Old galaxies in the young Universe

More than half of all stars in the local Universe are found in massive spheroidal galaxies, which are characterized by old stellar populations with little or no current star formation. In present models, such galaxies appear rather late in the history of the Universe as the culmination of a hierarchical merging process, in which larger galaxies are assembled through mergers of smaller precursor galaxies. But observations have not yet established how, or even when, the massive spheroidals formed, nor if their seemingly sudden appearance when the Universe was about half its present age (at redshift z approximately 1) results from a real evolutionary effect (such as a peak of mergers) or from the observational difficulty of identifying them at earlier epochs. Here we report the spectroscopic and morphological identification of four old, fully assembled, massive (10(11) solar masses) spheroidal galaxies at l.6 < z < 1.9, the most distant such objects currently known. The existence of such systems when the Universe was only about one-quarter of its present age shows that the build-up of massive early-type galaxies was much faster in the early Universe than has been expected from theoretical simulations.     

A substantial population of low-mass stars in luminous elliptical galaxies

The stellar initial mass function (IMF) describes the mass distribution of stars at the time of their formation and is of fundamental importance for many areas of astrophysics. The IMF is reasonably well constrained in the disk of the Milky Way but we have very little direct information on the form of the IMF in other galaxies and at earlier cosmic epochs. Here we report observations of the Na?(I) doublet and the Wing-Ford molecular FeH band in the spectra of elliptical galaxies. These lines are strong in stars with masses less than 0.3M(⊙) (where M(⊙) is the mass of the Sun) and are weak or absent in all other types of stars. We unambiguously detect both signatures, consistent with previous studies that were based on data of lower signal-to-noise ratio. The direct detection of the light of low-mass stars implies that they are very abundant in elliptical galaxies, making up over 80% of the total number of stars and contributing more than 60% of the total stellar mass. We infer that the IMF in massive star-forming galaxies in the early Universe produced many more low-mass stars than the IMF in the Milky Way disk, and was probably slightly steeper than the Salpeter form in the mass range 0.1M(⊙) to 1M(⊙).     

The star-formation history of the Universe from the stellar populations of nearby galaxies

The determination of the star-formation history of the Universe is a key goal of modern cosmology, as it is crucial to our understanding of how galactic structures form and evolve. Observations of young stars in distant galaxies at different times in the past have indicated that the stellar birthrate peaked some eight billion years ago before declining by a factor of around ten to its present value. Here we report an analysis of the 'fossil record' of the current stellar populations of 96,545 nearby galaxies, from which we obtained a complete star-formation history. Our results broadly support those derived from high-redshift galaxies. We find, however, that the peak of star formation was more recent--around five billion years ago. We also show that the bigger the stellar mass of the galaxy, the earlier the stars were formed, which indicates that high- and low-mass galaxies have very different histories.     

The suppression of star formation by powerful active galactic nuclei

The old, red stars that constitute the bulges of galaxies, and the massive black holes at their centres, are the relics of a period in cosmic history when galaxies formed stars at remarkable rates and active galactic nuclei (AGN) shone brightly as a result of accretion onto black holes. It is widely suspected, but unproved, that the tight correlation between the mass of the black hole and the mass of the stellar bulge results from the AGN quenching the surrounding star formation as it approaches its peak luminosity. X-rays trace emission from AGN unambiguously, whereas powerful star-forming galaxies are usually dust-obscured and are brightest at infrared and submillimetre wavelengths. Here we report submillimetre and X-ray observations that show that rapid star formation was common in the host galaxies of AGN when the Universe was 2-6 billion years old, but that the most vigorous star formation is not observed around black holes above an X-ray luminosity of 10(44) ergs per second. This suppression of star formation in the host galaxy of a powerful AGN is a key prediction of models in which the AGN drives an outflow, expelling the interstellar medium of its host and transforming the galaxy's properties in a brief period of cosmic time.     

