That the Universe is largely composed of a cosmic web consisting of narrow filaments upon which galaxies and intergalactic gas and dust are concentrated has been known for more than a decade. While a great deal of evidence for this has accumulated, visual evidence has been difficult to find. Astronomers have now photographed what appears to be a segment of a cosmic filament stimulated into fluorescence by irradiation from a nearby quasar.
The filaments of the cosmic web are difficult to see visually. They consist primarily of dark matter and intergalactic gas and dust, none of which have a visible signature detectable across billions of light years. As a result, our knowledge of filaments primarily comes from gravitational lensing studies, radio observations, and x-ray telescopes.
Now a team, led by researchers at the University of California, Santa Cruz (UCSC), has found an unusual configuration of celestial objects that appears to make visible a part of a filament that is ten billion light years distant. The section of the filament that is visible takes the form of a huge asymmetric nebula of diffuse intergalactic gas.
Normally this gas would not emit significant amounts of light, but in this case the intergalactic gas is being irradiated by extreme UV light from a nearby quasar; the active center of a galaxy. This irradiation ionized the gas (mostly consisting of atomic hydrogen), which then emits the characteristic light of atomic hydrogen (Lyman-alpha radiation) when the ionized atoms regain their electrons. When redshift (z~2.27) is taken into account, the Lyman-alpha radiation appears to our instruments as a violet glow, as seen in the lead photograph.
Let's start by taking a look at the observational evidence for a web of cosmic filaments in the Universe. (A cosmic web also appears in simulations of the evolution of the Universe, but data are better than simulations any time.)
The astrophoto shown above is of a huge and nearly one-dimensional cluster of galaxies that was found by the Sloan Digital Sky Survey (SDSS) along a cosmic filament.
The map above is also a product of the SDSS, which used a 2.5 meter telescope to image and determine redshift (and thereby distance) for galaxies in the cosmic vicinity of the Milky Way galaxy. It includes galaxies and quasars located in a thin slice of the sky above the Earth's equator out to a distance of two billion light years. One's first impression is of a slice through a foam of luminous bodies that lay on the boundary of huge voids.
Rather solid evidence also exists for the existence of filaments with a goodly share of dark matter, as illustrated in the above figure of just such a dark matter filament. This filament stretches about sixty million light years between the galaxy clusters Abell 222 and 223. X-ray emissions from the filament suggest that nearly 10 percent of the filament's mass consists of hot gas. This filament comprises at least dark matter and intergalactic gas.
The team published a report in the January 19 issue of Nature of their discovery of a rather unusual configuration of celestial objects in the early history of the Universe (about three billion years after the Big Bang) that provides additional evidence for the existence of the cosmic web. The lead photo is an astrophoto, taken in redshifted Lyman-alpha radiation from hydrogen gas using the 10-meter Keck I telescope in Hawaii. It shows a distant (z~2.27) quasar named UM 287 that appears to be surrounded by a glowing nebula of diffuse gas. However, this is no ordinary nebula.
The above map of the region surrounding the nebula appears to show no object near the nebula that could cause it to be excited into fluorescence other than UM 287. (The small quasar at approximately the same distance as UM 287 is far too dim to light up this nebula.)
Taking as a hypothesis that UM 287 and the nebula are at the same distance, the researchers were able to test the idea. The figure above shows the neighborhood of UM 287 taken on the left in redshifted hydrogen (Lyman-alpha) light, and on the right in a broad range of wavelengths that excludes redshifted hydrogen emissions. That the nebula appears in the redshifted Lyman-alpha light on the left image and not on the right tells us that the nebular light is line emission; specifically hydrogen emissions, and that the redshift of the nebula and the quasar are essentially identical.
The difference in the Lyman-alpha and broad spectrum images allowed the team to subtract out the background radiation, giving the above detailed image of the nebula including only the Lyman-alpha emissions. The purported filament, while far smaller than that observed between Abell 222 and 223, is much clearer in this view.
