So many out-of-the-way things had happened lately, that Alice had begun to think that very few things indeed were really impossible.
—LEWIS CARROLL, ALICE’S ADVENTURES IN WONDERLAND
Trying to convey the constant low background rumble of gravitational waves without mathematical terms is, as Paul Lasky discovered, almost as difficult as detecting that rumble.
“So, if we’re envisioning the universe as the surface of a lake or something, and there’s this flock of geese landing all around you,” he attempted. “But far enough away that you’re not picking up any individual goose . . .”
He trailed off.
“They should be penguins,” I offered unhelpfully.
“Actually, penguins are probably better because they’re below the surface, and they’re coming up every now and then.” He stopped and thought. “But penguins aren’t in lakes, are they? I guess we’ll be in the ocean. . . . But now we’ve got other animals and ocean waves to battle against.”
He stopped again, considering his options. “I think we’ve probably taken the analogy a bit too far.”
My son, half listening to the conversation, interjected, “And it’s raining.”
“Ooh, rain is actually probably better,” Lasky said, brightening, and he returned metaphorically to the lake that would ripple with what scientists call “stochastic gravitational waves.”
“It’s raining on the surface of the lake, and I get the same ripple pattern coming off every single raindrop, but I don’t resolve any individual raindrops.”
I could finally see it. The raindrops—not geese, not penguins—are innumerable disturbances sloshing the lake of spacetime, and they’re happening everywhere all the time. In my mind, the lake is small, the raindrops hitting quickly, but I found that I needed to expand my mind to grasp what Lasky was talking about. A single tiny splash represented the eons-long in-spiraling dance of two supermassive black holes, each containing anywhere from millions to billions of times the mass of the Sun.
We are less than a water molecule—no, we are less than a fraction of a subatomic particle—in this picture. The wavelengths and time scales of the ripples are immense compared to our tiny realm of experience.
It is a big, old, bumpy universe indeed.
Although tiny, we are ingenious. Astronomers like Lasky have realized that somewhere on that lake’s surface is a way not only to detect, but also to characterize the chaos of waves perpetually rippling past. LIGO, Virgo, and other ground-based gravitational wave observatories with short kilometer-scale arms won’t cut it. Those observatories are best suited to picking up only gravitational waves with frequencies ranging from about 10 to 2,000 hertz.
Even the future space-based LISA, which Naoz patiently awaits, with three arms spanning millions of kilometers will be too small, a fact that’s astonishing when you consider that LISA will be able to detect gravitational wave frequencies down to about one ten-thousandth of a hertz. At such frequencies, only one wave would pass every few hours, corresponding to gravitational wavelengths of 3 trillion meters, or 3 billion kilometers. These ripples in spacetime could span the entire orbit of Saturn, and they arise from interactions that take hours to cycle. Such are the screams of unassuming compact objects, like small black holes or neutron stars, caught in the raging whirlpool near a supermassive black hole. Occasionally gravitational waves in LISA’s range of vision will arise from the dizzying final circuits of two supermassive black holes.
When you get two supermassive black holes wrestling in a galactic core, though, the dance is more powerful but so . . . much . . . slower. Forget hours. It can take decades for these objects to make a full orbit around each other, and the result is gravitational wave frequencies that are measured in nanohertz. That’s billionths of a hertz. It would take a billion seconds—nearly 30 years—for just one of these waves traveling at the speed of light to pass you. With waves so large—100 quadrillion meters or so—you need gravitational wave observatories with arms well beyond LISA’s scale of millions of kilometers. You need arms with lengths of hundreds of trillions of kilometers.
These are interstellar scales, and given that we haven’t even managed to get a spacecraft to travel more than a few light-hours away, the prospects for detecting such low-frequency gravitational waves might look to be pretty bleak.
But humans are clever, and as mystifying as the universe can be sometimes, it has a propensity for handing us some surprisingly useful tools.
The key to any gravitational wave detector is knowing what to expect if a gravitational wave sweeps through and changes the distance between objects. If the distance becomes compressed, a signal arrives a bit too early. If the distance is stretched, a signal arrives late. Conveniently, nature has placed dozens of objects with extremely well-timed signals in our midst: millisecond pulsars.
