Additional information: Use policyThe full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-prot purposes provided that:• a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders.Please consult the full DRO policy for further details. ABSTRACTWe present a new version of the GALFORM semi-analytical model of galaxy formation. This brings together several previous developments of GALFORM into a single unified model, including a different initial mass function (IMF) in quiescent star formation and in starbursts, feedback from active galactic nuclei supressing gas cooling in massive halos, and a new empirical star formation law in galaxy disks based on their molecular gas content. In addition, we have updated the cosmology, introduced a more accurate treatment of dynamical friction acting on satellite galaxies, and updated the stellar population model. The new model is able to simultaneously explain both the observed evolution of the K-band luminosity function and stellar mass function, and the number counts and redshift distribution of sub-mm galaxies selected at 850µm. This was not previously achieved by a single physical model within the ΛCDM framework, but requires having an IMF in starbursts that is somewhat top-heavy. The new model is tested against a wide variety of observational data covering wavelengths from the far-UV to sub-mm, and redshifts from z = 0 to z = 6, and is found to be generally successful. These observations include the optical and near-IR luminosity functions, HI mass function, fraction of early type galaxies, Tully-Fisher, metallicity-luminosity and size-luminosity relations at z = 0, as well as far-IR number counts, and far-UV luminosity functions at z ∼ 3 − 6. Discrepancies are however found in galaxy sizes and metallicities at low luminosities, and in the abundance of low mass galaxies at high-z, suggesting the need for a more sophisticated model of supernova feedback.

We aim to measure the average dust and molecular gas content of massive star-forming galaxies (>3 × 10 10 M ) up to z = 4 in the COSMOS field to determine if the intense star formation observed at high redshift is induced by major mergers or is caused by large gas reservoirs. Firstly, we measured the evolution of the average spectral energy distributions as a function of redshift using a stacking analysis of Spitzer, Herschel, LABOCA, and AzTEC data for two samples of galaxies: normal star-forming objects and strong starbursts, as defined by their distance to the main sequence. We found that the mean intensity of the radiation field U heating the dust (strongly correlated with dust temperature) increases with increasing redshift up to z = 4 in main-sequence galaxies. We can reproduce this evolution with simple models that account for the decrease in the gas metallicity with redshift. No evolution of U with redshift is found in strong starbursts. We then deduced the evolution of the molecular gas fraction (defined here as M mol /(M mol + M )) with redshift and found a similar, steeply increasing trend for both samples. At z ∼ 4, this fraction reaches ∼60%. The average position of the main-sequence galaxies is on the locus of the local, normal star-forming disks in the integrated Schmidt-Kennicutt diagram (star formation rate versus mass of molecular gas), suggesting that the bulk of the star formation up to z = 4 is dominated by secular processes.

We exploit ALMA 870-µm (345 GHz) observations of sub-millimetre sources in the Extended Chandra Deep Field South to investigate the far-infrared properties of highredshift sub-millimetre galaxies (SMGs). Using the precisely located 870 µm ALMA positions of 99 SMGs, together with 24µm and radio imaging of this field, we deblend the Herschel / SPIRE imaging of this region to extract their far-infrared fluxes and colours. The median redshifts for ALMA LESS (ALESS) SMGs which are detected in at least two SPIRE bands increases as expected with wavelength of the peak in their SEDs, with z = 2.3 ± 0.2, 2.5 ± 0.3 and 3.5 ± 0.5 for the 250, 350 and 500-µm peakers respectively. We find that 34 ALESS SMGs do not have a >3 σ counterpart at 250, 350 or 500µm. These galaxies have a median photometric redshift derived from the rest-frame UV-mid-infrared SEDs of z = 3.3 ± 0.5, which is higher than the full ALESS SMG sample; z = 2.5 ± 0.2. Using the photometric redshifts together with the 250-870µm photometry, we estimate the far-infrared luminosities and characteristic dust temperature of each SMG. The median infrared luminosity and characteristic dust temperature of the S 870µm > 2 mJy SMGs is L IR = (3.0 ± 0.3) × 10 12 L ⊙ (star formation rate of SFR = 300 ± 30 M ⊙ yr −1 ) and T d = 32 ± 1 K (λ peak = 93 ± 4 µm). At a fixed luminosity, the characteristic dust temperature of these high-redshift SMGs is ∆T d = 3-5 K lower than comparably luminous galaxies at z = 0, reflecting the more extended star formation occurring in these systems. By extrapolating the 870µm number counts to S 870 = 1 mJy, we show that the contribution of S 870µm ≥ 1 mJy SMGs to the cosmic star formation budget is 20% of the total over the redshift range z ∼ 1-4. We derive a median dust mass for these galaxies of M d = (3.6 ± 0.3) × 10 8 M ⊙ and by adopting an appropriate gas-to-dust ratio, we estimate that the typical molecular mass of the ALESS SMGs in our sample is M H2 = (4.2 ± 0.4) × 10 10 M ⊙ . Together with the average stellar masses of SMGs, M ⋆ = (8 ± 1) × 10 10 M ⊙ , this suggests an average molecular gas fraction of ∼ 40%. Finally, we use our estimates of the H 2 masses to show that SMGs with S 870µm > 1 mJy (L IR > ∼ 10 12 L ⊙ ) contain ∼ 10% of the z ∼ 2 volume-averaged H 2 mass density at this epoch.

