Search

The Nancy Grace Roman Space Telescope, NASA's next flagship observatory, will redefine deep-field galaxy survey with a field of view two orders of magnitude larger than Hubble and an angular resolution of matching quality. These future deep-wide galaxy surveys necessitate new simulations to forecast their scientific output and to optimise survey strategies. In this work, we present five realizations of 2-deg^2 lightcones, containing a total of >25 million simulated galaxies with -16 < MUV < -25 spanning z ~ 0 to 10. This dataset enables a new set of experiments with the impacts of survey size on the derived galaxy formation and cosmological constraints. The intrinsic and observable galaxy properties are predicted using a well-established, physics-based semi-analytic modelling approach. We provide forecasts for number density, cosmic SFR, field-to-field variance, and angular two-point correlation functions, and demonstrate how the future wide-field surveys will be able to improve these measurements relative to current generation surveys. We also present a comparison between these lightcones and others that have been constructed with empirical models. The mock lightcones are designed to facilitate the exploration of multi-instrument synergies and connecting with current generation instruments and legacy surveys. In addition to Roman, we also provide photometry for a number of other instruments on upcoming facilities, including Euclid and Rubin, as well as the instruments that are part of many legacy surveys. Full object catalogues and data tables for the results presented in this work are made available through a web-based, interactive portal https://www.simonsfoundation.org/semi-analytic-forecasts., 25 pages, 16 figures. Accepted for publication in MNRAS

Wide, deep, blind continuum surveys at submillimetre/millimetre (submm/mm) wavelengths are required to provide a full inventory of the dusty, distant Universe. However, conducting such surveys to the necessary depth, with sub-arcsec angular resolution, is prohibitively time-consuming, even for the most advanced submm/mm telescopes. Here, we report the most recent results from the ALMACAL project, which exploits the 'free' calibration data from the Atacama Large Millimetre/submillimetre Array (ALMA) to map the lines of sight towards and beyond the ALMA calibrators. ALMACAL has now covered 1,001 calibrators, with a total sky coverage around 0.3 deg2, distributed across the sky accessible from the Atacama desert, and has accumulated more than 1,000h of integration. The depth reached by combining multiple visits to each field makes ALMACAL capable of searching for faint, dusty, star-forming galaxies (DSFGs), with detections at multiple frequencies to constrain the emission mechanism. Based on the most up-to-date ALMACAL database, we report the detection of 186 DSFGs with flux densities down to S870um ~ 0.2mJy, comparable with existing ALMA large surveys but less susceptible to cosmic variance. We report the number counts at five wavelengths between 870um and 3mm, in ALMA bands 3, 4, 5, 6 and 7, providing a benchmark for models of galaxy formation and evolution. By integrating the observed number counts and the best-fitting functions, we also present the resolved fraction of the cosmic infrared background (CIB) and the CIB spectral shape. Combining existing surveys, ALMA has currently resolved about half of the CIB in the submm/mm regime., Comment: 22 pages, 16 figures, accepted for publication in MNRAS

We present the first detection of the [N II] 122 $\mu$m and [O III] 52 $\mu$m lines for a reionisation-epoch galaxy. Based on these lines and previous [C II] 158 $\mu$m and [O III] 88 $\mu$m measurements, using two different radiative transfer models of the interstellar medium, we estimate an upper limit on electron density of $\lesssim$ 500 cm$^{-3}$ and a gas-phase metallicity $Z/Z_\odot \sim 1.1 \pm 0.2$ for A1689-zD1, a gravitationally-lensed, dusty galaxy at $z$ = 7.133. Other measurements or indicators of metallicity so far in galaxy interstellar media at $z \gtrsim$ 6 are typically an order of magnitude lower than this. The unusually high metallicity makes A1689-zD1 inconsistent with the fundamental metallicity relation, although there is likely significant dust obscuration of the stellar mass, which may partly resolve the inconsistency. Given a solar metallicity, the dust-to-metals ratio is a factor of several lower than expected, hinting that galaxies beyond $z \sim$ 7 may have lower dust formation efficiency. Finally, the inferred nitrogen enrichment compared to oxygen, on which the metallicity measurement depends, indicates that star-formation in the system is older than about 250 Myr, pushing the beginnings of this galaxy to $z >$ 10., Comment: 10 pages, 7 figures; Published MNRAS: 06 March 2023

Detecting the line-intensity mapping (LIM) signal from the galaxies of the Epoch of Reionization is an emerging tool to constrain their role in reionization. Ongoing and upcoming experiments target the signal fluctuations across the sky to reveal statistical and astrophysical properties of these galaxies via signal statistics, e.g., the power spectrum. Here, we revisit the [C II]$_{158 \mu \text{m}}$ LIM power spectrum under non-uniform line-luminosity scatter, which has a halo-mass variation of statistical properties. Line-luminosity scatter from a cosmological hydrodynamic and radiative transfer simulation of galaxies at $z=6$ is considered in this study. We test the robustness of different model frameworks that interpret the impact of the line-luminosity scatter on the signal statistics. We use a simple power-law model to fit the scatter and demonstrate that the mean luminosity-halo mass correlation fit cannot preserve the mean intensity of the LIM signal (hence the clustering power spectrum) under non-uniform scatter. In our case, the mean intensity changes by $\sim 48$ per cent compared to the mean correlation fit in contrast to the general case with semi-analytic scatter. However, we find that the prediction for the mean intensity from the most-probable fit can be modelled robustly, considering the generalized and more realistic non-uniform scatter. We also explore the possibility of diminishing luminosity bias under non-uniform scatter, affecting the clustering power spectrum, although this phenomenon might not be statistically significant. Therefore, we should adopt appropriate approaches that can consistently interpret the LIM power spectrum from observations., Comment: 9 pages, 8 figures, comments are welcome, accepted in MNRAS

