We present colour-colour diagrams of detected sources in the Herschel-ATLAS science demonstration field from 100 to 500 μm using both PACS and SPIRE. We fit isothermal modified black bodies to the spectral energy distribution (SED) to extract the dust temperature of sources with counterparts in Galaxy And Mass Assembly (GAMA) or SDSS surveys with either a spectroscopic or a photometric redshift. For a subsample of 330 sources detected in at least three FIR bands with a significance greater than 3σ, we find an average dust temperature of (28 ± 8) K. For sources with no known redshift, we populate the colour-colour diagram with a large number of SEDs generated with a broad range of dust temperatures and emissivity parameters, and compare to colours of observed sources to establish the redshift distribution of this sample. For another subsample of 1686 sources with fluxes above 35 mJy at 350 μm and detected at 250 and 500 μm with a significance greater than 3σ, we find an average redshift of 2.2 ± 0.6.

Aims. The Herschel-ATLAS survey (H-ATLAS) will be the largest area survey to be undertaken by the Herschel Space Observatory. It will cover 550 sq. deg. of extragalactic sky at wavelengths of 100, 160, 250, 350 and 500µm when completed, reaching flux limits (5σ) from 32 to 145mJy. We here present galaxy number counts obtained for SPIRE observations of the first ∼14 sq. deg. observed at 250, 350 and 500µm. Methods. Number counts are a fundamental tool in constraining models of galaxy evolution. We use source catalogs extracted from the H-ATLAS maps as the basis for such an analysis. Correction factors for completeness and flux boosting are derived by applying our extraction method to model catalogs and then applied to the raw observational counts. Results. We find a steep rise in the number counts at flux levels of 100-200mJy in all three SPIRE bands, consistent with results from BLAST. The counts are compared to a range of galaxy evolution models. None of the current models is an ideal fit to the data but all ascribe the steep rise to a population of luminous, rapidly evolving dusty galaxies at moderate to high redshift.

We present a self-consistent, semi-analytical cold dark matter ( CDM) model of star formation and reionization. For the cosmological parameters favoured by the Wilkinson Microwave Anisotropy Probe (WMAP) data, our models consistently reproduce the electron scattering optical depth to reionization, redshift of reionization and the observed luminosity functions (LF) and hence the star formation rate (SFR) density at 3 z 6 for a reasonable range of model parameters. While simple photoionization feedback produces the correct shape of LF at z = 6, for z = 3 we need additional feedback that suppresses star formation activities in haloes with 10 10 (M/M ) 10 11 . Models with prolonged continuous star formation activities are preferred over those with short bursts as they are consistent with the existence of a Balmer break in considerable fraction of observed galaxies even at z ∼ 6. The halo number density evolution from the standard CDM structure formation model that fits LF up to z = 6 is consistent with the upper limits on z 7 LF and source counts at 8 z 12 obtained from the Hubble Ultra Deep Field (HUDF) observations without requiring any dramatic change in the nature of star formation. However, to reproduce the observed LF at 6 z 10, obtained from the near-IR observations around strong lensing clusters, we need a strong evolution in the initial mass function, reddening correction and the mode of star formation at z 8. We show that low-mass molecular cooled haloes, which may be important for reionizing the universe, are not detectable in the present deep field observations even if a considerable fraction of its baryonic mass goes through a star burst phase. However, their presence and contribution to reionization can be inferred indirectly from the redshift evolution of the LF in the redshift range 6 z 12. In our model calculations, the contribution of low-mass haloes to global SFR density prior to reionization reveals itself in the form of second peak at z 6. However, this peak will not be visible in the observed SFR density as a function of z as most of these galaxies have luminosity below the detection threshold of various ongoing deep field surveys. Accurately measuring the LF at high redshifts can be used to understand the nature of star formation in the dark ages and probe the history of reionization.

We study winds in high-redshift galaxies driven by a relativistic cosmic ray (proton) component in addition to the hot thermal gas component. Cosmic rays (CRs) are likely to be efficiently generated in supernova shocks inside galaxies. We obtain solutions of such CR-driven free winds in a gravitational potential of the Navarro-Frenk-White form, relevant to galaxies. CRs naturally provide the extra energy and/or momentum input to the system, needed for a transonic wind solution in a gas with adiabatic index γ = 5/3. We show that CRs can effectively drive winds even when the thermal energy of the gas is lost due to radiative cooling. These wind solutions predict an asymptotic wind speed closely related to the circular velocity of the galaxy. Furthermore, the mass outflow rate per unit star formation rate (η w ) is predicted to be ∼0.2-0.5 for massive galaxies, with masses M ∼ 10 11 -10 12 M . We show η w to be inversely proportional to the square of the circular velocity. Magnetic fields at the μG levels are also required in these galaxies to have a significant mass loss. A large η w for small mass galaxies implies that CR-driven outflows could provide a strong negative feedback to the star formation in dwarf galaxies. Further, our results will also have important implications to the metal enrichment of the intergalactic medium. These conclusions are applicable to the class of free wind models where the source region is confined to be within the sonic point.

