A rare alignment of a quasar’s “flashlight” beam and a filament of the cosmic web illuminates the universe’s large-scale structure.

By now, most of us have seen the universe evolve — in simulations, that is. (See an example in the video below.) In these movies, gravity transforms the primordial soup into weblike structures of material, in which growing galaxy clusters are fed by gas that spools in along thin filaments of dark matter.

The simulations match the observed universe remarkably well. But although we can see galaxies throughout the universe, we rarely see the webs’ filaments. What little visible matter they contain is sparse gas — there are virtually no stars.

But Sebastiano Cantalupo (University of California, Santa Cruz and Lick Observatory) and colleagues think they’ve found a rare cosmic flashlight to illuminate a piece of the web.

Cosmic web filament

The Lyman-alpha blob (blue fuzz) surrounding the quasar UM 287 (white dot in center) extends 1.5 million light-years long, far too big to be contained within the quasar's host galaxy or the host galaxy's halo. The blue fuzz is likely to be part of a cosmic web filament.

S. Cantalupo / UCSC

As the team reported online January 19 in Nature, the astronomers observed UM 287, a quasar that lived 3 billion years after the Big Bang. Using the 10-meter Keck I telescope in Hawaii, the team witnessed cold hydrogen emitting Lyman-alpha radiation underneath the spotlight of the quasar’s intense ultraviolet beam.

Lyman-alpha blobs (yes, that’s a technical term) have been found before, but their origin is still mysterious. Generally, they’re thought to be associated with a single galaxy. But this blob is twice as big as previous discoveries. And at 1.5 million light-years across, it’s far too long to be contained within UM 287’s host galaxy, or even the galaxy’s much larger dark matter halo. That’s why Cantalupo’s team thinks this blob may be a cosmic filament, aligned by chance with the quasar’s beam.

Simulations of cosmic web

Computer simulations suggest that matter in the universe is distributed in a "cosmic web" of filaments, as seen in the image above from a large-scale dark-matter simulation (Bolshoi simulation, by Anatoly Klypin and Joel Primack). The inset is a zoomed-in, high-resolution image of a smaller part of the cosmic web, 10 million light-years across, from a simulation that includes gas as well as dark matter (credit: S. Cantalupo). The intense radiation from a quasar can, like a flashlight, illuminate part of the surrounding cosmic web (highlighted in the image) and make a filament of gas glow, as was observed in the case of quasar UM287.

S. Cantalupo (UCSC); Joel Primack (UCSC); Anatoly Klypin (NMSU)

Observers have seen hints of filaments before. Some reveal themselves indirectly when their gas absorbs light from distant quasars — different than the new study, which is of gas emission. And in a 2012 paper, Jörg Dietrich (University Observatory Munich, Germany) and colleagues reported the direct detection of a filament connecting two galaxy clusters, made visible by its gravitational distortion of background galaxies’ light and also by X-rays emitted from its hot gas. Unlike Dietrich’s study, Cantalupo’s study is looking at relatively cold gas.

“I believe the authors make a convincing case that they have indeed observed relatively cold filamentary gas,” Dietrich says. He adds that his team’s results and those in the new study are complementary, because they deal with gas in different states — one hot, one cold. “This will lead to a fuller understanding of filaments and their make-up on all scales.”

The UM 287 observations already point to a potential flaw in simulations. Going by its Lyman-alpha radiation, the blob stores an astonishing amount of mass — a trillion Suns’ worth — in its cold gas. But that amount is 10 times more than expected from cosmology simulations.

That’s not necessarily a surprise because “universe in a box” simulations tend to watch the cosmos evolve on extremely large scales, with “boxes” about 1 billion light-years across. “Small scales are not well resolved in simulations and gas processes on small scales are notoriously difficult to simulate,” Dietrich explains. So observations of individual filaments are the perfect laboratory for testing theory.


Image of Anthony Barreiro

Anthony Barreiro

January 28, 2014 at 12:59 pm

As an interested amateur with no formal education in astrophysics, I find one actual observation, even if it's just a fuzzy blob, much more interesting and compelling than a dozen beautiful computer simulations. I know that the models are helpful in generating hypotheses, but the proof is always in the pudding, or the Lyman-alpha blob, as the case may be.

You must be logged in to post a comment.

