It’s a basic scary-movie rule: Never show the big, bad monster. Show his shadow, his tooth marks, his trail of slime. But as soon as you show his face, the fright bubble pops.
The universe seems to understand this rule. Its biggest, baddest monsters—supermassive black holes that haunt the center of nearly every galaxy, containing as much mass as millions or billions of stars—are totally invisible. Sure, by looking at the way stars whip around the center of the Milky Way and observing the electromagnetic fireworks going off in faraway galaxies, astrophysicists infer that supermassive black holes are lurking there. But they can never see the black holes themselves. And that’s not just storytelling—it’s physics.
While black holes don’t give off light, they do make a lot of noise in the form of gravitational waves, ripples in space-time that radiate out from massive objects when they change speed or direction.
Until recently, we couldn’t sense those ripples. But thanks to LIGO, the billion-dollar gravitational wave detector based in Hanford, Washington and Livingston, Louisiana, physicists can finally “hear” and study gravitational waves from black holes and neutron stars. Yet the objects LIGO can discern, exotic as they are, are paltry compared to supermassive black holes. They are like the flies that buzz around the scraps the monsters leaves behind. Now, physicists may be beginning to close in on the monsters themselves.
Lost and Found
In September 2015, two things happened: LIGO, the National Science Foundation’s flagship discovery machine, detected gravitational waves for the very first time, a headline-making, Nobel-winning discovery that set champagne corks popping in laboratories all over the world when it was announced publicly the following February. Meanwhile, on the other side of the globe, radio astronomers at the Parkes Observatory in Australia declared gravitational waves officially missing.
It might sound like a contradiction, but it’s not. The two experiments look for very different kinds of gravitational waves, with very different equipment, though they exploit the same basic physics. As gravitational waves ripple across the universe, they stretch and squeeze space-time like a cosmic accordion, pulling some objects closer together and pushing others farther apart. LIGO uses lasers and mirrors to measure those changes in distance along two perpendicular, two-and-a-half-mile-long light paths. The Parkes experiment uses radio stars called millisecond pulsars to measure the distance change across the entire Milky Way.
Just as there’s a whole spectrum sound waves, from the low rumble of the double-bass to the high quaver of the violin, there’s also a spectrum of gravitational waves. The frequency of a gravitational wave keeps the beat of the orbit that’s generating it: two black holes in a quick, tight orbit squeak a high note, while a pair in a wide, slow orbits groan a low tone.
LIGO is tuned to hear high notes, the one-time, now-you-see-it-now-you-don’t waves that cycle thousands of times a second, like keys plinked at the top of a piano. This sort of gravitational wave is most likely to come from black holes and other dead stars that have about as much mass as the Sun.
Pulsar timing arrays like the experiment at Parkes, on the other hand, are all about bass. They “listen” for low-drone gravitational waves that oscillate anywhere from once a year to once a decade. These gravitational waves come from supermassive black holes hundreds of millions of times heavier than LIGO’s finds, and they are impossible to detect with an instrument like LIGO because they get drowned out by seismic motion on Earth.
Astrophysicists think that one of these behemoth black holes lurks at the center of nearly every galaxy in the universe. When galaxies merge—a process that’s happening all the time, all over the universe—their central black holes should merge, too, releasing a brief, space-rattling crash of gravitational waves.
The odds of detecting the collision itself are low, at least with today’s technology. But for tens of thousands of years before the supermassive black holes actually collide, they orbit each other in leisurely loops, giving off low-frequency gravitational waves that cycle every year or so. These gravitational waves aren’t as strong as the sudden blast that comes out of the final smash-up, but because they go on for so long, and because there are so many merging galaxies in the universe, they add up, creating “an amazing, cacophonous noise” like an orchestra tuning up, says astronomer Scott Ransom, one of the founding members of the North American Nanohertz Observatory for Gravitational Waves (NANOGrav).
This noise, called the gravitational wave background, comes from “the superposition of gravitational wave signals from all the supermassive black holes that ever merged in the history of the universe,” says Xavier Siemens, director of the NANOGrav Physics Frontiers Center and an associate professor at the University of Wisconsin–Milwaukee. That’s the signal the Parkes team was trying to find.
A Galaxy-Sized Telescope
Scientists first floated the idea of using pulsars to measure the gravitational wave background in the 1970s, soon after pulsars were discovered. Pulsars are dead stars that cram a sun’s-worth of mass into a ball the size of a small city—not quite dense enough for the full black-hole cave-in, but close. They are draped in strong magnetic fields that channel charged particles, setting up beams of radio waves at the pulsar’s north and south magnetic poles. As a pulsar rotates, those beams sweep through space like lighthouse beacons, showing up at radio telescopes on Earth as startlingly regular pulses.
Typical pulsars rotate about once a second, too slow for the kind of precision measurements that could make out the gravitational wave background. But in 1982, astronomers discovered a new kind of pulsar, now known as a millisecond pulsar, that spins about 1,000 times faster than the “standard” variety. With millisecond pulsars as their clocks, astronomers can figure relative distances with exquisite precision. “That dramatic improvement in the precision made people rethink this idea,” Ransom says.
Today, there are three active pulsar timing arrays—the Parkes Pulsar Timing Array, the European Pulsar Timing Array (EPTA), and NANOGrav, which is currently the most sensitive detection experiment of the three. Each uses different telescopes to monitor different sets of millisecond pulsars, but they all operate on the same principle. And compared with LIGO, which stands as the National Science Foundation’s biggest-ever hardware buy, they are all dirt-cheap. The Zipcar to LIGO’s Bugatti, they share time on already-built telescopes with dozens or hundreds of other experiments.
“We get the pulsars for free,” Siemens says. “And that’s what makes this experiment so beautiful.”