No abstract

We present semi‐analytic models of galactic outflows that are constrained by available observations on high‐redshift star formation and reionization. Galactic outflows are modelled in a manner akin to models of stellar wind‐blown bubbles. Large‐scale outflows can generically escape from low‐mass haloes (M≲ 109 M⊙) for a wide range of model parameters while this is not the case in high‐mass haloes (M≳ 1011 M⊙). The flow generically accelerates within the halo virial radius, then starts to decelerate, and traverses well into the intergalactic medium (IGM), before freezing to the Hubble flow. The acceleration phase can result in shell fragmentation due to the Rayleigh–Taylor instability, although the final outflow radius is not significantly altered. The gas‐phase metallicities of the outflow and within the galaxy are computed assuming uniform instantaneous mixing. Ionization states of different metal species are calculated and used to examine the detectability of metal lines from the outflows. The global influence of galactic outflows is also investigated using porosity‐weighted averages and probability density functions of various physical quantities. Models with only atomic cooled haloes significantly fill the IGM at z∼ 3 with metals (with −2.5 ≳[Z/Z⊙]≳−3.7), the actual extent depending on the efficiency of winds, the initial mass function and the fractional mass that goes through star formation. The reionization history has a significant effect on the volume filling factor, due to radiative feedback. In these models, a large fraction of outflows at z∼ 3 are supersonic, hot (T≥ 105 K) and have low density, making metal lines difficult to detect. They may also result in significant perturbations in the IGM gas on scales probed by the Lyman α forest. On the contrary, models including molecular cooled haloes with a normal mode of star formation can potentially volume fill the universe at z≥ 8 without drastic dynamic effects on the IGM, thereby setting up a possible metallicity floor (−4.0 ≤[Z/Z⊙]≤−3.6). The fluctuations of order unity at z∼ 8 that become the mildly non‐linear fluctuations traced by Lyman α forest at z < 4 will then have this metallicity. Interestingly, molecular cooled haloes with a ‘top‐heavy’ mode of star formation are not very successful in establishing the metallicity floor because of the additional radiative feedback that they induce.

We present a derivation of the star formation rate per comoving volume of quasar host galaxies, derived from stacking analyses of far-infrared to mm-wave photometry of quasars with redshifts 0 < z < 6 and absolute I-band magnitudes −22 > I AB > −32 We use the science demonstration observations of the first ∼ 16 deg 2 from the Herschel Astrophysical Terahertz Large Area Survey (H-ATLAS) in which there are 240 quasars from the Sloan Digital Sky Survey (SDSS) and a further 171 from the 2dF-SDSS LRG and QSO (2SLAQ) survey. We supplement this data with a compilation of data from IRAS, ISO, Spitzer, SCUBA and MAMBO. H-ATLAS alone statistically detects the quasars in its survey area at > 5σ at 250, 350 and 500 μm. From the compilation as a whole we find striking evidence of downsizing in quasar host galaxy formation: low-luminosity quasars with absolute magnitudes in the range −22 > I AB > −24 have a comoving star formation rate (derived from 100 μm rest-frame luminosities) peaking between redshifts of 1 and 2, while high-luminosity quasars with I AB < −26 have a maximum contribution to the star formation density at z ∼ 3. The volume-averaged star formation rate of −22 > I AB > −24 quasars evolves as (1 + z) 2.3±0.7 at z < 2, but the evolution at higher luminosities is much faster reaching (1 + z) 10±1 at −26 > I AB > −28. We tentatively interpret this as a combination of a declining major merger rate with time and gas consumption reducing fuel for both black hole accretion and star formation.

We present measurements of the angular correlation function of galaxies selected from the first field of the H-ATLAS survey. Careful removal of the background from galactic cirrus is essential, and currently dominates the uncertainty in our measurements. For our 250 μm-selected sample we detect no significant clustering, consistent with the expectation that the 250 μm-selected sources are mostly normal galaxies at z < ∼ 1. For our 350 μm and 500 μm-selected samples we detect relatively strong clustering with correlation amplitudes A of 0.2 and 1.2 at 1 , but with relatively large uncertainties. For samples which preferentially select high redshift galaxies at z ∼ 2−3 we detect significant strong clustering, leading to an estimate of r 0 ∼ 7−11 h −1 Mpc. The slope of our clustering measurements is very steep, δ ∼ 2. The measurements are consistent with the idea that sub-mm sources consist of a low redshift population of normal galaxies and a high redshift population of highly clustered star-bursting galaxies.