The nebula is about a minute of arc across in the photograph. (It may actually be much larger, if it extends beyond the quasar's ability to make the nebular gas fluoresce.) Using the standard cosmological model to trace back through the expanding spacetime to that period in the distant past, the size of the nebula is about 1.5 MLy (million light years). This is larger than any previously known Lyman-alpha source, leading to the question of just what the nebula might be.
Indeed, the extent of the nebula is too large to be gas gravitationally trapped by the quasar, even taking into account the quasar's halo of dark matter. UM 287 is a radio-quiet quasar, and such quasars with luminosity equal to that of UM 287 have never been seen to have such extensive halos. Such a halo would have a mass over ten times larger than the largest known quasar halo.
So large a halo causes clumping of the matter contained therein, and observationally this appears as a high density of localized Lyman-alpha sources. The level of clumping observed around UM 287 also indicates a much smaller halo, in all likelihood extending no more than 300 thousand light years or so. The evidence supports that the nebula found in the region surrounding UM 287 is formed of intergalactic gas.
“We have studied other quasars this way without detecting such extended gas,” said Dr. Sebastiano Cantalupo, a member of the UCSC team. “The light from the quasar is like a flashlight beam, and in this case we were lucky that the flashlight is pointing toward the nebula and making the gas glow. We think this is part of a filament that may be even more extended than this, but we only see the part of the filament that is illuminated by the beamed emission from the quasar.”
So how much gas did they find? Any hydrogen atoms that can emit Lyman-alpha radiation must be cool enough to recapture their electrons once ionized, as this recapture is what generates the radiation. This calls for a temperature smaller than about 50,000 K. At larger temperatures, the electrons have too much energy to be captured, so "hot" gas does not add to the observed nebular emissions; we can only see cold hydrogen.
The researchers estimated the amount of gas in the nebula based on the ability of the quasar's radiation to penetrate the nebular gas and then to effectively stimulate an hydrogen atom to emit a Lyman-alpha photon. These estimates included two extreme cases as shown above, if all the nebular gas is ionized (left), and if none of the nebular gas is ionized (right). In both cases, the mass of the filament is found to be about a trillion solar masses.
This mass is at least ten times more than indicated by computer simulations of the Universe's evolution, and the filamental web that results. It is somewhat embarrassing; if the entire cosmic web were ten times more massive than expected, the Universe would have too much mass to be in its current configuration.
There are at least four explanations for this discrepancy. Obviously, and perhaps most likely, the simulation may not include all the physical mechanisms required to accurately model the details of the filaments.
Another limitation of the simulation is that it has a smallest size of about 30,000 light years, which is similar to the optical resolution of these observations. The amount of gas required to explain the observations of UM 287's nebula becomes smaller if that gas is concentrated into clumps of higher density. If such clumps do evolve and are typically smaller than, say, 10,000 light years, the observations would not detect them, and the simulations would not reproduce them. However, the clumping needs to be rather extreme to reconcile the observations with the filament simulation.
Third, the particular region of the filament that is illuminated by the quasar might be atypically dense for some reason that is presently unknown. The study authors have done a good job of ruling out a past gravitational interaction between UM 287 and the dimmer quasar, which are at very nearly the same distance and are separated by less than some ten million light years.
Finally, the nebula may not be a cosmic filament at all. While it is tempting to make this identification, and most of the observational evidence is at least compatible with the conjecture, there is no smoking gun that converts this conjecture into a "most likely explanation". Ideally, a smoking gun would allow one to argue backward from the observational data to conclude that a web of cosmic filaments exists. One observation of a configuration that is suggestive of being a filament doesn't cut it.
In the end, the observations of the atypical nebula surrounding UM 287 reveal an unusual situation calling out for attempts to obtain more data and to seek out other examples of the phenomenon, but do not lead to the conclusion that the intergalactic gas in a filament has been observed. A reasonable conjecture? Yes, but nothing even close to a smoking gun. This will be an interesting story to follow.