A large population of galaxies 9 to 12 billion years back in the history of the Universe

To understand the evolution of galaxies, we need to know as accurately as possible how many galaxies were present in the Universe at different epochs. Galaxies in the young Universe have hitherto mainly been identified using their expected optical colours, but this leaves open the possibility that a significant population remains undetected because their colours are the result of a complex mix of stars, gas, dust or active galactic nuclei. Here we report the results of a flux-limited I-band survey of galaxies at look-back times of 9 to 12 billion years. We find 970 galaxies with spectroscopic redshifts between 1.4 and 5. This population is 1.6 to 6.2 times larger than previous estimates, with the difference increasing towards brighter magnitudes. Strong ultraviolet continua (in the rest frame of the galaxies) indicate vigorous star formation rates of more than 10-100 solar masses per year. As a consequence, the cosmic star formation rate representing the volume-averaged production of stars is higher than previously measured at redshifts of 3 to 4.     

A correlation between the cosmic microwave background and large-scale structure in the Universe

Observations of distant supernovae and the fluctuations in the cosmic microwave background (CMB) indicate that the expansion of the Universe may be accelerating under the action of a 'cosmological constant' or some other form of 'dark energy'. This dark energy now appears to dominate the Universe and not only alters its expansion rate, but also affects the evolution of fluctuations in the density of matter, slowing down the gravitational collapse of material (into, for example, clusters of galaxies) in recent times. Additional fluctuations in the temperature of CMB photons are induced as they pass through large-scale structures and these fluctuations are necessarily correlated with the distribution of relatively nearby matter. Here we report the detection of correlations between recent CMB data and two probes of large-scale structure: the X-ray background and the distribution of radio galaxies. These correlations are consistent with those predicted by dark energy, indicating that we are seeing the imprint of dark energy on the growth of structure in the Universe.     

The formation of cluster elliptical galaxies as revealed by extensive star formation

The most massive galaxies in the present-day Universe are found to lie in the centres of rich clusters. They have old, coeval stellar populations suggesting that the bulk of their stars must have formed at early epochs in spectacular starbursts, which should be luminous phenomena when observed at submillimetre wavelengths. The most popular model of galaxy formation predicts that these galaxies form in proto-clusters at high-density peaks in the early Universe. Such peaks are indicated by massive high-redshift radio galaxies. Here we report deep submillimetre mapping of seven high-redshift radio galaxies and their environments. These data confirm not only the presence of spatially extended regions of massive star-formation activity in the radio galaxies themselves, but also in companion objects previously undetected at any wavelength. The prevalence, orientation, and inferred masses of these submillimetre companion galaxies suggest that we are witnessing the synchronous formation of the most luminous elliptical galaxies found today at the centres of rich clusters of galaxies.     

Spectroscopic confirmation of a galaxy at redshift z = 8.6

Galaxies had their most significant impact on the Universe when they assembled their first generations of stars. Energetic photons emitted by young, massive stars in primeval galaxies ionized the intergalactic medium surrounding their host galaxies, cleared sightlines along which the light of the young galaxies could escape, and fundamentally altered the physical state of the intergalactic gas in the Universe continuously until the present day. Observations of the cosmic microwave background, and of galaxies and quasars at the highest redshifts, suggest that the Universe was reionized through a complex process that was completed about a billion years after the Big Bang, by redshift z?≈?6. Detecting ionizing Lyman-α photons from increasingly distant galaxies places important constraints on the timing, location and nature of the sources responsible for reionization. Here we report the detection of Lyα photons emitted less than 600?million years after the Big Bang. UDFy-38135539 (ref. 5) is at a redshift of z = 8.5549?±?0.0002, which is greater than those of the previously known most distant objects, at z = 8.2 (refs 6 and 7) and z = 6.96 (ref. 8). We find that this single source is unlikely to provide enough photons to ionize the volume necessary for the emission line to escape, requiring a significant contribution from other, probably fainter galaxies nearby.     

