The Feedback In Realistic Environments (FIRE) project explores feedback in cosmological galaxy formation simulations. Previous FIRE simulations used an identical source code ("FIRE-1") for consistency. Motivated by the development of more accurate numerics -including hydrodynamic solvers, gravitational softening, and supernova coupling algorithms -and exploration of new physics (e.g. magnetic fields), we introduce "FIRE-2", an updated numerical implementation of FIRE physics for the GIZMO code. We run a suite of simulations and compare against FIRE-1: overall, FIRE-2 improvements do not qualitatively change galaxy-scale properties. We pursue an extensive study of numerics versus physics. Details of the star-formation algorithm, cooling physics, and chemistry have weak effects, provided that we include metal-line cooling and star formation occurs at higher-than-mean densities. We present new resolution criteria for high-resolution galaxy simulations. Most galaxy-scale properties are robust to numerics we test, provided: (1) Toomre masses are resolved; (2) feedback coupling ensures conservation, and (3) individual supernovae are time-resolved. Stellar masses and profiles are most robust to resolution, followed by metal abundances and morphologies, followed by properties of winds and circum-galactic media (CGM). Central (∼kpc) mass concentrations in massive (> L * ) galaxies are sensitive to numerics (via trapping/recycling of winds in hot halos). Multiple feedback mechanisms play key roles: supernovae regulate stellar masses/winds; stellar mass-loss fuels late star formation; radiative feedback suppresses accretion onto dwarfs and instantaneous star formation in disks. We provide all initial conditions and numerical algorithms used.

zCOSMOS is a large redshift survey that is being undertaken in the COSMOS field using 600 hours of observation with the VIMOS spectrograph on the 8-m VLT. The survey is designed to characterise the environments of COSMOS galaxies from the 100 kpc scales of galaxy groups up to the 100 Mpc scale of the cosmic web and to produce diagnostic information on galaxies and active galactic nuclei. The zCOSMOS survey consists of two parts: (a) zCOSMOS-bright, a magnitude-limited I-band I AB < 22.5 sample of about 20,000 galaxies with 0.1 < z < 1.2 covering the whole 1.7 deg 2 COSMOS ACS field, for which the survey parameters at z ~ 0.7 are designed to be directly comparable to those of the 2dFGRS at z ~ 0.1; and (b) zCOSMOS-deep, a survey of approximately 10,000 galaxies selected through colourselection criteria to have 1.4 < z < 3.0, within the central 1 deg 2 . This paper describes the survey design and the construction of the target catalogues, and briefly outlines the observational program and the data pipeline. In the first observing season, spectra of 1303 zCOSMOS-bright targets and of 977 zCOSMOS-deep targets have been obtained. These are briefly analysed to demonstrate the characteristics that may be expected from zCOSMOS, and particularly zCOSMOS-bright, when it is finally completed between 2008-2009. The power of combining spectroscopic and photometric redshifts is demonstrated, especially in correctly identifying the emission line in single-line spectra and in determining which of the less reliable spectroscopic redshifts are correct and which are incorrect. These techniques bring the overall success rate in the zCOSMOS-bright so far to almost 90% and to above 97% in the 0.5 < z < 0.8 redshift range. Our zCOSMOS-deep spectra demonstrate the power of our selection techniques to isolate high redshift galaxies at 1.4 < z < 3.0 and of VIMOS to measure their redshifts using ultraviolet absorption lines.