We study the link between gas flow events and key galaxy scaling relations: the relations between star formation rate (SFR) and stellar mass (the main sequence, MS), gas metallicity and stellar mass (the mass-metallicity relation, MZR) and gas metallicity, stellar mass and SFR (the fundamental metallicity relation, FMR). Using all star-forming galaxies (SFGs) in the 22 MUSE fields of the MusE GAs FLOw and Wind (MEGAFLOW) survey, we derive the MS, MZR and FMR scaling relations for 385 SFGs with $M = 10^8 - 10^{11.5}$ $M_\odot$ at redshifts 0.35 < z < 0.85. Using the MUSE data and complementary X-Shooter spectra at 0.85 < z < 1.4, we determine the locations of 21 SFGs associated with inflowing or outflowing circumgalactic gas (i.e. with strong MgII absorption in background quasar spectra) relative to these scaling relations. Compared to a control sample of galaxies without gas flows (i.e., without MgII absorption within 70 kpc of the quasar), SFGs with inflow events (i.e., MgII absorption along the major axis) are preferentially located above the MS, while SFGs with ouflow events (i.e., MgII absorption along the minor axis) are preferentially more metal rich. Our observations support the scenario in which gas accretion increases the SFR while diluting the metal content and where circumgalactic outflows are found in more metal-rich galaxies., 13 pages, 8 figures

In anticipation of the new era of high-redshift exploration marked by the commissioning of the James Webb Space Telescope (JWST), we present two sets of galaxy catalogues that are designed to aid the planning and interpretation of observing programs. We provide a set of 40 wide-field lightcones with footprints spanning approximately ~ 1,000 sq. arcmin, containing galaxies up to z = 10, and a new set of 8 ultra-deep lightcones with 132 sq. arcmin footprints, containing galaxies up to z ~ 12 down to the magnitudes expected to be reached in the deepest JWST surveys. These mock lightcones are extracted from dissipationless N-body simulations and populated with galaxies using the well-established, computationally efficient Santa Cruz semi-analytic model for galaxy formation. We provide a wide range of predicted physical properties, and simulated photometry from NIRCam and many other instruments. We explore the predicted counts and luminosity functions and angular two-point correlation functions for galaxies in these simulated lightcones. We also explore the predicted field-to-field variance using multiple lightcone realizations. We find that these lightcones reproduce the available measurements of observed clustering from 0.2 < z < 7.5 very well. We provide predictions for galaxy clustering at high redshift that may be obtained from future JWST observations. All of the lightcones presented here are made available through a web-based, interactive data release portal https://flathub.flatironinstitute.org/group/sam-forecasts., Final paper of the Semi-analytic forecasts for JWST series. 24 pages, 14 figures, many years of exciting science to come, accepted for publication in MNRAS

Active galactic nuclei (AGN) forming in the early universe are thought to be the primary source of hard ionizing photons contributing to the reionization of intergalactic helium. However, the number density and spectral properties of high-redshift AGN remain largely unconstrained. In this work, we make use of physically-informed models calibrated with a wide variety of available observations to provide estimates for the role of AGN throughout the Epoch of Reionization. We present AGN luminosity functions in various bands between z = 2 to 7 predicted by the well-established Santa Cruz semi-analytic model, which includes modelling of black hole accretion and AGN feedback. We then combine the predicted AGN populations with a physical spectral model for self-consistent estimates of ionizing photon production rates, which depend on the mass and accretion rate of the accreting supermassive black hole. We then couple the predicted comoving ionizing emissivity with an analytic model to compute the subsequent reionization history of intergalactic helium and hydrogen. This work demonstrates the potential of coupling physically motivated analytic or semi-analytic techniques to capture multi-scale physical processes across a vast range of scales (here, from AGN accretion disks to cosmological scales). Our physical model predicts an intrinsic ionizing photon budget well above many of the estimates in the literature, meaning that helium reionization can comfortably be accomplished even with a relatively low escape fraction. We also make predictions for the AGN populations that are expected to be detected in future \emph{James Webb Space Telescope} surveys., Comment: 28 pages, 23 figures, accepted for publication in MNRAS

We present a multiwavelength study of RXC J2014.8-2430, the most extreme cool-core cluster in the Representative $XMM-Newton$ Cluster Structure Survey (REXCESS), using $Chandra$ X-ray, Southern Astrophysical Research (SOAR) Telescope, Atacama Large Millimeter/submillimeter Array (ALMA), Very Large Array (VLA), and Giant Metrewave Radio Telescope (GMRT) observations. While feedback from an active galactic nucleus (AGN) is thought to be the dominant mechanism by which a cooling flow is suppressed, the $Chandra$ imaging observations surprisingly do not reveal the bi-lateral X-ray cavities expected in the intracluster medium (ICM) of an extreme cool core hosting a powerful radio source. We discuss the limits on the presence of any radio bubbles associated with any undetected X-ray cavities. We place upper limits on any significant X-ray AGN in the brightest cluster galaxy, and show that the X-ray peak is offset from the central radio source, which exhibits a steep low frequency radio spectrum indicative of electron ageing. The SOAR data reveal an extended, luminous emission line source. From our narrowband H$\alpha$ imaging of the BCG, the central H$\alpha$ peak is coincident with the radio observations, yet offset from the X-ray peak, consistent with sloshing found previously in this cluster. ALMA observations reveal a large reservoir of molecular gas that traces the extended H$\alpha$ emission. We conclude either that the radio source and its cavities in the X-ray gas are nearly aligned along the line of sight, or that ram pressure induced by sloshing has significantly displaced the cool molecular gas feeding it, perhaps preempting the AGN feedback cycle. We argue that the sloshing near the core is likely subsonic, as expected, given the co-location of the H$\alpha$, CO(1-0), radio continuum, and stellar emission peaks and their proximity to the intact cool core seen in X-ray., Comment: 20 pages (including the appendix), 15 figures. A&A accepted on 18 July 2022

We aim to quantify the relation between the dust-to-gas mass ratio (DTG) and gas-phase metallicity of $z=$2.1-2.5 luminous galaxies and contrast this high-redshift relation against analogous constraints at z$=$0. We present a sample of ten star-forming main-sequence galaxies in the redshift range $2.1, Comment: Version accepted for publication in A&A