The rapid formation of a large rotating disk galaxy three billion years after the Big Bang

Observations and theoretical simulations have established a framework for galaxy formation and evolution in the young Universe. Galaxies formed as baryonic gas cooled at the centres of collapsing dark-matter haloes; mergers of haloes and galaxies then led to the hierarchical build-up of galaxy mass. It remains unclear, however, over what timescales galaxies were assembled and when and how bulges and disks--the primary components of present-day galaxies--were formed. It is also puzzling that the most massive galaxies were more abundant and were forming stars more rapidly at early epochs than expected from models. Here we report high-angular-resolution observations of a representative luminous star-forming galaxy when the Universe was only 20% of its current age. A large and massive rotating protodisk is channelling gas towards a growing central stellar bulge hosting an accreting massive black hole. The high surface densities of gas, the high rate of star formation and the moderately young stellar ages suggest rapid assembly, fragmentation and conversion to stars of an initially very gas-rich protodisk, with no obvious evidence for a major merger.     

A probable stellar solution to the cosmological lithium discrepancy

The measurement of the cosmic microwave background has strongly constrained the cosmological parameters of the Universe. When the measured density of baryons (ordinary matter) is combined with standard Big Bang nucleosynthesis calculations, the amounts of hydrogen, helium and lithium produced shortly after the Big Bang can be predicted with unprecedented precision. The predicted primordial lithium abundance is a factor of two to three higher than the value measured in the atmospheres of old stars. With estimated errors of 10 to 25%, this cosmological lithium discrepancy seriously challenges our understanding of stellar physics, Big Bang nucleosynthesis or both. Certain modifications to nucleosynthesis have been proposed, but found experimentally not to be viable. Diffusion theory, however, predicts atmospheric abundances of stars to vary with time, which offers a possible explanation of the discrepancy. Here we report spectroscopic observations of stars in the metal-poor globular cluster NGC 6397 that reveal trends of atmospheric abundance with evolutionary stage for various elements. These element-specific trends are reproduced by stellar-evolution models with diffusion and turbulent mixing. We thus conclude that diffusion is predominantly responsible for the low apparent stellar lithium abundance in the atmospheres of old stars by transporting the lithium deep into the star.     

A measurement of the cosmological mass density from clustering in the 2dF Galaxy Redshift Survey

The large-scale structure in the distribution of galaxies is thought to arise from the gravitational instability of small fluctuations in the initial density field of the Universe. A key test of this hypothesis is that forming superclusters of galaxies should generate a systematic infall of other galaxies. This would be evident in the pattern of recessional velocities, causing an anisotropy in the inferred spatial clustering of galaxies. Here we report a precise measurement of this clustering, using the redshifts of more than 141,000 galaxies from the two-degree-field (2dF) galaxy redshift survey. We determine the parameter beta = Omega0.6/b = 0.43 +/- 0.07, where Omega is the total mass-density parameter of the Universe and b is a measure of the 'bias' of the luminous galaxies in the survey. (Bias is the difference between the clustering of visible galaxies and of the total mass, most of which is dark.) Combined with the anisotropy of the cosmic microwave background, our results favour a low-density Universe with Omega approximately 0.3.     

A galaxy at a redshift z = 6.96

When galaxy formation started in the history of the Universe remains unclear. Studies of the cosmic microwave background indicate that the Universe, after initial cooling (following the Big Bang), was reheated and reionized by hot stars in newborn galaxies at a redshift in the range 6 < z < 14 (ref. 1). Though several candidate galaxies at redshift z > 7 have been identified photometrically, galaxies with spectroscopically confirmed redshifts have been confined to z < 6.6 (refs 4-8). Here we report a spectroscopic redshift of z = 6.96 (corresponding to just 750 Myr after the Big Bang) for a galaxy whose spectrum clearly shows Lyman-alpha emission at 9,682 A, indicating active star formation at a rate of approximately 10M(o) yr(-1), where M(o) is the mass of the Sun. This demonstrates that galaxy formation was under way when the Universe was only approximately 6 per cent of its present age. The number density of galaxies at z approximately 7 seems to be only 18-36 per cent of the density at z = 6.6.     