Context. Photometric redshifts (photo-z's) have become an essential tool in extragalactic astronomy. Many current and upcoming observing programmes require great accuracy of photo-z's to reach their scientific goals. Aims. Here we introduce PHAT, the PHoto-z Accuracy Testing programme, an international initiative to test and compare different methods of photo-z estimation. Methods. Two different test environments are set up, one (PHAT0) based on simulations to test the basic functionality of the different photo-z codes, and another one (PHAT1) based on data from the GOODS survey including 18-band photometry and ∼2000 spectroscopic redshifts.Results. The accuracy of the different methods is expressed and ranked by the global photo-z bias, scatter, and outlier rates. While most methods agree very well on PHAT0 there are differences in the handling of the Lyman-α forest for higher redshifts. Furthermore, different methods produce photo-z scatters that can differ by up to a factor of two even in this idealised case. A larger spread in accuracy is found for PHAT1. Few methods benefit from the addition of mid-IR photometry. The accuracy of the other methods is unaffected or suffers when IRAC data are included. Remaining biases and systematic effects can be explained by shortcomings in the different template sets (especially in the mid-IR) and the use of priors on the one hand and an insufficient training set on the other hand. Some strategies to overcome these problems are identified by comparing the methods in detail. Scatters of 4-8% in Δz/(1 + z) were obtained, consistent with other studies. However, somewhat larger outlier rates (>7.5% with Δz/(1 + z) > 0.15; >4.5% after cleaning) are found for all codes that can only partly be explained by AGN or issues in the photometry or the spec-z catalogue. Some outliers were probably missed in comparisons of photo-z's to other, less complete spectroscopic surveys in the past. There is a general trend that empirical codes produce smaller biases than template-based codes. Conclusions. The systematic, quantitative comparison of different photo-z codes presented here is a snapshot of the current state-ofthe-art of photo-z estimation and sets a standard for the assessment of photo-z accuracy in the future. The rather large outlier rates reported here for PHAT1 on real data should be investigated further since they are most probably also present (and possibly hidden) in many other studies. The test data sets are publicly available and can be used to compare new, upcoming methods to established ones and help in guiding future photo-z method development.

Galaxy formation models are now able to reproduce observed relations such as the relation between galaxies' star formation rates (SFRs) and stellar masses (M * ) and the stellar masshalo mass relation. We demonstrate that comparisons of the short-timescale variability in galaxy SFRs with observational data provide an additional useful constraint on the physics of galaxy formation feedback. We apply SFR indicators with different sensitivity timescales to galaxies from the Feedback in Realistic Environments (FIRE) simulations. We find that the SFR-M * relation has a significantly greater scatter when the Hα-derived SFR is considered compared with when the far-ultraviolet (FUV)-based SFR is used. This difference is a direct consequence of bursty star formation because the FIRE galaxies exhibit order-of-magnitude SFR variations over timescales of a few Myr. We show that the difference in the scatter between the simulated Hα-and FUV-derived SFR-M * relations at z = 2 is consistent with observational constraints. We also find that the Hα/FUV ratios predicted by the simulations at z = 0 are similar to those observed for local galaxies except for a population of low-mass (M * 10 9.5 M ) simulated galaxies with lower Hα/FUV ratios than observed. We suggest that future cosmological simulations should compare the Hα/FUV ratios of their galaxies with observations to constrain the feedback models employed.

We introduce massive black holes (BHs) in the Feedback In Realistic Environments project and perform high-resolution cosmological hydrodynamic simulations of quasar-mass halos (M halo (z = 2) ≈ 10 12.5 M ) down to z = 1. These simulations model stellar feedback by supernovae, stellar winds, and radiation, and BH growth using a gravitational torque-based prescription tied to resolved properties of galactic nuclei. We do not include BH feedback. We show that early BH growth occurs through short ( 1 Myr) accretion episodes that can reach or even exceed the Eddington rate. In this regime, BH growth is limited by bursty stellar feedback continuously evacuating gas from galactic nuclei, and BHs remain under-massive relative to the local M BH -M bulge relation. BH growth is more efficient at later times, when the nuclear stellar potential retains a significant gas reservoir, star formation becomes less bursty, and galaxies settle into a more ordered state, with BHs rapidly converging onto the scaling relation when the host reaches M bulge ∼ 10 10 M . Our results are not sensitive to the details of the accretion model so long as BH growth is tied to the gas content within ∼100 pc of the BH. Our simulations imply that bursty stellar feedback has strong implications for BH and AGN demographics, especially in the early Universe and for low-mass galaxies.