We present observations of CO(3-2) in 13 main-sequence $z=2.0-2.5$ star-forming galaxies at $\log(M_*/M_{\odot})=10.2-10.6$ that span a wide range in metallicity (O/H) based on rest-optical spectroscopy. We find that CO(3-2)/SFR decreases with decreasing metallicity, implying that the CO luminosity per unit gas mass is lower in low-metallicity galaxies at $z\sim2$. We constrain the CO-to-H$_2$ conversion factor ($\alpha_{\text{CO}}$) and find that $\alpha_{\text{CO}}$ inversely correlates with metallicity at $z\sim2$. We derive molecular gas masses ($M_{\text{mol}}$) and characterize the relations among $M_*$, SFR, $M_{\text{mol}}$, and metallicity. At $z\sim2$, $M_{\text{mol}}$ increases and molecular gas fraction ($M_{\text{mol}}$/$M_*$) decrease with increasing $M_*$, with a significant secondary dependence on SFR. Galaxies at $z\sim2$ lie on a near-linear molecular KS law that is well-described by a constant depletion time of 700 Myr. We find that the scatter about the mean SFR-$M_*$, O/H-$M_*$, and $M_{\text{mol}}$-$M_*$ relations is correlated such that, at fixed $M_*$, $z\sim2$ galaxies with larger $M_{\text{mol}}$ have higher SFR and lower O/H. We thus confirm the existence of a fundamental metallicity relation at $z\sim2$ where O/H is inversely correlated with both SFR and $M_{\text{mol}}$ at fixed $M_*$. These results suggest that the scatter of the $z\sim2$ star-forming main sequence, mass-metallicity relation, and $M_{\text{mol}}$-$M_*$ relation are primarily driven by stochastic variations in gas inflow rates. We place constraints on the mass loading of galactic outflows and perform a metal budget analysis, finding that massive $z\sim2$ star-forming galaxies retain only 30% of metals produced, implying that a large mass of metals resides in the circumgalactic medium., Comment: 35 pages, 17 figures, submitted to ApJ

We study the relation between the UV-slope, $\beta$, and the ratio between the infrared- and UV-luminosities (IRX) of galaxies from TNG50, the latest installment of the IllustrisTNG galaxy formation simulations. We select 7280 star-forming main-sequence (SFMS) galaxies with stellar mass $10^9M_\odot$ at redshifts $0 \leq z \leq 4$ and perform radiative transfer with SKIRT to model effects of interstellar medium dust on the emitted stellar light. Assuming a Milky Way (MW) dust type and a dust-to-metal ratio of 0.3, we find that TNG50 SFMS galaxies generally agree with observationally-derived IRX-$\beta$ relations at $z \lesssim 1$. However, we find a redshift-dependent systematic offset with respect to empirically-derived local relations, with the TNG50 IRX-$\beta$ relation shifting towards lower $\beta$ and steppening at higher redshifts. This is partially driven by variations in the dust-uncorrected UV-slope of galaxies, due to different star-formation histories of galaxies selected at different cosmic epochs; we suggest the remainder of the effect is caused by differences in the effective dust attenuation curves (EDACs) of galaxies as a function of redshift. We find a typical galaxy-to-galaxy variation of 0.3 dex in IRX at fixed $\beta$, correlated with intrinsic galaxy properties: galaxies with higher star-formation rates, star-formation efficiencies, gas metallicities and stellar masses exhibit larger IRX values. We demonstrate a degeneracy between stellar age, dust geometry and dust composition: $z=4$ galaxies with a Small Magellanic Cloud dust type follow the same IRX-$\beta$ relation as low-redshift galaxies with MW dust. We provide a redshift-dependent fitting function for the IRX-$\beta$ relation for MW dust based on our models., Comment: The paper has been accepted for publishing in MNRAS

Galaxies forming in low-mass halos are thought to be primarily responsible for reionizing the Universe during the first billion years after the Big Bang. Yet, these halos are extremely inefficient at forming stars in the nearby Universe. In this work, we address this apparent tension, and ask whether a physically motivated model of galaxy formation that reproduces the observed abundance of faint galaxies in the nearby Universe is also consistent with available observational constraints on the reionization history. By interfacing the Santa Cruz semi-analytic model for galaxy formation with an analytic reionization model, we constructed a computationally efficient pipeline that connects `ground-level' galaxy formation physics to `top-level' cosmological-scale observables. Based on photometric properties of the galaxy populations predicted up to $z=15$, we compute the reionization history of intergalactic hydrogen. We quantify the three degenerate quantities that influence the total ionizing photon budget, including the abundance of galaxies, the intrinsic production rate of ionizing photons, and the LyC escape fraction. We explore covariances between these quantities using a Markov chain Monte Carlo method. We find that our locally calibrated model is consistent with all currently available constraints on the reionization history, under reasonable assumptions about the LyC escape fraction. We quantify the fraction of ionizing photons contributed by galaxies of different luminosities and find that the galaxies expected to be detected in JWST NIRCam wide and deep surveys are responsible for producing $\sim 40$-$80\%$ of ionizing photons throughout the EoR. All results presented in this work are available at https://www.simonsfoundation.org/semi-analytic-forecasts-for-jwst/., 21 pages, 15 figures, accepted for publication in MNRAS

The UV bump is a broad absorption feature centered at 2175{\AA} that is seen in the attenuation/extinction curve of some galaxies, but its origin is not well known. Here, we use a sample of 86 star-forming galaxies at z=1.7-2.7 with deep rest-frame UV spectroscopy from the MUSE HUDF Survey to study the connection between the strength of the observed UV 2175{\AA} bump and the Spitzer/MIPS 24 micron photometry, which at the redshift range of our sample probes mid-IR polycyclic aromatic hydrocarbon (PAH) emission at ~6-8 micron. The sample has robust spectroscopic redshifts and consists of typical main-sequence galaxies with a wide range in stellar mass (log(Mstar/Msun) ~ 8.5-10.7) and star formation rates (SFRs; SFR ~ 1-100 Msun/yr). Galaxies with MIPS detections have strong UV bumps, except for those with mass-weighted ages younger than ~150 Myr. We find that the UV bump amplitude does not change with SFR at fixed stellar mass but increases with mass at fixed SFR. The UV bump amplitude and the PAH strength (defined as mid-IR emission normalized by SFR) are highly correlated and both also correlate strongly with stellar mass. We interpret these correlations as the result of the mass-metallicity relationship, such that at low metallicities PAH emission is weak due to a lower abundance of PAH molecules. The weak or complete absence of the 2175{\AA} bump feature on top of the underlying smooth attenuation curve at low mass/metallicities is then expected if the PAH carriers are the main source of the additional UV absorption., Comment: 9 pages, 7 figures, accepted for publication in mnras