Gravitational Wave No-Show
By 2015, the Parkes team had been tracking 24 pulsars for 11 years, looking for timing inconsistencies that could be due to the gravitational wave background. But they hadn’t found any. To the Parkes team, that suggested a flaw in our understanding of supermassive black holes and how they merge.
The leading theory is that galaxies form from the bottom up. Little galaxies combine to form bigger ones, which merge to make even bigger ones, and so on. When two galaxies join up, their central supermassive black holes should do the same. These aren’t head-on collisions: The black holes orbit around each other, first wide apart, then edging closer until, in a final cataclysmic swirl, they merge into one, releasing a great blast of gravitational waves.
Exactly how that final plunge happens still rankles astrophysicists. The black holes must be losing orbital energy; otherwise they would circle forever, like a perpetual motion machine. Simulations suggest that they should be shedding most of their energy in the form of gravitational waves. So why hasn’t anyone found the gravitational wave background yet?
There are plenty of possible explanations. Maybe supermassive black holes aren’t so common after all, or perhaps it takes galaxies and their black holes longer to merge than astrophysicists thought. Maybe nearby gas or the kinetic slingshot action of nearby stars speeds up the process, and the gravitational waves to quickly slip out of detectable range. Or perhaps the black holes prefer egg-shaped orbits over circular ones, leading them to merge more quickly. Pulsar timing array measurements aren’t yet able to discriminate between these scenarios, but as the experiments continue, researchers hope to be able to zero in on the likeliest explanations.
“We were very naïve at first,” says Maura McLaughlin, a professor of astronomy at West Virginia University and chair of the NANOGrav collaboration. “We generally worked with this assumption of completely gravitational-wave driven mergers, in which case we would likely have made the detection earlier..”
But Ransom says that talk of “missing” gravitational waves is “much ado about nothing.” The gravitational wave background is out there, he says. We just haven’t found it yet.
Are We There Yet?
This week, the NANOGrav team released a fresh helping of data, and there are hints that they may be getting close. The latest data including more than 3000 hours’ worth of pulses from 45 pulsars taken over 11 years at the Arecibo Observatory in Puerto Rico and the Green Bank Telescope in West Virginia. The team hasn’t found the gravitational-wave background yet, but they have established a new limit on how strong the background could be while still escaping detection. (Their analysis is posted online and has been submitted to the Astrophysical Journal.)
Their limit is lower than it was two years ago, the last time NANOGrav released data, but it’s not quite as low as astronomers were expecting. “That’s actually very interesting, because that’s what you would expect to see if you have gravitational waves in your data,” Ransom says, even as he cautions that it’s just too soon to say whether they are closing in on the signal.
At the same time, the NANOGrav team uncovered a flaw in all the previous analyses, including the controversial Parkes paper. To figure out if pulsar pulses are arriving early or late, astrophysicists have to know when pulses should arrive, to a matter of tens of nanoseconds. To calculate that expected arrival time, researchers make corrections for everything from inconsistencies in the pulsar’s spin to slowdowns from interstellar gas to the motion of the Earth relative to the pulsar.
That’s where they spotted a problem. To account for the Earth’s motion, pulsar astronomers had been using the same data tables that NASA uses to plot spacecraft trajectories. Those calculations, called ephemerides, are published by the Jet Propulsion Laboratory (JPL), and they are accurate to a few hundreds of meters. But it turned out that systematic errors in the positions were skewing the pulsar timing results. “They were good enough for them to get a spacecraft to Mars, but not good enough for us,” McLaughlin says.
The NANOGrav team delayed publishing their results for almost a year as they consulted with the JPL experts and tried to find a workaround. Finally, they developed a new way to account for the positions of the Solar System’s planets to within tens of meters. When they reanalyzed their data using the new technique, they found that the formerly rigid limits loosened. “Our previous limits—and presumably those set by other groups—were artificially small because of this issue,” McLaughlin says.
Ryan Shannon, an astrophysicist at the Swinburne University of Technology in Australia who was the first author of the 2015 Parkes analysis, thinks that new technique may be unduly conservative. “It is unlikely that an error in the solar system ephemeris would cancel out a gravitational wave signal and make a limit lower,” he says.
But Siemens thinks that all the old conclusions will need to be rethought. “Everyone needs to reanalyze their data in light of these findings,” he says. “We cannot trust the upper limits that had been reported previously.”
The new NANOGrav results still put a fairly tight limit on the magnitude of the gravitational wave background. “We can say that these supermassive black holes binaries are not merging solely due to gravitational wave emission. There’s other physics at play,” McLaughlin says.
Astrophysicists can’t yet nail down what that other physics is, but they hope that they won’t have to wait too much longer. One reason for optimism: Pulsar timing arrays get better with age. “As our data sets get longer and longer, we’re sensitive to longer periods, and models predict that the gravitational wave background also gets stronger at longer periods,” Ransom says. “We get this two-pronged improvement to our sensitivity: We’re getting more data, and we’re getting access to longer-period gravitational waves.”
At the same time, astronomers are finding more millisecond pulsars and adding them to their timing arrays; NANOGrav is currently following more than 70 pulsars. “If we double the number of pulsars, we double our sensitivity,” McLaughlin says.
But NANOGrav is counting on the Arecibo Observatory and the Green Bank Telescope, both of which have been teetering on the edge of the financial abyss. Green Bank is hustling for private contracts to fill the funding gaps left behind by the National Science Foundation, and leaders at Arecibo, which just escaped a full zeroing-out, are still trying to figure out how to manage a sharp cut in federal funding.
“We’re pretty concerned, frankly,” Siemens says. “If we lose either of these instruments, we lose a factor of two in sensitivity.”
And for pulsar astronomers, that’s a scary monster that is all too real.