Discovery of Four Apparently Cold Dusty Galaxies at z = 3.62–5.85 in the COSMOS Field: Direct Evidence of Cosmic Microwave Background Impact on High-redshift Galaxy Observables

Abstract

We report Atacama Large Millimetre Array (ALMA) observations of four high-redshift dusty star-forming galaxy candidates selected from far-Infrared (FIR)/submm observations in the COSMOS field. We securely detect all galaxies in the continuum and spectroscopically confirm them at z=3.62-5.85 using ALMA 3mm line scans, detecting multiple CO and/or [CI] transitions. This includes the most distant dusty galaxy currently known in the COSMOS field, ID85001929 at z=5.847. These redshifts are lower than we had expected as these galaxies have substantially colder dust temperatures than most literature sources at z>4. We provide direct evidence that, given their cold spectral energy distributions, CMB plays a significant role biasing their observed Rayleigh-Jeans (RJ) slopes to unlikely steep values and, possibly, reducing their CO fluxes by a factor of two. We recover standard RJ slopes when the CMB contribution is taken into account. High resolution ALMA imaging shows compact morphology and evidence for mergers. This work reveals a population of cold dusty star-forming galaxies that were under-represented in current surveys, and are even colder than typical Main Sequence galaxies at the same redshift. The observed cold dust temperatures could be a result of low star formation efficiency with rapid metal enrichment or, more likely, evidence for optically thick dust continuum in the FIR. High FIR dust optical depth might be a widespread feature of compact starbursts at any redshift.

COLDz: A High Space Density of Massive Dusty Starburst Galaxies ∼1 Billion Years after the Big Bang

“…j Values obtained by fitting the de-blended fluxes at 100 μm-3 mm using a method virtually identical to that adopted for GN10. The T dust is compatible with the value of 61±8 K reported by Jin et al (2019). k Blended with a lower-z source that contributes ∼30% of the flux at 870 μm and likely a higher fraction at 500 μm.…”

“…Adopting the measured CO(J=1→0) luminosity instead of that inferred from the LVG modeling would yield a ratio of ;5400 for GN10.52 We caution that a direct comparison in terms of physical properties to lower redshifts (as is typically done with photon dominated region models calculated at z=0) is not straight forward due to the reduction in the intensity of low-J CO line emission at z>5 due to the warmer CMB.53 Note that an optically thin fitting procedure suggests a more modest dust temperature of (53±5) K compared to the general fitting result adopted here, but it also yields a worse fit to the data.54 Jin et al (2019) report ID 20010161, a source with a candidate redshift of z=5.051 based on a single emission line. While the identification is plausible, we do not include it in the current data compilation until the redshift is confirmed.55 Uncertainties are given as the median absolute deviation.…”

“…Dust temperature (top left panel) and FIR luminosity (top right panel) of COLDz z>5 DSFGs, compared to literature DSFGs, showing that they cover a broad range in dust temperatures and that they are less luminous than other DSFGs known at z>5. They also probe a luminosity range that is only accessible through gravitational lensing in other current samples with spectroscopic redshifts (dotted dividing line in the top right panel; see Table 6 for references), with the exception of one source discovered recently from de-blended single-dish catalogs (Jin et al 2019). There is a weak trend between dust temperature and FIR luminosity within the current sample, which is largely consistent with a standard L∝T 4 scaling (dashed line in the bottom left panel, shown scaled to the median values and with ±0.5 dex scatter as the shaded region for reference).…”

The ALPINE-ALMA [CII] survey: Data processing, catalogs, and statistical source properties

“…At z > 4, most of our constraints come from redshifted ultraviolet (UV) light, which probes the unobscured star formation rate (SFR). Except for a few very bright objects (e.g., Walter et al 2012;Riechers et al 2013;Watson et al 2015;Capak et al 2015;Strandet et al 2017;Zavala et al 2018;Jin et al 2019;Casey et al 2019), we have much less information about dust-obscured star formation, that is, the UV light absorbed by dust and re-emitted in the far infrared. To accurately measure the star formation history in the Universe, we need to know both the obscured and unobscured parts (e.g., Madau & Dickinson 2014;Maniyar et al 2018).…”

ALMA Deep Field in SSA22

“…For instance, Simpson et al (2014) found ten SMGs which were detected at 3.6 µm, but not detected at shorter wavelengths. While obtaining a spectroscopic redshift of these types of near-infrared-dark SMGs is still not easy, A&A 640, L8 (2020) a handful of these galaxies are now placed at z ∼ 4−5 via molecular and atomic line detections at (sub)millimeter wavelengths and the number is increasing (e.g., Swinbank et al 2012;Riechers et al 2017;Oteo et al 2018;Casey et al 2019;Jin et al 2019;Jiménez-Andrade et al 2020;Riechers et al 2020).…”