The Baldwin, Philips, & Terlevich diagram of [O III]/H$\beta$ vs. [N II]/H$\alpha$ (hereafter N2-BPT) has long been used as a tool for classifying galaxies based on the dominant source of ionizing radiation. Recent observations have demonstrated that galaxies at $z\sim2$ reside offset from local galaxies in the N2-BPT space. In this paper, we conduct a series of controlled numerical experiments to understand the potential physical processes driving this offset. We model nebular line emission in a large sample of galaxies, taken from the SIMBA cosmological hydrodynamic galaxy formation simulation, using the CLOUDY photoionization code to compute the nebular line luminosities from H II regions. We find that the observed shift toward higher [O III]/H$\beta$ and [N II]/H$\alpha$ values at high redshift arises from sample selection: when we consider only the most massive galaxies $M_* \sim 10^{10-11} M_\odot$, the offset naturally appears, due to their high metallicities. We predict that deeper observations that probe lower-mass galaxies will reveal galaxies that lie on a locus comparable to $z\sim 0$ observations. Even when accounting for sample selection effects, we find that there is a subtle mismatch between simulations and observations. To resolve this discrepancy, we investigate the impact of varying ionization parameters, H II region densities, gas-phase abundance patterns, and increasing radiation field hardness on N2-BPT diagrams. We find that either decreasing the ionization parameter or increasing the N/O ratio of galaxies at fixed O/H can move galaxies along a self-similar arc in N2-BPT space that is occupied by high-redshift galaxies., Comment: 25 pages, 15 figures. Accepted for publication in ApJ

The dust abundance of the interstellar medium plays an important role in galaxy physics, the chemical evolution of matter and the absorption and re-emission of stellar light. The last years have seen a surge in observational and theoretical studies constraining the dust-abundance of galaxies up to $z\sim5$. In this work we gather the latest observational measurements (with a focus on absorption studies covering metallicities in the range $6.8 < 12 + \log{(O/H)}, Comment: Accepted for publication in MNRAS. Some corrections added

One of the most fundamental baryonic matter components of galaxies is the neutral atomic hydrogen (H i). At low redshifts, this component can be traced directly through the 21 cm transition, but to infer the H i gas content of the most distant galaxies, a viable tracer is needed. We here investigate the fidelity of the fine-structure transition of the (2 P 3/2 − 2 P 1/3) transition of singly ionized carbon C ii at 158 μm as a proxy for H i in a set simulated galaxies at z ≈ 6, following the work by Heintz et al. We select 11,125 star-forming galaxies from the simba simulations, with far-infrared line emissions postprocessed and modeled within the Sigame framework. We find a strong connection between C ii and H i, with the relation between this C ii-to-H i relation (β [C II]) being anticorrelated with the gas-phase metallicity of the simulated galaxies. We further use these simulations to make predictions for the total baryonic matter content of galaxies at z ≈ 6, and specifically the H i gas mass fraction. We find mean values of M H I/M ⋆ = 1.4 and M H I/M bar,tot = 0.45. These results provide strong evidence for H i being the dominant baryonic matter component by mass in galaxies at z ≈ 6.

We present predictions for the extent of the dust-continuum emission of thousands of main-sequence galaxies drawn from the TNG50 simulation between $z=1-5$. To this aim, we couple the radiative transfer code SKIRT to the output of the TNG50 simulation and measure the dust-continuum half-light radius of the modeled galaxies, assuming a Milky Way dust type and a metallicity dependent dust-to-metal ratio. The dust-continuum half-light radius at observed-frame 850 $\mu$m is up to $\sim$75 per cent larger than the stellar half-mass radius, but significantly more compact than the observed-frame 1.6 $\mu$m (roughly corresponding to H-band) half-light radius, particularly towards high redshifts: the compactness compared to the 1.6 $\mu$m emission increases with redshift. This is driven by obscuration of stellar light from the galaxy centres, which increases the apparent extent of 1.6 $\mu$m disk sizes relative to that at 850 $\mu$m. The difference in relative extents increases with redshift because the observed-frame 1.6 $\mu$m emission stems from ever shorter wavelength stellar emission. These results suggest that the compact dust-continuum emission observed in $z>1$ galaxies is not (necessarily) evidence of the buildup of a dense central stellar component. We also find that the dust-continuum half-light radius very closely follows the radius containing half the star formation in galaxies, indicating that single band dust-continuum emission is a good tracer of the location of (obscured) star formation. The dust-continuum emission is more compact than the H2 mass (for galaxies at $z\geq 2$) and the underlying dust mass. The dust emission strongly correlates with locations with the highest dust temperatures, which do not need to be the locations where most H$_2$ and/or dust is located. The presented results are a common feature of main-sequence galaxies., Comment: Submitted to MNRAS

A number of new $z>11$ galaxy candidates have recently been identified based on public James Webb Space Telescope (JWST) NIRCam observations. Spectroscopic confirmation of these candidates is necessary to robustly measure their redshift and put them in the context of our understanding of the buildup of galaxies in the early Universe. GLASS-z13 is one of these candidates, with a reported photometric redshift $z>11.9$. I present publicly available Atacama Large Millimeter/submillimeter Array (ALMA) band 6 Director's Discretionary Time observations (project 2021.A.00020.S; PI T. Bakx), taken to acquire a spectroscopic redshift for GLASS-z13 by searching for [OIII] line emission in the redshift range $z=11.9-13.5$. No [OIII] emission is detected in integrated spectra extracted within an aperture around GLASS-z13, nor when using an automated line finding algorithm (applying different uv-weighting strategies for the imaging). 1.2 mm continuum emission associated to GLASS-z13 is not detected either. If GLASS-z13 is at z$\approx$12-13, this implies a 3-$\sigma$ upper limit on the [OIII] and rest-frame $\sim$90 $\mu$m continuum emission of $\sim1\times10^8\,\rm{L}_\odot$ and 10.8 $\mu$Jy, respectively. The non-detection of [OIII] and continuum emission does not necessarily imply that GLASS-z13 is not at $z\approx12-13$. It can also be explained by a low metallicity ($\sim 0.2\,\rm{Z}_\odot$ or lower) and/or high-density (at least 100 $\rm{cm}^{-3}$) interstellar medium. This work demonstrates the synergy between ALMA and JWST to study the properties of the first galaxies, although JWST/NIRSpec spectroscopy will be necessary to confirm or reject the high photometric-redshift of GLASS-z13., Comment: Version accepted for publication in A&A

In this invited talk, we discuss the physics of the lifecycle of dust in the context of galaxy formation simulations. After outlining the basic physical processes, we apply algorithms for the formation, growth, and destruction of dust in the ISM to a state-of-the-art cosmological simulation to develop a model for the evolution of the dust to gas and dust to metals ratios in galaxies. We show that while modern simulations are able to match the observed dust mass function at redshift z = 0, most models underpredict the observed mass function at high-redshift (z = 2). We then show the power of these techniques by expanding our model to include a spectrum of dust sizes, and make initial predictions for extinction laws in local galaxies.

