes where most galaxies today are still forming stars. Thus the evolution of the galaxies which are destined to lie in clusters today, which will also lie in overdense regions at earlier times, must be accelerated compared to those which are lower-density environments. This local evidence therefore points to a striking reversal of the behaviour we see today, with the star formation activity been higher in dense regions of the Universe at high redshifts, compared to the lower-density field at the same epoch.

We have used the SCUBA-2 camera on the JCMT to search for evidence of this accelerated activity in dense regions in the early Universe by studying the dusty starburst population in protoclusters at high redshifts. At a redshift of z~1.5, around 9 billion years ago, we find a variety of behaviour in clusters: some show an absence of star-forming galaxies in the very highest-density regions, which are already populated by dead, passive galaxies; other clusters in contrast show strong excesses of submillimetre starburst galaxies in their higher density regions. An example of the latter is shown here, the circle shows a 1-Mpc radius from the cluster centre which contains 11 ultraluminous submillimetre galaxies, a factor of 4x more than expected in a blank field. This variation in the starburst populations in clusters suggests that this epoch is the point where the environmental trends we see in galaxies we see today are first being created.

 


Theoretical models of SMGs

The observational programmes being pursued by the DustyGal team provide the fundamental data needed for phenomenological interpretation of the evolutionary behaviour of extreme starburst galaxies in the distant Universe, the triggering mechanisms responsible for their activity, their relationship to less-extreme star forming galaxies, their response to environmental influences and some of the detailed physics of the star formation process occuring with these galaxies. However, to turn these observations into quantitative tests of the evolution of these galaxies and in particular to tie them directly to the formation of the galaxy populations we see around us in the Universe today, we need to reproduce their behaviour in theoretical models. These models can then be used to link the different populations of galaxies seen at different epochs within a single theoretical framework for galaxy formation and evolution.

Theoretical galaxy formation models have had significant success in reproducing the bulk properties of ``typical'' galaxies over much of the history of the Universe. However, they have had much less success in predicting the properties of the more active starburst sources, such as submillimetre galaxies, where the models consistently underestimate how extreme these systems are. This most likely represents a failure to capture some of the critical small-scale physics occuring in these galaxies, relating to the structural stability of the gas disks or the timescale or frequency of mergers in high redshift galaxies. It is hoped that the latest generation of numerical models of galaxy formation, such as the EAGLE simulation shown here, which have both higher spatial resolution and more accurate microphysics, may address some of these shortcomings and so we are also working to compare our observational results to these new simulations to test their predictions for the most luminous starburst galaxies.