Submillimetre-bright galaxies at high redshift are the most luminous, heavily star-forming galaxies in the Universe and are characterized by prodigious emission in the far-infrared, with a flux of at least five millijanskys at a wavelength of 850 micrometres. They reside in haloes with masses about 10(13) times that of the Sun, have low gas fractions compared to main-sequence disks at a comparable redshift, trace complex environments and are not easily observable at optical wavelengths. Their physical origin remains unclear. Simulations have been able to form galaxies with the requisite luminosities, but have otherwise been unable to simultaneously match the stellar masses, star formation rates, gas fractions and environments. Here we report a cosmological hydrodynamic galaxy formation simulation that is able to form a submillimetre galaxy that simultaneously satisfies the broad range of observed physical constraints. We find that groups of galaxies residing in massive dark matter haloes have increasing rates of star formation that peak at collective rates of about 500-1,000 solar masses per year at redshifts of two to three, by which time the interstellar medium is sufficiently enriched with metals that the region may be observed as a submillimetre-selected system. The intense star formation rates are fuelled in part by the infall of a reservoir gas supply enabled by stellar feedback at earlier times, not through major mergers. With a lifetime of nearly a billion years, our simulations show that the submillimetre-bright phase of high-redshift galaxies is prolonged and associated with significant mass buildup in early-Universe proto-clusters, and that many submillimetre-bright galaxies are composed of numerous unresolved components (for which there is some observational evidence).

We present the Zurich Extragalactic Bayesian Redshift Analyzer (zebra). The current version of zebra combines and extends several of the classical approaches to produce accurate photometric redshifts down to faint magnitudes. In particular, zebra uses the template‐fitting approach to produce Maximum Likelihood and Bayesian redshift estimates based on the following points. An automatic iterative technique to correct the original set of galaxy templates to best represent the Spectral Energy Distributions (SEDs) of real galaxies at different redshifts. A training set of spectroscopic redshifts for a small fraction of the photometric sample to improve the robustness of the photometric redshift estimates. An iterative technique for Bayesian redshift estimates, which extracts the full two‐dimensional redshift and template probability function for each galaxy. We demonstrate the performance of zebra by applying it to a sample of 866 IAB≤ 22.5 COSMOS galaxies with available u*, B, V, g′, r′, i′, z′ and Ks photometry and zCOSMOS spectroscopic redshifts in the range 0 < z < 1.3. Adopting a 5σ clipping that excludes ≤10 galaxies, both the Maximum Likelihood and Bayesian zebra estimates for this sample have an accuracy σΔz/(1+z) smaller than 0.03. Similar accuracies are recovered using mock galaxies. zebra is made available at http://www.exp-astro.phys.ethz.ch/ZEBRA.

Dust temperature is an important property of the interstellar medium (ISM) of galaxies. It is required when converting (sub)millimetre broad-band flux to total infrared luminosity (LIR), and hence star formation rate, in high-redshift galaxies. However, different definitions of dust temperatures have been used in the literature, leading to different physical interpretations of how ISM conditions change with, e.g. redshift and star formation rate. In this paper, we analyse the dust temperatures of massive ($M_{\rm star} \gt 10^{10}\, \mathrm{M}_{\odot }$) $z$ = 2–6 galaxies with the help of high-resolution cosmological simulations from the Feedback in Realistic Environments (fire) project. At $z$ ∼ 2, our simulations successfully predict dust temperatures in good agreement with observations. We find that dust temperatures based on the peak emission wavelength increase with redshift, in line with the higher star formation activity at higher redshift, and are strongly correlated with the specific star formation rate. In contrast, the mass-weighted dust temperature, which is required to accurately estimate the total dust mass, does not strongly evolve with redshift over $z$ = 2–6 at fixed IR luminosity but is tightly correlated with LIR at fixed $z$. We also analyse an ‘equivalent’ dust temperature for converting (sub)millimetre flux density to total IR luminosity, and provide a fitting formula as a function of redshift and dust-to-metal ratio. We find that galaxies of higher equivalent (or higher peak) dust temperature (‘warmer dust’) do not necessarily have higher mass-weighted temperatures. A ‘two-phase’ picture for interstellar dust can explain the different scaling relations of the various dust temperatures.