We present the first results from the 2mm Mapping Obscuration to Reionization (MORA) survey, the largest ALMA contiguous blank-field survey to-date with a total area of 184 sq. arcmin and the only at 2mm to search for dusty star-forming galaxies (DSFGs). We use the 13 sources detected above 5sigma to estimate the first ALMA galaxy number counts at this wavelength. These number counts are then combined with the state-of-the-art galaxy number counts at 1.2mm and 3mm and with a backward evolution model to place constraints on the evolution of the IR luminosity function and dust-obscured star formation in the last 13 billion years. Our results suggest a steep redshift evolution on the space density of DSFGs and confirm the flattening of the IR luminosity function at faint luminosities, with a slope of $\alpha_{LF} = -0.42^{+0.02}_{-0.04}$. We conclude that the dust-obscured component, which peaks at z=2-2.5, has dominated the cosmic history of star formation for the past ~12 billion years, back to z~4. At z=5, the dust-obscured star formation is estimated to be ~35% of the total star formation rate density and decreases to 25%-20% at z=6-7, implying a minor contribution of dust-enshrouded star formation in the first billion years of the Universe. With the dust-obscured star formation history constrained up to the end of the epoch of reionization, our results provide a benchmark to test galaxy formation models, to study the galaxy mass assembly history, and to understand the dust and metal enrichment of the Universe at early times., Comment: Accepted for publication in The Astrophysical Journal

The Millimeter-wave Intensity Mapping Experiment (mmIME) recently reported a detection of excess spatial fluctuations at a wavelength of 3 mm, which can be attributed to unresolved emission of several CO rotational transitions between $z\sim1-5$. We study the implications of this data for the high-redshift interstellar medium using a suite of state-of-the-art semianalytic simulations which have successfully reproduced many other sub-millimeter line observations across the relevant redshift range. We find that the semianalytic predictions are mildly in tension with the mmIME result, with a predicted CO power $\sim3.5\sigma$ below what was observed. We explore some simple modifications to the models which could resolve this tension. Increasing the molecular gas abundance at the relevant redshifts to $\sim10^8\ M_\odot\ \rm{Mpc}^{-3}$, a value well above that obtained from directly imaged sources, would resolve the discrepancy, as would assuming a CO-$H_2$ conversion factor $\alpha_{\rm{CO}}$ of $\sim1.5\ M_{\odot}$ K$^{-1}$ $(\rm{km}/\rm{s})^{-1}$ pc$^{2}$, a value somewhat lower than is commonly assumed. We go on to demonstrate that these conclusions are quite sensitive to the detailed assumptions of our simulations, highlighting the need for more careful modeling efforts as more intensity mapping data become available., Comment: 15 pages, 8 figures, submitted to ApJ

Sub-millimeter emission lines produced by the interstellar medium (ISM) are strong tracers of star formation and are some of the main targets of line intensity mapping (LIM) surveys. In this work we present an empirical multi-line emission model that simultaneously covers the mean, scatter, and correlations of [CII], CO J=1-0 to J=5-4, and [CI] lines in redshift range $1\leq z\leq9$. We assume the galaxy ISM line emission luminosity versus halo mass relations can be described by double power laws with redshift-dependent log normal scatter. The model parameters are then derived by fitting to the state of the art semi-analytic simulation results that have successfully reproduced multiple sub-millimeter line observations at $0\leq z\lesssim6$. We cross check the line emission statistics predicted by the semi-analytic simulation and our empirical model, finding that at $z\geq1$ our model reproduces the simulated line intensities with fractional error less than about 10%. The fractional difference is less than 25% for the power spectra. Grounded on physically-motivated and self-consistent galaxy simulations, this computationally efficient model will be helpful in forecasting ISM emission line statistics for upcoming LIM surveys., 16 pages, 10 figures. Correct a typo in Eq 12. Add Figure 7

We report molecular gas mass estimates obtained from a stacking analysis of CO line emission in the ALMA Spectroscopic Survey (ASPECS) using the spectroscopic redshifts from the optical integral field spectroscopic survey by the Multi Unit Spectroscopic Explorer (MUSE) of the Hubble Ultra Deep Field (HUDF). Stacking was performed on subsets of the sample of galaxies classified by their stellar mass and position relative to the main-sequence relation (on, above, below). Among all the CO emission lines, from CO(2-1) to CO(6-5), with redshifts accessible via the ASPECS Band 3 and the MUSE data, CO(2-1) provides the strongest constraints on the molecular gas content. We detect CO(2-1) emission in galaxies down to stellar masses of $\mathrm{log}({M}_{* }/{M}_{\odot })=10.0$ . Below this stellar mass, we present a new constraint on the molecular gas content of $z\sim 1.5$ main-sequence galaxies by stacking based on the MUSE detections. We find that the molecular gas mass of main-sequence galaxies continuously decreases with stellar mass down to $\mathrm{log}({M}_{* }/{M}_{\odot })\approx 9.0$ . Assuming a metallicity-based CO-to-H2 conversion factor, the molecular gas-to-stellar mass ratio from $\mathrm{log}({M}_{* }/{M}_{\odot })\sim 9.0$ to ˜10.0 does not seem to decrease as fast as for $\mathrm{log}({M}_{* }/{M}_{\odot })\gt 10.0$ , which is in line with simulations and studies at lower redshift. The inferred molecular gas density $\rho ({{\rm{H}}}_{2})=(0.49\pm 0.09)\times {10}^{8}\,{{\rm{M}}}_{\odot }\,{\mathrm{Mpc}}^{-3}$ of MUSE-selected galaxies at $z\sim 1.5$ is comparable with the one derived in the HUDF with a different CO selection. Using the MUSE data we recover most of the CO emission in our deep ALMA observations through stacking, demonstrating the synergy between volumetric surveys obtained at different wave bands.