Characterizing the nonlinear growth of large-scale structure in the Universe

The local Universe displays a rich hierarchical pattern of galaxy clusters and superclusters. The early Universe, however, was almost smooth, with only slight 'ripples' as seen in the cosmic microwave background radiation. Models of the evolution of cosmic structure link these observations through the effect of gravity, because the small initially overdense fluctuations are predicted to attract additional mass as the Universe expands. During the early stages of this expansion, the ripples evolve independently, like linear waves on the surface of deep water. As the structures grow in mass, they interact with each other in nonlinear ways, more like waves breaking in shallow water. We have recently shown how cosmic structure can be characterized by phase correlations associated with these nonlinear interactions, but it was not clear how to use that information to obtain quantitative insights into the growth of structures. Here we report a method of revealing phase information, and show quantitatively how this relates to the formation of filaments, sheets and clusters of galaxies by nonlinear collapse. We develop a statistical method based on information entropy to separate linear from nonlinear effects, and thereby are able to disentangle those aspects of galaxy clustering that arise from initial conditions (the ripples) from the subsequent dynamical evolution.     

A median redshift of 2.4 for galaxies bright at submillimetre wavelengths

A significant fraction of the energy emitted in the early Universe came from very luminous galaxies that are largely hidden at optical wavelengths (because of interstellar dust grains); this energy now forms part of the cosmic background radiation at wavelengths near 1 mm (ref. 1). Some submillimetre (submm) galaxies have been resolved from the background radiation, but they have been difficult to study because of instrumental limitations. This has impeded the determination of their redshifts (z), which is a crucial element in understanding their nature and evolution. Here we report spectroscopic redshifts for ten submm galaxies that were identified using high-resolution radio observations. The median redshift for our sample is 2.4, with a quartile range of 1.9-2.8. This population therefore coexists with the peak activity of quasars, suggesting a close relationship between the growth of massive black holes and luminous dusty galaxies. The space density of submm galaxies at redshifts over 2 is about 1,000 times greater than that of similarly luminous galaxies in the present-day Universe, so they represent an important component of star formation at high redshifts.     

The formation of the first low-mass stars from gas with low carbon and oxygen abundances

The first stars in the Universe are predicted to have been much more massive than the Sun. Gravitational condensation, accompanied by cooling of the primordial gas via molecular hydrogen, yields a minimum fragmentation scale of a few hundred solar masses. Numerical simulations indicate that once a gas clump acquires this mass it undergoes a slow, quasi-hydrostatic contraction without further fragmentation; lower-mass stars cannot form. Here we show that as soon as the primordial gas--left over from the Big Bang--is enriched by elements ejected from supernovae to a carbon or oxygen abundance as small as approximately 0.01-0.1 per cent of that found in the Sun, cooling by singly ionized carbon or neutral oxygen can lead to the formation of low-mass stars by allowing cloud fragmentation to smaller clumps. This mechanism naturally accommodates the recent discovery of solar-mass stars with unusually low iron abundances (10(-5.3) solar) but with relatively high (10(-1.3) solar) carbon abundance. The critical abundances that we derive can be used to identify those metal-poor stars in our Galaxy with elemental patterns imprinted by the first supernovae. We also find that the minimum stellar mass at early epochs is partially regulated by the temperature of the cosmic microwave background.     

Type II supernovae as a significant source of interstellar dust

Large amounts of dust (>10(8)M(o)) have recently been discovered in high-redshift quasars and galaxies corresponding to a time when the Universe was less than one-tenth of its present age. The stellar winds produced by stars in the late stages of their evolution (on the asymptotic giant branch of the Hertzsprung-Russell diagram) are thought to be the main source of dust in galaxies, but they cannot produce that dust on a short enough timescale (&<1 Gyr) to explain the results in the high-redshift galaxies. Supernova explosions of massive stars (type II) are also a potential source, with models predicting 0.2-4M(o) of dust. As massive stars evolve rapidly, on timescales of a few Myr, these supernovae could be responsible for the high-redshift dust. Observations of supernova remnants in the Milky Way, however, have hitherto revealed only 10(-7)-10(-3)M(o) each, which is insufficient to explain the high-redshift data. Here we report the detection of approximately 2-4M(o) of cold dust in the youngest known Galactic supernova remnant, Cassiopeia A. This observation implies that supernovae are at least as important as stellar winds in producing dust in our Galaxy and would have been the dominant source of dust at high redshifts.     

