We present new measurements of the cosmic cold molecular gas evolution out to redshift 6 based on systematic mining of the ALMA public archive in the COSMOS deep field (A 3 COSMOS). Our A 3 COSMOS dataset contains ∼ 700 galaxies (0.3 z 6) with high-confidence ALMA detections in the (sub-)millimeter continuum and multi-wavelength spectral energy distributions (SEDs). Multiple gas mass calibration methods are compared and biases in band conversions (from observed ALMA wavelength to rest-frame Rayleigh-Jeans(RJ)-tail continuum) have been tested. Combining our A 3 COSMOS sample with ∼ 1, 000 CO-observed galaxies at 0 z 4 (75% at z < 0.1), we parameterize galaxies' molecular gas depletion time (τ depl ) and molecular gas to stellar mass ratio (µ molgas ) each as a function of the stellar mass (M ), offset from the star-forming main sequence (∆MS) and cosmic age (or redshift). Our proposed functional form provides a statistically better fit to current data (than functional forms in the literature), and implies a "downsizing" effect (i.e., more-massive galaxies evolve earlier than less-massive ones) and "mass-quenching" (gas consumption slows down with cosmic time for massive galaxies but speeds up for low-mass ones). Adopting galaxy stellar mass functions and applying our µ molgas function for gas mass calculation, we for the first time infer the cosmic cold molecular gas density evolution out to redshift 6 and find agreement with CO blind surveys as well as semi-analytic modeling. These together provide a coherent picture of cold molecular gas, SFR and stellar mass evolution in galaxies across cosmic time.

Dusty star-forming galaxies at high redshift (1 < z < 3) represent the most intense star-forming regions in the universe. Key aspects to these processes are the gas heating and cooling mechanisms, and although it is well known that these galaxies are gas-rich, little is known about the gas excitation conditions. Only a few detailed radiative transfer studies have been carried out owing to a lack of multiple line detections per galaxy. Here we examine these processes in a sample of 24 strongly lensed star-forming galaxies identified by the Planck satellite (LPs) at z ∼ 1.1–3.5. We analyze 162 CO rotational transitions (ranging from J up = 1 to 12) and 37 atomic carbon fine-structure lines ([C i]) in order to characterize the physical conditions of the gas in the sample of LPs. We simultaneously fit the CO and [C i] lines and the dust continuum emission, using two different non-LTE, radiative transfer models. The first model represents a two-component gas density, while the second assumes a turbulence-driven lognormal gas density distribution. These LPs are among the most gas-rich, IR-luminous galaxies ever observed (μ L L IR ( 8 − 1000 μ m ) ∼ 10 13 − 14.6 L ⊙; 〈 μ L M ISM 〉  = (2.7 ± 1.2) × 1012 M ⊙, with μ L ∼ 10–30 the average lens magnification factor). Our results suggest that the turbulent interstellar medium present in the LPs can be well characterized by a high turbulent velocity dispersion ( 〈 ΔV turb 〉  ∼ 100 km s−1) and ratios of gas kinetic temperature to dust temperature 〈 T kin/T d 〉  ∼ 2.5, sustained on scales larger than a few kiloparsecs. We speculate that the average surface density of the molecular gas mass and IR luminosity, Σ M ISM  ∼ 103–4 M ⊙ pc−2 and Σ L IR  ∼ 1011–12 L ⊙ kpc−2, arise from both stellar mechanical feedback and a steady momentum injection from the accretion of intergalactic gas.