We make use of sensitive (9.3 microJy/beam RMS) 1.2mm-continuum observations from the ASPECS ALMA large program of the Hubble Ultra Deep Field (HUDF) to probe dust-enshrouded star formation from 1362 Lyman-break galaxies spanning the redshift range z=1.5-10 (to ~7-28 Msolar/yr at 4 sigma over the entire range). We find that the fraction of ALMA-detected galaxies in our z=1.5-10 samples increases steeply with stellar mass, with the detection fraction rising from 0% at 10^9 Msolar to 85(-18)(+9)% at >10^{10} Msolar. Moreover, stacking all 1253 low-mass (10^{9.5} M_{solar} and a SMC-like relation at lower masses. Using stellar-mass and beta measurements for z~2 galaxies over CANDELS, we derive a new empirical relation between beta and stellar mass and then use this correlation to show that our IRX-beta and IRX-stellar mass relations are consistent with each other. We then use these constraints to express the infrared excess as a bivariate function of beta and stellar mass. Finally, we present updated estimates of star-formation rate density determinations at z>3, leveraging current improvements in the measured infrared excess and recent probes of ultra-luminous far-IR galaxies at z>2., 32 pages, 22 figures, 11 pages, accepted for publication in the Astrophysical Journal, key results shown in Figures 7, 8, 11, and 20, updated to match version in press

We present the first predictions for the $L_{\rm [CII]}$ - SFR relation and [CII] luminosity function (LF) in the Epoch of Reionization (EoR) based on cosmological hydrodynamics simulations using the SIMBA suite plus radiative transfer calculations via S\'IGAME. The sample consists of 11,137 galaxies covering halo mass $\log M_{\rm halo}\in$[9, 12.4] $M_\odot$, star formation rate SFR$\in$[0.01, 330] $M_\odot$ yr$^{-1}$, and metallicity $_{\rm SFR}\in$[0.1, 0.9] $Z_\odot$. The simulated $L_{\rm [CII]}$-SFR relation is consistent with the range observed, but with a spread of $\simeq$0.3 dex at the high end of SFR ($>$100 $M_\odot$ yr$^{-1}$) and $\simeq$0.6 dex at the lower end, and there is tension between our predictions and the values of $L_{\rm [CII]}$ above 10$^{8.5}$ $L_\odot$ observed in some galaxies reported in the literature. The scatter in the $L_{\rm [CII]}$-SFR relation is mostly driven by galaxy properties, such that at a given SFR, galaxies with higher molecular gas mass and metallicity have higher $L_{\rm [CII]}$. The [CII] LF predicted by SIMBA is consistent with the upper limits placed by the only existing untargeted flux-limited [CII] survey at the EoR (ASPECS) and those predicted by semi-analytic models. We compare our results with existing models and discuss differences responsible for the discrepant slopes in the $L_{\rm [CII]}$-SFR relatiion., Comment: 19 pages, 14 figures, Accepted by ApJ

Using the deepest 1.2 mm continuum map to date in the Hubble Ultra Deep Field obtained as part of the ALMA Spectroscopic Survey (ASPECS) large program, we measure the cosmic density of dust and implied gas (H$_{2}+$H I) mass in galaxies as a function of look-back time. We do so by stacking the contribution from all $H$-band selected galaxies above a given stellar mass in distinct redshift bins, $��_{\rm dust}(M_\ast>M,z)$ and $��_{\rm gas}(M_\ast>M,z)$. At all redshifts, $��_{\rm dust}(M_\ast>M,z)$ and $��_{\rm gas}(M_\ast>M,z)$ grow rapidly as $M$ decreases down to $10^{10}\,M_\odot$, but this growth slows down towards lower stellar masses. This flattening implies that at our stellar mass-completeness limits ($10^8\,M_\odot$ and $10^{8.9}\,M_\odot$ at $z\sim0.4$ and $z\sim3$), both quantities converge towards the total cosmic dust and gas mass densities in galaxies. The cosmic dust and gas mass densities increase at early cosmic time, peak around $z\sim2$, and decrease by a factor $\sim4$ and 7, compared to the density of dust and molecular gas in the local universe, respectively. The contribution of quiescent galaxies -- i.e., with little on-going star-formation -- to the cosmic dust and gas mass densities is minor ($\lesssim10\%$). The redshift evolution of the cosmic gas mass density resembles that of the star-formation rate density, as previously found by CO-based measurements. This confirms that galaxies have relatively constant star-formation efficiencies (within a factor $\sim2$) across cosmic time. Our results also imply that by $z\sim0$, a large fraction ($\sim90\%$) of dust formed in galaxies across cosmic time has been destroyed or ejected to the intergalactic medium., 32 pages, 13 figures, Accepted for publication in ApJ

We present a new semi-empirical model for the dust continuum number counts of galaxies at 1.1 millimeter and 850 \micron. Our approach couples an observationally motivated model for the stellar mass and SFR distribution of galaxies with empirical scaling relations to predict the dust continuum flux density of these galaxies. Without a need to tweak the IMF, the model reproduces the currently available observations of the 1.1 millimeter and 850 \micron number counts, including the observed flattening in the 1.1 millimeter number counts below 0.3 mJy \citep{Gonzalez2019numbercounts} and the number counts in discrete bins of different galaxy properties. Predictions of our work include : (1) the galaxies that dominate the number counts at flux densities below 1 mJy (3 mJy) at 1.1 millimeter (850 $\mu$m) have redshifts between $z=1$ and $z=2$, stellar masses of $\sim 5\times10^{10}~\rm{M}_\odot$, and dust masses of $\sim 10^{8}~\rm{M}_\odot$; (2) the flattening in the observed 1.1 millimeter number counts corresponds to the knee of the 1.1 millimeter luminosity function. A similar flattening is predicted for the number counts at 850 $\mu$m; (3) the model reproduces the redshift distribution of current 1.1 millimeter detections; (4) to efficiently detect large numbers of galaxies through their dust continuum, future surveys should scan large areas once reaching a 1.1 millimeter flux density of 0.1 mJy rather than integrating to fainter fluxes. Our modeling framework also suggests that the amount of information on galaxy physics that can be extracted from the 1.1 millimeter and 850 $\mu$m number counts is almost exhausted., Comment: Accepted for publication in ApJ

Measurements of the cosmic far-infrared background (CIB) indicate that emission from many extragalactic phenomena, including star formation and black hole accretion, in the Universe can be obscured by dust. Resolving the CIB to study the population of galaxies in which this activity takes place is a major goal of submillimetre astronomy. Here, we present interferometric 650$\mu$m submillimetre number counts. Using the Band 8 data from the ALMACAL survey, we have analysed 81 ALMA calibrator fields together covering a total area of 5.5~arcmin$^2$. The typical central rms in these fields is $\sim 100 \mu$Jy~beam$^{-1}$ with the deepest maps reaching $\sigma = 47 \mu$Jy~beam$^{-1}$ at sub-arcsec resolution. Multi-wavelength coverage from ALMACAL allows us to exclude contamination from jets associated with the calibrators. However, residual contamination by jets and lensing remain a possibility. Using a signal-to-noise threshold of $4.5\sigma$, we find 21 dusty, star-forming galaxies with 650$\mu$m flux densities of $\geq 0.7 $mJy. At the detection limit we resolve $\simeq 100$ per cent of the CIB at 650$\mu$m, a significant improvement compared to low resolution studies at similar wavelength. We have therefore identified all the sources contributing to the EBL at 650 microns and predict that the contribution from objects with flux 0.7, Comment: 10 pages, 5 figures, accepted for publication in MNRAS

