Newly discovered fast-growing galaxies could solve cosmic riddle – and show ancient cosmic merger
Astronomers have discovered a new kind of galaxy in the early universe, less than a billion years after the Big Bang. These galaxies are forming stars more than a hundred times faster than our own Milky Way. The discovery could explain an earlier finding: a population of suprisingly massive galaxies at a time 1.5 billion years after the Big Bang, which would require such hyper-productive precursors to grow their hundreds of billions of stars. The observations also show what appears to be the earliest image of galaxies merging. The results, by a group of astronomers led by Roberto Decarli of the Max Planck Institute for Astronomy, have been published in the 25 May issue of the journal Nature.
In-depth description: Newly discovered fast-growing galaxies could solve cosmic riddle – and show ancient cosmic merger
A group of astronomers led by Roberto Decarli at the Max Planck Institute has discovered surprisingly productive galaxies in the very early universe. These galaxies, which we see as they were less than a billion years after the Big Bang, produce more than hundred solar masses worth of stars every year – and could be the key to explaining a population of somewhat later unusually massive galaxies that other astronomers had discovered in the early universe, about 1.5 billion years after the Big Bang. Those later massive galaxies posed a particular kind of puzzle: While less than a billion years old themselves, they contain numerous reddish stars almost as old as these galaxies themselves, indicating that they must have been forming stars at a high rate for almost all of their existence.
Understanding cosmic history
On the one hand, the history of the universe as a whole is simpler than the history of Earth's human inhabitants. Cosmological history directly follows simple fundamental laws, namely the laws of physics. On the other hand, this ups the ante for cosmologists: They should be able to explain in terms of physical processes how the universe has reached its present state from a fairly boring, almost homogeneous beginning directly after the Big Bang, 13.8 billion years ago.
There are several key classes of objects whose properties and evolution need explaining. First of all, there is dark matter, which does not interact with light and other forms of electromagnetic radiation at all. Over the past 13.8 billion years, dark matter has clumped together under its own gravity, forming the gigantic filaments of the cosmic web, the backdrop or framework of cosmic history. On smaller scales, dark matter has formed loose, almost spherical associations known as halos. Gas collecting in those halos has formed galaxies: collections of between hundreds of thousands and hundreds of billions of stars, suffused with (mostly hydrogen) gas.
To the best of current astronomical knowledge, every massive galaxy contains a supermassive black hole in its central regions, with masses between a few hundred thousand and a few billion times the mass of the Sun. (The central black hole of our own galaxy has a mass of 4 million solar masses.) When sufficient amounts of matter fall into such a supermassive black hole, it turns into a quasar: directly before falling into the black hole, matter collects in a swirling disk; this "accretion disk" is heated up as more and more infalling matter deposits its energy; the extreme temperature of the disk (think "incandescent light bulb") and additional effects make the quasar into one of the brightest objects in the universe, as bright as all the stars of a large galaxy combined.
In addition to stars, and rare and transient phenomena like quasars, there is intergalactic gas – again, mostly hydrogen, both in the galaxies themselves and filling the void between galaxies, and between the filaments of the cosmic web.
Cosmic history on display
Cosmic history describes the formation and the evolution of these objects, including their interactions. How and when did galaxies form their stars? Is intergalactic gas funneled into galaxies, providing new raw material for star formation? Does quasar activity hinder or encourage star formation? Is star formation the same throughout history, or did galaxies become less productive, or more productive, over time? By now, the field of cosmic historiography can provide at least some answers. Open questions are pursued using modeling, simulations, and observations – including recent massive surveys that enable statistics with samples of hundreds of thousands of objects.
Astronomical distances are so large that it takes the light of distant objects an impressive time to reach us here on Earth. That provides astronomers with a cross section of cosmic history. For instance, we see the Andromeda galaxy as it was 2.5 million years ago, since Andromeda's light has taken 2.5 million years to reach us. Other galaxies, we see as they were billions of years ago.
Thus, while we cannot follow the entire history of any single object, astronomical observations do show us the different stages of cosmic history. Assuming that at least on average, no location within the universe is markedly different from any other – for instance, that we will find the same numbers of galaxies, or quasars, with the same average properties –, we can observe distant objects as they once were, and draw conclusions about our own past.