We provide a coherent, uniform measurement of the evolution of the logarithmic star formation rate (SFR) -stellar mass (M * ) relation, called the main sequence of star-forming galaxies (MS), for starforming and all galaxies out to z ∼ 5. We measure the MS using mean stacks of 3 GHz radio continuum images to derive average SFRs for ∼ 200,000 mass-selected galaxies at z > 0.3 in the COSMOS field. We describe the MS relation adopting a new model that incorporates a linear relation at low stellar mass (log(M * /M )<10) and a flattening at high stellar mass that becomes more prominent at low redshift (z < 1.5). We find that the SFR density peaks at 1.5 < z < 2 and at each epoch there is a characteristic stellar mass (M * = 1−4×10 10 M ) that contributes the most to the overall SFR density. This characteristic mass increases with redshift, at least to z ∼ 2.5. We find no significant evidence for variations in the MS relation for galaxies in different environments traced by the galaxy number density at 0.3 < z < 3, nor for galaxies in X-ray groups at z ∼ 0.75. We confirm that massive bulgedominated galaxies have lower SFRs than disk-dominated galaxies at a fixed stellar mass at z < 1.2. As a consequence, the increase in bulge-dominated galaxies in the local star-forming population leads to a flattening of the MS at high stellar masses. This indicates that "mass-quenching" is linked with changes in the morphological composition of galaxies at a fixed stellar mass.

We combine high-resolution ALMA and HST/CANDELS observations of 20 submillimeter galaxies (SMGs) predominantly from the AS2UDS survey at z ≃ 2 with bright rest-frame optical counterparts (K s 22.9) to investigate the resolved structural properties of their dust and stellar components. We derive two-dimensional stellar-mass distributions that are inferred from spatial mass-to-light ratio (M/L * ) corrections based on rest-frame optical colors. Due to the high central column densities of dust in our SMGs, our mass distributions likely represent a lower limit to the true central mass density. The centroid positions between the inferred stellar-mass and the dust distributions agree within 1.1 kpc, indicating an overall good spatial agreement between the two components. The majority of our sources exhibit compact dust configurations relative to the stellar component (with a median ratio of effective radii R e,dust /R e, * = 0.6). This ratio does not change with specific star-formation rate (sSFR) over the factor of 30 spanned by our targets, sampling the locus of 'normal' main sequence galaxies up to the starburst regime, log (sSFR/sSFR MS ) ≥ 0.5. Our results imply that massive SMGs are experiencing centrally enhanced star formation unlike typical spiral galaxies in the local Universe. The sizes and stellar densities of our SMGs are in agreement with those of the passive population at z = 1.5, consistent with these systems being the descendants of z ≃ 2 SMGs.

Dust-enshrouded, starbursting, submillimeter galaxies (SMGs) at z3 have been proposed as progenitors of z2 compact quiescent galaxies (cQGs). To test this connection, we present a detailed spatially resolved study of the stars, dust, and stellar mass in a sample of six submillimeter-bright starburst galaxies at z∼4.5. The stellar UV emission probed by HST is extended and irregular and shows evidence of multiple components. Informed by HST, we deblend Spitzer/ IRAC data at rest-frame optical, finding that the systems are undergoing minor mergers with a typical stellar mass ratio of 1:6.5. The FIR dust continuum emission traced by ALMA locates the bulk of star formation in extremely compact regions (median r e =0.70±0.29 kpc), and it is in all cases associated with the most massive component of the mergers (median M M log 10.49 0.32We compare spatially resolved UV slope (β) maps with the FIR dust continuum to study the infrared excess (IRX=L IR /L UV )-β relation. The SMGs display systematically higher IRX values than expected from the nominal trend, demonstrating that the FIR and UV emissions are spatially disconnected. Finally, we show that the SMGs fall on the mass-size plane at smaller stellar masses and sizes than the cQGs at z=2. Taking into account the expected evolution in stellar mass and size between z=4.5 and z=2 due to the ongoing starburst and mergers with minor companions, this is in agreement with a direct evolutionary connection between the two populations.