Image of P. Law

P. Law

January 31, 2014 at 5:22 pm

The Scientific Method, the hallmark of the Rennaisance of Science (as opposed to this New Science), has been completely abandoned, or turned on its head, if one follows any of the articles here and elsewhere. The New Scientific Method appears to regard simulation as the observational data, which may or may not correlate with real observations. Much is made to save a computer simulation at any cost, reminiscent of epicycles of Ptolemy.

Bunk. A clear sign of the disaster when scientists are left to do Philosophy.

You must be logged in to post a comment.

Image of clifford wright

clifford wright

January 31, 2014 at 9:12 pm

I was greatly heartened by the 2 first posters. How nice to see someone who understands what Science is all about.
These days we are overwhelmed by layers of weird hypotheses
based on nothing more than simulations. Indeed I have seen for myself the attempts by some "Scientists" to save a simulation by ignoring or discounting observation.
Remember GIGO!!!!!
At present we have at least 4 Totally unproven and (so far) totally unprovable hypothetical forces or unobservable matter all over the universe (except where we can detect it).

To think that string theory is criticised for its lack of experimental proof! It has nothing at all on 21st century Cosmology.
It's about time that Occam's razor got wielded!

You must be logged in to post a comment.

Image of Mike W. Herberich

Mike W. Herberich

February 7, 2014 at 11:56 am

Provided we have the knowledge and capability to do either, namely to observe AND to model, we (who IS we, actually?) should have AND do both. All else would be an inexcusable abdication of acquiring more knowledge, an inexplicable self-imposed limitation. --- Sure, it all is in the emphasis, weight or, worse, exclusiveness or playing off one against the other where the rub is. --- I suppose that people specializing in modelling do not do much more than just that, all day long ... which is still okay. It's not their fault if and when someone gives to much or all the weight to only their results. Conversely, people specializing in observing, maybe just one little particular thing, all day long: what are THEY doing wrong? Nothing either, I guess. It's in bringing these two exceptional and all-important things (and more!) together in just the "right" proportions (incidentally, what are they?) that conclusions are drawn and that knowledge originates, usually and nowadays. At least, this is what I hear: less and less we have just this one person in the ivory tower "doing science" (in the "right" way!?) all by him- or herself, is it?

You must be logged in to post a comment.

Image of Mike W. Herberich

Mike W. Herberich

February 7, 2014 at 12:14 pm

By the way, it is exactly by the method of Occam's razor that many of these "strange" and relatively new and "non-intuitive" things arose, such as the infamous dark entities (there is little in science that is intuitive AND correct!). All other explanations would be more complicated, secondary, or whatever one pleases to call it. --- Sure, next to string theories, etc., there are models like Robert's "(discrete) fractality all over" and one may fight over the exact amount of probability of each. But, I think, it is inherent to almost any scientifically inclined, let alone trained, individual to presuppose the Occam principle. In any case, it would turn out very hard to prove (scientifically!) the opposite! No scientist in his right mind would posit, say, a dark flow or energy or matter, just for the jest and fun of it or just to aggravate or annoy others. On the contrary, he or she would do all to obviate introducing new entities, so long as there is no absolute need to. Only as a last resort would one posit, say, a new particle, if and when all else fails.

You must be logged in to post a comment.

Image of Hamsterdam


February 7, 2014 at 2:24 pm

P. Law....you are Bang-on in your assessment. They are getting to areas in which Actual experiment cannot be done, so they do what I call "simulated experiments".

I tend to stick to Feynman's short, yet eloquent description of the process.

Good Show! This exposes the kind of confirmation bias that leads to mistakes, e.g., the faster than light neutrinos last yr.

You must be logged in to post a comment.

Image of Mike W. Herberich

Mike W. Herberich

February 10, 2014 at 6:26 pm

Weren't the faster than light neutrinos just a plain -and embarrassing!- measuring or calculating fault? Wasn't it detected immediately just for what we're discussing here: the first peer-review or re-enacting the experiment, respectively? And -boom! dead it was! --- Moreover, that was the opposite of modelling: just a plain and concrete experiment (falsely interpreted or executed, sure!)? --- And, sure, what could be MORE tempting for a scientist to -allegedly- perform an experiment that proves EINSTEIN WRONG!? --- That's where the confirmation bias came in, most likely! Throwing a roving eye on success, too much, too soon!

You must be logged in to post a comment.

You must be logged in to post a comment.