Cosmic gamma-ray background from structure formation in the intergalactic medium

The Universe is filled with a diffuse background of gamma-ray radiation, the origin of which remains one of the unsolved puzzles of cosmology. Less than one-quarter of the gamma-ray flux can be attributed to unresolved discrete sources, such as active galactic nuclei; the remainder appears to constitute a truly diffuse background. Here we show that the shock waves induced by gravity in the gas of the intergalactic medium, during the formation of large-scale structures like filaments and sheets of galaxies, produce a population of highly relativistic electrons. These electrons scatter a small fraction of the cosmic microwave background photons in the local Universe up to gamma-ray energies, thereby providing the gamma-ray background. The predicted diffuse flux agrees with the observed background across more than four orders of magnitude in photon energy, and the model predicts that the gamma-ray background, though generated locally, is isotropic to better than five per cent on angular scales larger than a degree. Moreover, the agreement between the predicted and observed background fluxes implies a mean cosmological density of baryons that is consistent with Big Bang nucleosynthesis.     

The obscuration by dust of most of the growth of supermassive black holes

Supermassive black holes underwent periods of exponential growth during which we see them as quasars in the distant Universe. The summed emission from these quasars generates the cosmic X-ray background, the spectrum of which has been used to argue that most black-hole growth is obscured. There are clear examples of obscured black-hole growth in the form of 'type-2' quasars, but their numbers are fewer than expected from modelling of the X-ray background. Here we report the direct detection of a population of distant type-2 quasars, which is at least comparable in size to the well-known unobscured type-1 population. We selected objects that have mid-infrared and radio emissions characteristic of quasars, but which are faint at near-infrared and optical wavelengths. We conclude that this population is responsible for most of the black-hole growth in the young Universe and that, throughout cosmic history, black-hole growth occurs in the dusty, gas-rich centres of active galaxies.     

Rapid evolution of the most luminous galaxies during the first 900 million years

The first 900 million years (Myr) to redshift z approximately 6 (the first seven per cent of the age of the Universe) remains largely unexplored for the formation of galaxies. Large samples of galaxies have been found at z approximately 6 (refs 1-4) but detections at earlier times are uncertain and unreliable. It is not at all clear how galaxies built up from the first stars when the Universe was about 300 Myr old (z approximately 12-15) to z approximately 6, just 600 Myr later. Here we report the results of a search for galaxies at z approximately 7-8, about 700 Myr after the Big Bang, using the deepest near-infrared and optical images ever taken. Under conservative selection criteria we find only one candidate galaxy at z approximately 7-8, where ten would be expected if there were no evolution in the galaxy population between z approximately 7-8 and z approximately 6. Using less conservative criteria, there are four candidates, where 17 would be expected with no evolution. This demonstrates that very luminous galaxies are quite rare 700 Myr after the Big Bang. The simplest explanation is that the Universe is just too young to have built up many luminous galaxies at z approximately 7-8 by the hierarchical merging of small galaxies.     

A characteristic size of approximately 10 Mpc for the ionized bubbles at the end of cosmic reionization

The first galaxies to appear in the Universe at redshifts z > 20 created ionized bubbles in the intergalactic medium of neutral hydrogen left over from the Big Bang. The ionized bubbles grew with time, surrounding clusters of dwarf galaxies and eventually overlapped quickly throughout the Universe over a narrow redshift interval near z approximately 6. This event signalled the end of the reionization epoch when the Universe was a billion years old. Measuring the size distribution of the bubbles at their final overlap phase is a focus of forthcoming programmes to observe highly redshifted radio emission from atomic hydrogen. Here we show that the combined constraints of cosmic variance and light travel time imply an observed bubble size at the end of the overlap epoch of approximately 10 physical Mpc, and a scatter in the observed redshift of overlap along different lines-of-sight of approximately 0.15. This scatter is consistent with observational constraints from recent spectroscopic data on the farthest known quasars. This implies that future radio experiments should be tuned to a characteristic angular scale of 0.5 degrees and have a minimum frequency bandwidth of approximately 8 MHz for an optimal detection of 21-cm flux fluctuations near the end of reionization.