An unusual population of massive galaxies
Cosmology must take the many observations that represent different epochs of cosmic history and weave them into a consistent physical narrative: Objects that have been found in one particular epoch must have formed in some earlier epoch. One example is the discovery of a substantial population of very massive galaxies, each with hundreds of billions of stars and a total mass of hundreds of billions of solar masses, in an epoch around 1.5 billion years after the Big Bang (z ∼ 4) by Caroline Straatman (then Leiden University, now at MPIA) and collaborators in 2014.
Once this observation has been made, it needs to be explained. For there to be galaxies that rich in stars at a time of 1.5 billion years after the Big Bang, when the universe was a bit more than 10% its present age, the precursors of these galaxies must have formed stars at an enormous rate at earlier epochs.
But do we see evidence for such actively star-forming galaxies in the very early universe?
A serendipitous discovery
The new results by Roberto Decarli and collaborators described here have shed new light on this question – albeit serendipitously, as the astronomers' initial aim had been somewhat different. Using the ALMA observatory, they were looking for very distant star-forming host galaxies of quasars. Since quasars are galactic nuclei, each is embedded in what is known as its host galaxy. There have long been questions about the interaction of quasars with their host galaxies – do they, for instance, inhibit star formation in the galaxy surrounding them?
More generally, what are the properties of these host galaxies – and are they related to the fact that the galaxy is hosting a quasar? To address such questions, Decarli and his colleagues studied known quasars so distant they represent the first billion years of cosmic history – and in targeting these quasars, they looked specifically for emission associated with star-forming activity.
Signs of star formation activity
Star formation involves gas clouds collapsing under their own gravity. If gravity is strong enough to compress the central regions to such high densities, and heat them to such high temperatures, that nuclear fusion sets in, turning hydrogen nuclei (protons) into helium. The result is, by definition, a star: an object bound by its own gravity, with nuclear fusion in its core region, shining brightly as the energy liberated during the fusion processes is transported outwards. But in order to reach these high densities, and such an advanced state of collapse, the cloud needs to cool down during the collapse.
That is surprisingly difficult: Hydrogen molecule, it turns out, are not very efficient in radiating away heat in the form of light. Most of the cooling-down is mediated by a kind of atom that occurs only very rarely in such collapsing clouds, but is able to radiate energy very efficiently: carbon. There are typically only three carbon atoms for each 100,000 hydrogen atoms in a modern-day star-forming environment, but in particular in its singly ionized form, with one electron having broken free from the atom, carbon is a highly efficient radiator, shining brightly in a very narrow frequency range known among astronomers as the [CII] line.
(The square brackets indicate that this is a line that is only visible under the rarified conditions of outer space – in laboratory experiments at higher gas density, the atoms in question are more likely to lose their energy by colliding with other atoms, before they can radiate [CII] light.)
Starforming regions are the main source of [CII] light in galaxies. Conversely, by measuring the amount of [CII] light emitted by a galaxy, one can estimate the rate at which that galaxy is forming new stars.
Distant star formation with ALMA
For close-up objects, the [CII] line has a wavelength of 158 μm, in the far infrared range of the spectrum. Unfortunately, the Earth's atmosphere is virtually opaque for light at that wavelength, and observations of this kind can only be made by airborne or space observatory, most recently SOFIA and Herschel.
For very distant objects, though, there is an additional effect that makes ground-based observations possible. For an observer on Earth, the light of very distant objects is stretched by the so-called cosmological redshift, an effect of the expansion of the universe. For the galaxies and quasars that Decarli and his colleagues were aiming at, light is stretched by a factor of about seven (corresponding to a z value z ~ 6), bringing the line into the millimeter wave regime, which is observable using ground-based telescopes like ALMA. That allows for high-resolution, sensitive observations.
ALMA is a telescope array composed of about 50 high-precision antennas, operated by an international consortium in the Atacama desert in Chile, and represents a significant increase in sensitivity over previous such observatories. Before the present study, [CII] studies on high redshift ('high-z') quasar host galaxies had only been done in small samples (with up to four quasars per study). With ALMA, bigger samples became feasible: Decarli and his colleagues obtained sensitive [CII] data for 25 galaxies.