The concept of using pulsars as tools to detect extremely low-frequency gravitational waves predates the discovery of the first millisecond pulsar itself, so astronomers had no idea just how stable the timing of some of these objects could be. What if, Steven Detweiler mused in a 1979 paper, pulsar astronomers turned their attention from the troublemakers of the pulsar population—the glitchers and the nullers—and instead looked at the usefulness of the peaceful, quiet, well-behaved ones? There are, he argued, a few that might be up to the challenge of becoming a cosmic gravitational wave observatory, and if that is the case, then astronomers might find, er . . .
Why, they could find . . .
Admittedly, the exact source of the gravitational waves they might detect using pulsars was still a bit fuzzy. It would have to be something with an extraordinarily long wavelength. “The close encounter of two supermassive black holes at a cosmological distance might generate such a wave,” Detweiler suggested, with the caveat “but of course the existence of black holes and their interactions are matters of speculation.”
The case had nevertheless been made. If there are a number of pulsars with exquisitely precise timing, astronomers could simply watch for any changes to the pulse arrival times. If pulsar A’s pulses become unusually delayed and pulsar B’s pulses arrive unusually early, it is possible that a gravitational wave from a powerful but lengthy process has washed over us.
The idea percolated in the back of astronomers’ minds for a couple of decades. By the late 1990s, the Arecibo Radio Telescope and the Green Bank Telescope had been steadily adding millisecond pulsars to the roster, and astronomers had chosen a particularly well-behaved subset of them in a first stab at the concept. In mid-2004, the Parkes Pulsar Timing Array (PPTA) officially ramped up, using the Dish to look for any hints that gradual spacetime deformations were betrayed by the universe’s most stable clocks. In 2007, the North American Nanohertz Observatory for Gravitational Waves (NANOGrav) threw its hat into the ring, and the European Pulsar Timing Array added the power of several more radio telescopes to what would ultimately become the International Pulsar Timing Array (IPTA).
Tragically, because it’s located in the dynamic mess of a crowded globular cluster, Zippy did not make the cut. Teasing apart any putative gravitational wave signal from minute discrepancies in pulsar timing is hard enough without worrying about the effects of the pulsar’s neighborhood. Two dozen millisecond pulsars did make the initial cut, a number that grew with each passing year, giving the pulsar timing arrays far more arms than LIGO. But those arms had problems that LIGO’s didn’t.
I’m not saying that LIGO’s observations don’t come with their own challenges. When LIGO sees the merger of a binary black hole system, the signal carries in it information about the two masses: three numbers to describe the spin of each one (think: pitch, yaw, and roll) and how elongated or circular the orbit of each black hole is.
Lasky rattled off some other considerations. “Then there are some polarization angles, sky location, time of coalescence. . . . I’ve lost count.” The computer models need to tease out well over a dozen parameters. “It’s tough,” he admitted. “But it’s doable.”
Doable, I thought, as long as you have supercomputers churning through the data.
“If I’m doing pulsar timing, though, each pulsar has lots of parameters. And that’s just for one pulsar. What we’re looking for are the correlations between all the pulsars, and the PPTA is looking at dozens.”
LIGO’s work was beginning to sound like child’s play in comparison.
He continued, “The problem becomes absolutely intractable, so we don’t do it properly. We analyze it in sort of a piecemeal way, looking at each pulsar, getting the best timing model for each pulsar, and then combining all that after the fact.”
Lasky hesitated, as though preparing to tell me the really bad news. “Then, of course, there are the solar system parameters.”
These are the sorts of things that most people never think about because they’re irrelevant to everyday life. The reality is, though, that someone living on the equator travels the entire circumference of Earth every day as our planet whips around its axis. This means that they’re moving nearly 1,700 kilometers (1,000 miles) per hour, while someone in the mid-latitudes is covering perhaps half that. The planet itself is racing around the Sun at a whopping 30 kilometers (18 miles) per second and not in a perfectly circular orbit. Then there’s the added complication of the Moon, which acts a bit like Earth’s binary companion, jostling Earth to and fro as the two waltz around the Sun. The Sun is similarly pulled around by the gravitational tugs of the planets in the solar system. Jupiter nudges it on a 12-year cycle, while Saturn’s weaker tugs take nearly 30 years to play out.