We present an analysis of the dust attenuation of star forming galaxies at z = 2.5 − 4.0 through the relationship between the UV spectral slope (β), stellar mass (M * ) and the infrared excess (IRX= L IR /L UV ) based on far-infrared continuum observations from the Atacama Large Millimeter/sub-millimeter Array (ALMA). Our study exploits the full ALMA archive over the COSMOS field processed by the A 3 COSMOS team, which includes an unprecedented sample of ∼ 1500 galaxies at z ∼ 3 as primary or secondary targets in ALMA band 6 or 7 observations with a median continuum sensitivity of 126 µJy/beam (1σ). The detection rate is highly mass dependent, decreasing drastically below log(M * /M ) = 10.5. The detected galaxies show that the IRX-β relationship of massive (log M * /M > 10) main sequence galaxies at z = 2.5 − 4.0 is consistent with that of local galaxies, while starbursts are generally offset by ∼ 0.5 dex to larger IRX values. At the low mass end, we derive upper limits on the infrared luminosities through stacking of the ALMA data. The combined IRX-M * relation at log (M * /M ) > 9 exhibits a significantly steeper slope than reported in previous studies at similar redshifts, implying little dust obscuration at log M * /M < 10. However, our results are consistent with early measurements at z ∼ 5.5, indicating a potential redshift evolution between z ∼ 2 and z ∼ 6. Deeper observations targeting low mass galaxies will be required to confirm this finding.

We report the detection of CO(1 − 0) line emission from seven Planck and Herschel selected hyper luminous (L IR(8−1000µm) > 10 13 L ) infrared galaxies with the Green Bank Telescope (GBT ). CO(1 − 0) measurements are a vital tool to trace the bulk molecular gas mass across all redshifts. Our results place tight constraints on the total gas content of these most apparently luminous high-z star-forming galaxies (apparent IR luminosities of L IR > 10 13−14 L ), while we confirm their predetermined redshifts measured using the Large Millimeter Telescope, LMT (z CO = 1.33 − 3.26). The CO(1 − 0) lines show similar profiles as compared to J up = 2 − 4 transitions previously observed with the LMT. We report enhanced infrared to CO line luminosity ratios of < L IR /L CO(1−0) >= 110 ± 22 L (K km s −1 pc −2 ) −1 compared to normal star-forming galaxies, yet similar to those of well-studied IR-luminous galaxies at high-z. We find average brightness temperature ratios of < r 21 >= 0.93 (2 sources), < r 31 >= 0.34 (5 sources), and < r 41 >= 0.18 (1 source). The r 31 and r 41 values are roughly half the average values for SMGs. We estimate the total gas mass content as µM H2 = (0.9 − 27.2) × 10 11 (α CO /0.8)M , where µ is the magnification factor and α CO is the CO line luminosity to molecular hydrogen gas mass conversion factor. The rapid gas depletion times, < τ depl >= 80 Myr, reveal vigorous starburst activity, and contrast the Gyr depletion timescales observed in local, normal star-forming galaxies.

To better constrain the physical mechanisms driving star formation, we present the first systematic study of the radio continuum size evolution of star-forming galaxies (SFGs) over the redshift range 0.35 < z < 2.25. We use the VLA COSMOS 3GHz map (noise rms = 2.3 µJy beam −1 , θ beam = 0.75 arcsec) to construct a mass-complete sample of 3184 radio-selected SFGs that reside on and above the main-sequence (MS) of SFGs. We constrain the overall extent of star formation activity in galaxies by applying a 2D-Gaussian model to their radio continuum emission. Extensive Monte Carlo simulations are used to validate the robustness of our measurements and characterize the selection function. We find no clear dependence between the radio size and stellar mass, M , of SFGs with 10.5 log(M /M ) 11.5. Our analysis suggests that MS galaxies are preferentially extended, while SFGs above the MS are always compact. The median effective radius of SFGs on (above) the MS of R eff = 1.5 ± 0.2 (1.0 ± 0.2) kpc remains nearly constant with cosmic time; a parametrization of the form R eff ∝ (1 + z) α yields a shallow slope of only α = −0.26 ± 0.08 (0.12 ± 0.14) for SFGs on (above) the MS. The size of the stellar component of galaxies is larger than the extent of the radio continuum emission by a factor ∼2 (1.3) at z = 0.5 (2), indicating star formation is enhanced at small radii. The galactic-averaged star formation rate surface density (Σ SFR ) scales with the distance to the MS, except for a fraction of MS galaxies ( 10%) that harbor starburst-like Σ SFR . These "hidden" starbursts might have experienced a compaction phase due to disk instability and/or merger-driven burst of star formation, which may or may not significantly offset a galaxy from the MS. We thus propose to jointly use Σ SFR and distance to the MS to better identify the galaxy population undergoing a starbursting phase.