Not the galaxies they were looking for
And for four of these targets, the astronomers were in for a surprise. Yes, there were quasars in those images, but there were galaxies as well. Not the quasars' host galaxies, but companion galaxies, each a little offset from the quasar target. And these were galaxies that were shining brightly in [CII], evidently forming more than a hundred solar masses' worth of stars per year. In galactic terms, that is quite a lot. Our home galaxy, for instance, forms no more than one solar mass per year. The other galaxies astronomers had previously found in this period of the early universe had star formation rates between one and ten solar masses per year.
The objects observed by Decarli and colleagues are so distant that we see them as they were a bit more than 900 million years after the Big Bang (z ∼ 6). But at that rate of forming new stars, these galaxies could indeed be the precursors of the star-rich galaxies found by Straatman and her colleagues at 1.5 billion years after the Big Bang (z ∼ 4).
The group around Decarli found a missing piece of the puzzle of cosmic history: A population of young, vigorously star-forming galaxies at a time 900 million years after the Big Bang. If this type of galaxy is sufficiently common, it could explain the unexpectedly star-rich galaxies about 600 million years later.
Quasars, overdensities and star formation
In all probability, finding these galaxies so close to quasars is no coincidence. The details will need to be examined much more thoroughly, including additional observations, but one general correlation suggests itself: In order to explain how the black holes driving quasars were able to amass a billion solar masses that early in the history of the universe, these quasars should be located in the highest-density regions of the universe at that time. It is plausible that the same overdense environment was conducive to the formation of the newly found, quickly star-forming galaxies as well. Thus, one would be more likely to find these galaxies in the neighbourhood of quasars.
Either alternatively or in addition, it is possible that the quasar's activity encouraged the nearby galaxy to form more stars, for instance by pushing on that galaxy's gas from the outside, setting off more local cloud collapses than would otherwise have happened. If these newly discovered active galaxies are representative of a more widespread population of vigorously star-forming galaxies in the very early universe, occurring even in the many regions where there are no quasars (albeit more rarely), they would be sufficient to account for the massive, evolved galaxies discovered by Straatman and collaborators.
The first known merger?
One of the four objects, the quasar with the catalogue number PJ308-21, is particularly interesting. Its star-forming companion galaxy is comparatively close to the quasar, and appears to be stretched out into a long shape towards the quasar. This kind of deformation is to be expected if the companion galaxy is interacting with the quasar host galaxy.
This kind of interaction, each galaxy distorted with tidal forces of the other galaxy's gravity, commonly is the prelude to the merger of these galaxies, resulting in the formation of a larger single galaxy. In the current models of galaxy evolution, this is a key mechanism for how galaxies have grown in the course of cosmic history. If the new observation indeed shows a galaxy merger, it would be the earliest known such merger.
All in all, the newly discovered population has shown us one piece of the cosmic narrative, namely how the somewhat later, star-rich galaxies formed. It is also pointing astronomers in a specific direction to find out more about the history of the early universe, namely towards an investigation of the role of overdensities, and of possible interactions, in the formation of the quasars and their companions.
Next, Decarli and his colleagues will need to fully characterize their newly discovered sources: Since these galaxies do not show obvious signs of accreting central black holes, which would outshine the faint stellar emission of the host galaxy, and which might influence star-formation in the galaxy, these newly discovered galaxies are ideal laboratory to study the first stages of the formation of massive galaxies. What kinds of stars do they contain, and in what proportion? What is their total mass, and how many stars have already been formed in these galaxies? What are the properties of the gas between the stars in these galaxies, the interstellar medium – how dense is it, what is its temperature, what fraction of it is ionized? And are these galaxies indeed only found very close to quasars, or do they exist in other environments, as well?
Answering these questions will require a whole battery of telescopes: from ALMA via the Hubble Space Telescope and the Spitzer Space Telescope to various ground-based telescopes and, in the immediate future, the James Webb Space Telescope. But by analyzing the data from these telescopes, with their different specializations and strengths, astronomers should be able to write a detailed version of this particular chapter of cosmic history: how the earliest massive galaxies came into being.