23 pages, 11 figures, 5 tables. Accepted for publication in the Astrophysical Journal; International audience; We present a CO and atomic fine-structure line luminosity function analysis using the ALMA Spectroscopic Survey in the Hubble Ultra Deep Field (ASPECS). ASPECS consists of two spatially-overlapping mosaics that cover the entire ALMA 3mm and 1.2mm bands. We combine the results of a line candidate search of the 1.2mm data cube with those previously obtained from the 3mm cube. Our analysis shows that $\sim$80% of the line flux observed at 3mm arises from CO(2-1) or CO(3-2) emitters at $z$=1-3 (`cosmic noon'). At 1.2mm, more than half of the line flux arises from intermediate-J CO transitions ($J_{\rm up}$=3-6); $\sim12$% from neutral carbon lines; and $< 1$% from singly-ionized carbon, [CII]. This implies that future [CII] intensity mapping surveys in the epoch of reionization will need to account for a highly significant CO foreground. The CO luminosity functions probed at 1.2mm show a decrease in the number density at a given line luminosity (in units of $L'$) at increasing $J_{\rm up}$ and redshift. Comparisons between the CO luminosity functions for different CO transitions at a fixed redshift reveal sub-thermal conditions on average in galaxies up to $z\sim 4$. In addition, the comparison of the CO luminosity functions for the same transition at different redshifts reveals that the evolution is not driven by excitation. The cosmic density of molecular gas in galaxies, $\rho_{\rm H2}$, shows a redshift evolution with an increase from high redshift up to $z\sim1.5$ followed by a factor $\sim 6$ drop down to the present day. This is in qualitative agreement with the evolution of the cosmic star-formation rate density, suggesting that the molecular gas depletion time is approximately constant with redshift, after averaging over the star-forming galaxy population.

International audience; We compare the extent of the dust, molecular gas and stars in three star-forming galaxies, at $z= 1.4, 1.6$ and $2.7$, selected from the Hubble Ultra Deep Field based on their bright CO and dust-continuum emission as well as their large rest-frame optical sizes. The galaxies have high stellar masses, $\mathrm{M}_*>10^{11}\mathrm{M}_\odot$, and reside on, or slightly below, the main sequence of star-forming galaxies at their respective redshifts. We probe the dust and molecular gas using subarcsecond Atacama Large Millimeter/submillimeter Array observations of the 1.3 mm continuum and CO line emission, respectively, and probe the stellar distribution using \emph{Hubble Space Telescope} observations at 1.6 \textmu m. We find that for all three galaxies the CO emission appears $\gtrsim 30\%$ more compact than the stellar emission. For the $z= 1.4$ and $2.7$ galaxies, the dust emission is also more compact, by $\gtrsim 50\%$, than the stellar emission, whereas for the $z=1.6$ galaxy, the dust and stellar emission have similar spatial extents. This similar spatial extent is consistent with observations of local disk galaxies. However, most high redshift observations show more compact dust emission, likely due to the ubiquity of central starbursts at high redshift and the limited sensitivity of many of these observations. Using the CO emission line, we also investigate the kinematics of the cold interstellar medium in the galaxies, and find that all three have kinematics consistent with a rotation-dominated disk.

Sub-millimeter emission lines are important tracers of the cold gas and ionized environments of galaxies and are the targets for future line intensity mapping surveys. Physics-based simulations that predict multiple emission lines arising from different phases of the interstellar medium are crucial for constraining the global physical conditions of galaxies with upcoming LIM observations. In this work we present a general framework for creating multitracer mock sub-millimeter line intensity maps based on physically grounded galaxy formation and sub-mm line emission models. We simulate a mock lightcone of 2 deg$^2$ over a redshift range $0\leq z\leq10$, comprising discrete galaxies and galaxy CII, CO, CI emission. We present simulated line intensity maps for two fiducial surveys with resolution and observational frequency windows representative of COMAP and EXCLAIM. We show that the star formation rate and line emission scaling relations predicted by our simulation significantly differ at low halo masses from widely used empirical relations, which are often calibrated to observations of luminous galaxies at lower redshifts. We show that these differences lead to significant changes in key summary statistics used in intensity mapping, such as the one point intensity probability density function and the power spectrum. It will be critical to use more realistic and complex models to forecast the ability of future line intensity mapping surveys to measure observables such as the cosmic star formation rate density., Comment: 19 pages, 11 figures

We present results from Atacama Large Millimeter/submillimeter Array (ALMA) 1.2 mm continuum observations of a sample of 27 star-forming galaxies at z = 2.1–2.5 from the MOSFIRE Deep Evolution Field survey with metallicity and star formation rate measurements from optical emission lines. Using stacks of Spitzer, Herschel, and ALMA photometry (rest frame ∼8–400 μm), we examine the infrared (IR) spectral energy distributions (SED) of z ∼ 2.3 subsolar-metallicity (∼0.5 Z ⊙) luminous infrared galaxies (LIRGs). We find that the data agree well with an average template of higher-luminosity local low-metallicity dwarf galaxies (reduced χ 2 = 1.8). When compared with the commonly used templates for solar-metallicity local galaxies or high-redshift LIRGs and ultraluminous IR galaxies, even in the most favorable case (with reduced χ 2 = 2.8), the templates are rejected at >98% confidence. The broader and hotter IR SED of both the local dwarfs and high-redshift subsolar-metallicity galaxies may result from different grain properties or a harder/more intense ionizing radiation field that increases the dust temperature. The obscured star formation rate (SFR) indicated by the far-IR emission of the subsolar-metallicity galaxies is only ∼60% of the total SFR, considerably lower than that of the local LIRGs with ∼96%–97% obscured fractions. Due to the evolving IR SED shape, the local LIRG templates fit to mid-IR data overestimate the Rayleigh–Jeans tail measurements by a factor of 2–20. These templates underestimate IR luminosities if fit to the observed ALMA fluxes by >0.4 dex. At a given stellar mass or metallicity, dust masses at z ∼ 2.3 are an order of magnitude higher than z ∼ 0. Given the predicted molecular gas fractions, the observed z ∼ 2.3 dust-to-stellar mass ratios suggest lower dust-to-molecular gas masses than in local galaxies with similar metallicities.

