Thesis Topic: Ultrared dusty starburst galaxies at the highest redshifts


Thesis Supervisors: Rob Ivison ( and Michele Cirasuolo



It is well established that there exists a population of massive, old, red-and-dead galaxies at z ∼ 2 whose progenitors formed the bulk of their stars in luminous starbursts at z > 4, a phase in their evolution consistent with the most distant dusty star–forming galaxies (DSFGs).  To follow the build-up of the bulk of the stellar mass in the Universe, as well as refining, confirming or refuting galaxy evolution models, we need to identify and confirm the z > 4 tail of DSFGs, determine how they are related to lower-redshift DSFGs, to luminous, red-and-dead galaxies at z ∼ 2 and, ultimately, to luminous ellipticals in the local Universe.  A decade ago, only a handful of DSFGs had been reported to lie at z > 4.  However, the situation changed dramatically with ESA’s Herschel satellite, which uncovered significant samples of DSFGs at z > 4, allowing the study of the relatively unexplored population of dusty starbursts at early epochs: ultrared DSFGs (Ivison et al. 2016).

Previous efforts have gathered a sample of ~100 DSFGs with FIR/submm/mm photometric redshifts of z ∼ 4, but the number of sources observed at a spatial resolution of ~1” remains low, ~50, a mixture of strongly lensed DSFGs, to luminous disk-like galaxies and extreme protoclusters of DSFGs (Oteo et al. 2017).  Having ∼1′′ resolution ALMA imaging at ∼1.2mm of a large sample of ultrared DSFGs is important because this is the only way to improve their photometric redshifts and select the reddest sources (those at the highest redshifts), determine and compare their number density with models, explore the fraction and spatial distribution of lensed sources, and to look for the most extreme starbursts and galaxy protoclusters.

This project will exploit a new A-ranked Cycle 6 ALMA programme which will gather dust continuum observations at a wavelength of 1.2mm (ALMA band 6) for a sample of ∼3000 ultrared DSFGs with z(photometric)  ~ 4.  These data will improve the photometric redshifts and isolate the reddest sources in the sample, those with the highest ALMA versus SPIRE flux density ratios and thus those at the highest redshifts.  These sources will deserve further investigation, e.g. we can measure spectroscopic redshifts via ALMA spectral scans and explore extreme star formation at the epoch of reionisation.  This will also naturally lead to further follow-up in optical and near-IR with the Very Large Telescope (e.g. with the KMOS instrument) to derive the physical conditions of the ionised gas.  Roughly 9% (270 sources) of the sample have Herschel colours redder than HFLS3 and G09–83808 (both at z ~ 6) so we will be observing a significant population of DSFGs at z > 6, probably breaking the current redshift record (SPT0311–58 at z = 6.900 +/- 0.002).  Indeed, 4% of the sources (120 sources) to be observed here have SPIRE colours redder than SPT0311–58, and some are therefore likely to be at even higher redshift.

The spatial resolution of the ALMA observations will also allow us to investigate the fraction of ultrared DSFGs which are resolved into several components and will thus reveal the existence of extreme over-densities of DSFGs, pinpointing likely protocluster cores at z~>4 with extreme star-formation rates, of which currently only two are known.  The observations should translate into at least four refereed publications - paper 1: photometric redshifts and description of the observations; paper 2: ultra-luminous unlensed starbursts; paper 3: extreme proto-clusters; paper 4: number counts and comparison with models, plus others reporting unexpected results which are likely to appear when observing a sample as large as this.