As much as we’ve explored it, we simply don’t know our own neighborhood as well as we’d like. “It’s remarkable,” Lasky said. “To land a spacecraft on Mars, you don’t need that much precision.”
But to succeed in using pulsars as very long gravitational wave observatory arms, researchers have to know exactly where all of their receivers are and what they are doing relative to the ends of those arms. Moreover, they have to understand precisely how local dents in spacetime or even coronal mass ejections from the Sun are affecting what they see.
But wait, there’s more!
“We then have time-related issues,” George Hobbs explained. “We measure our arrival times with an observatory clock, and we have to relate those times to the best terrestrial time standard.”
He took a breath. “And then we have the issue that our instruments do strange things. You only need to spend some time at Parkes to realize that everything isn’t perfect. Cables change. The receiver system gets upgraded. Somebody changes something.”
Amazingly, the physics behind the behavior of millisecond pulsars is the least of researchers’ worries. Pulsar timing arrays don’t hinge on understanding the particles compressed in a neutron star’s core or the exact mechanism for generating the complicated magnetic fields.
In other words, millisecond pulsars are puzzles, but at least they are useful puzzles.
By design, the International Pulsar Timing Array is in it for the long haul. Looking for variations over the course of decades, IPTA astronomers have no need for the sorts of instant triggers that LIGO uses to alert the electromagnetic community to a potentially observable mashup. Still, by the middle of 2015, they were hoping that our planetary ferry would have ridden over some obvious swells. With Advanced LIGO scheduled to be operational by the third quarter of the year, time was running out to be the first team to detect ripples in spacetime.
Then GW150914 crashed through.
So much for the running joke among PPTA members that showed Einstein at a chalkboard with “PPTA > LIGO” scrawled on it.
It was okay though.
No, really. It was fine.
For one thing, Lasky pointed out, the groups are complementary, not competitive, and each looks for a signal that the other can’t detect. For another, there is significant overlap in the personnel involved. Lasky, like many others, is a member of both the LIGO Scientific Collaboration and the PPTA.
Scientifically, not getting an answer is, itself, an answer.
The raindrops on the metaphorical lake are countless years-long interactions between binary supermassive black holes. These interactions arise because the galaxies harboring these beasts ever so slowly and ever so messily joined forces in a cosmic mashup deep in the past. After the dust settled, what used to be two galactic hearts spiraled into one. Astronomers have countless snapshots of these galactic collisions playing out all over the universe, implying that out there somewhere, there are supermassive black hole binaries in all possible stages of the dance, each pair a single raindrop making ripples on the lake.
Our own Milky Way is not immune. In approximately 5 billion years—about the time the Sun will engulf Earth and the other planets—our Galaxy will have a life-changing run-in with the Andromeda Galaxy, spraying out stars, gas, black holes, neutron stars, white dwarfs, and more. Will our two central black holes merge quickly, at least cosmically speaking, into a single superduper massive black hole? Or is there some physical process that could prevent this from happening, sentencing them instead to circling each other for untold eons?
Astronomers don’t know how things will play out once these powerhouses get within a few light-years of each other. Admittedly, they’re not entirely certain about the processes leading up to the formation of these central black holes in the first place. That’s one thing scientists are hoping to pin down using the pulsar timing arrays.
In 2016, after a decade of null results, NANOGrav member Justin Ellis valiantly attempted to put a positive spin on it: “We are now at a point where the nondetection of gravitational waves is actually improving our understanding of black hole binary evolution.”
It’s phenomenal that not witnessing something pushes astronomers to improve their models. Obviously, they’d prefer an unambiguous signal, but as NANOGrav’s Ryan Lynch proudly pointed out, “Our secondary science is better than most people’s primary science.”
In the meantime, ground-based gravitational wave observatories were just warming up.