The advent of new sub-millimeter observational facilities has stimulated the desire to model the sub-mm line emission of galaxies within cosmological galaxy formation models. This is typically done in post-processing by applying sub-resolution recipes to describe the properties of the unresolved interstellar medium. At the same time, while there is freedom in how one implements these sorts of recipes, the impact of various choices has yet to be systematically explored in simulations. In this paper, we do just that. We combine a semi-analytic model of galaxy formation with chemical equilibrium networks and numerical radiative transfer models and explore how different choices for the sub-grid modeling affect the predicted CO, [CI], and [CII] emission of galaxies. We find that a key component for a successful model includes a molecular cloud mass-size relation and scaling for the ultraviolet and cosmic ray radiation field that varies based on local ISM properties. Our most successful model adopts a Plummer profile for the radial density distribution of gas within molecular clouds. On the other hand, different assumptions for the clumping of gas within molecular clouds and changes in the slope of the molecular cloud mass distribution function hardly affect the CO, [CI], and [CII] luminosities of galaxies. At fixed star-formation rate the [CII]-SFR ratio of galaxies scales inversely with the pressure acting on molecular clouds, increasing the molecular clouds density and hence decreasing the importance of [CII] line cooling. Overall we find that it is essential that a wide range of sub-mm emission lines arising in vastly different phases of the ISM are used as model constraints in order to limit the freedom in sub-grid choices and we present the details of the most successful model variant., accepted for publication in MNRAS

The initial mass function (IMF) is a key ingredient in many studies of galaxy formation and evolution. Although the IMF is often assumed to be universal, there is continuing evidence that it is not universal. Spectroscopic studies that derive the IMF of the unresolved stellar populations of a galaxy often assume that this spectrum can be described by a single stellar population (SSP). To alleviate these limitations, in this paper we have developed a unique hierarchical Bayesian framework for modelling composite stellar populations (CSPs). Within this framework we use a parameterized IMF prior to regulate a direct inference of the IMF. We use this new framework to determine the number of SSPs that is required to fit a set of realistic CSP mock spectra. The CSP mock spectra that we use are based on semi-analytic models and have an IMF that varies as a function of stellar velocity dispersion of the galaxy. Our results suggest that using a single SSP biases the determination of the IMF slope to a higher value than the true slope, although the trend with stellar velocity dispersion is overall recovered. If we include more SSPs in the fit, the Bayesian evidence increases significantly and the inferred IMF slopes of our mock spectra converge, within the errors, to their true values. Most of the bias is already removed by using two SSPs instead of one. We show that we can reconstruct the variable IMF of our mock spectra for signal-to-noise ratios exceeding $\sim$75., Comment: Accepted for publication in MNRAS, 16 pages, 8 figures

We analyze the interstellar medium properties of a sample of sixteen bright CO line emitting galaxies identified in the ALMA Spectroscopic Survey in the Hubble Ultra Deep Field (ASPECS) Large Program. This CO$-$selected galaxy sample is complemented by a couple of additional CO line emitters in the UDF that are identified based on their MUSE optical spectroscopic redshifts. The ASPECS CO$-$selected galaxies cover a larger range of star-formation rates and stellar masses compared to literature CO emitting galaxies at $z>1$ for which scaling relations have been established previously. Most of ASPECS CO-selected galaxies follow these established relations in terms of gas depletion timescales and gas fractions as a function of redshift, as well as the star-formation rate-stellar mass relation (`galaxy main sequence'). However, we find that $\sim30\%$ of the galaxies (5 out of 16) are offset from the galaxy main sequence at their respective redshift, with $\sim12\%$ (2 out of 16) falling below this relationship. Some CO-rich galaxies exhibit low star-formation rates, and yet show substantial molecular gas reservoirs, yielding long gas depletion timescales. Capitalizing on the well-defined cosmic volume probed by our observations, we measure the contribution of galaxies above, below, and on the galaxy main sequence to the total cosmic molecular gas density at different lookback times. We conclude that main sequence galaxies are the largest contributor to the molecular gas density at any redshift probed by our observations (z$\sim$1$-$3). The respective contribution by starburst galaxies above the main sequence decreases from z$\sim$2.5 to z$\sim$1, whereas we find tentative evidence for an increased contribution to the cosmic molecular gas density from the passive galaxies below the main sequence., Accepted for publication in ApJ. Only minor differences respect to previous version

The ALMA Spectroscopic Survey in the Hubble Ultra Deep Field (ASPECS) provides new constraints for galaxy formation models on the molecular gas properties of galaxies. We compare results from ASPECS to predictions from two cosmological galaxy formation models: the IllustrisTNG hydrodynamical simulations and the Santa Cruz semi-analytic model (SC SAM). We explore several recipes to model the H$_2$ content of galaxies, finding them to be consistent with one another, and take into account the sensitivity limits and survey area of ASPECS. For a canonical CO-to-H$_2$ conversion factor of $��_{\rm CO} = 3.6\,\rm{M}_\odot/(\rm{K}\,\rm{km/s}\,\rm{pc}^{2})$ the results of our work include: (1) the H$_2$ mass of $z>1$ galaxies predicted by the models as a function of their stellar mass is a factor of 2-3 lower than observed; (2) the models do not reproduce the number of H$_2$-rich ($M_{\rm H2} > 3\times 10^{10}\,\rm{M}_\odot$) galaxies observed by ASPECS; (3) the H$_2$ cosmic density evolution predicted by IllustrisTNG (the SC SAM) is in tension (only just agrees) with the observed cosmic density, even after accounting for the ASPECS selection function and field-to-field variance effects. The tension between models and observations at $z>1$ can be alleviated by adopting a CO-to-H$_2$ conversion factor in the range $��_{\rm CO} = 2.0 - 0.8\,\rm{M}_\odot/(\rm{K}\,\rm{km/s}\,\rm{pc}^{2})$. Additional work on constraining the CO-to-H$_2$ conversion factor and CO excitation conditions of galaxies through observations and theory will be necessary to more robustly test the success of galaxy formation models., Re-submitted to ApJ after addressing the first round of comments by the referee, other comments welcome