We investigate the abundance of galactic molecular hydrogen (H 2 ) in the "Evolution and Assembly of GaLaxies and their Environments" (EAGLE) cosmological hydrodynamic simulations. We assign H 2 masses to gas particles in the simulations in post-processing using two different prescriptions that depend on the local dust-to-gas ratio and the interstellar radiation field. Both result in H 2 galaxy mass functions that agree well with observations in the local and high redshift Universe. The simulations reproduce the observed scaling relations between the mass of H 2 and the stellar mass, star formation rate and stellar surface density. Towards high redshifts, galaxies in the simulations display larger H 2 mass fractions and lower H 2 depletion timescales, also in good agreement with observations. The comoving mass density of H 2 in units of the critical density, Ω H2 , peaks at z ≈ 1.2 − 1.5, later than the predicted peak of the cosmic star formation rate activity, at z ≈ 2. This difference stems from the decrease in gas metallicity and increase in interstellar radiation field with redshift, both of which hamper H 2 formation. We find that the cosmic H 2 budget is dominated by galaxies with M H2 > 10 9 M ⊙ , star formation rates > 10 M ⊙ yr −1 and stellar masses M stellar > 10 10 M ⊙ , which are readily observable in the optical and near-IR. The match between the H 2 properties of galaxies that emerge in the simulations and observations is remarkable, particularly since H 2 observations were not used to adjust parameters in EAGLE. Leroy et al. 2008;Schruba et al. 2011). At a global level (i.e. integrated over the galaxy), the H2 mass correlates very well with the total SFR (Saintonge et al. 2011;Boselli et al. 2014). These observations place new constraints on galaxy formation models.Molecular hydrogen is very difficult to observe in the ISM of galaxies because it lacks a dipole moment, making its emission extremely weak at the typical temperature of the cold ISM. A widely-used tracer of H2 is the carbon monoxide (hereafter 'CO') molecule, which is the second most abundant molecule after H2, and is easily excited (see Carilli & Walter 2013 for a recent review on how CO has been used to trace H2 at z 6). Direct CO detections are available for relatively large samples in the local Universe (z < 0.1), from which it has been possible to derive the z ≈ 0 CO(1 − 0) luminosity function (Keres et al. 2003), where (1 − 0) is the lowest energy rotational transition. From this luminosity function, and adopting a Milky-Way like CO(1 − 0)-H2 conversion factor (Bolatto et al. 2013), it has been possible c 2012 RAS

We report the source size distribution, as measured by ALMA millimetric continuum imaging, of a sample of 13 AzTEC-selected submillimeter galaxies (SMGs) at z 3 phot ~-6. Their infrared luminosities and star formation rates (SFRs) are L IR ~2-6 10 12 ´L and ∼200-600 M  yr −1 , respectively. The sizes of these SMGs range from 0″. 10 to 0″. 38, with a median of 0″. 20 0. 05to that seen in local merger-driven (U)LIRGs rather than in extended disk galaxies at low and high redshifts. The discovery of compact starbursts in z 3  SMGs strongly supports a massive galaxy formation scenario wherein z 3 ~-6 SMGs evolve into the compact stellar components of z 2 ~cQGs. These cQGs are then thought to evolve into the most massive ellipticals in the local universe, mostly via dry mergers. Our results thus suggest that z 3  SMGs are the likely progenitors of massive local ellipticals, via cQGs, meaning that we can now trace the evolutionary path of the most massive galaxies over a period encompassing ∼90% of the age of the universe.