But the Lord was not in the wind; and after the wind an earthquake; but the Lord was not in the earthquake: And after the earthquake a fire; but the Lord was not in the fire: and after the fire a still small voice.
—1 KINGS 19:11–12 (KING JAMES VERSION)
To visit the Laser Interferometer Gravitational-Wave Observatory (LIGO) facility in Livingston, Louisiana, you need unwavering faith in your GPS system.* Tucked in thick piney woods 30 minutes east of Baton Rouge, this powerhouse of technological achievement—true to the phenomena it observes—announces its Nobel Prize–winning marvels quietly with a small roadside sign. This modesty makes perfect sense. Unlike the enormous hunting and fishing superstore I had passed on the interstate on my way there, LIGO has little interest in drawing traffic to its grounds.
Upon being ushered through the security gates, one begins to sense the significance of the work carried out there. Corridors of pine trees have been felled to make room for the enormous pipe-like tunnels that reach four kilometers to the south-southeast and the west-southwest, a perfect right angle on the landscape. These tunnels converge on the science facility, its stark blue and white architecture practically an homage to right angles.
Across the parking lot is the Science Education Center, open to anyone willing to make the trek on the third Saturday of any month. Here, the LI (laser interferometer) part of LIGO is illustrated by dozens of hands-on exhibits. Laser interferometry hinges on a well-known wave property: interference. Periodic waves consist of regularly alternating peaks and troughs, and when waves interact, occasionally the peak of one will correspond to the trough of another, canceling out the waves in that patch of space and time.
LIGO’s observations start with 200 watts of laser light at a wavelength outside the range of human perception. The thin pencil of infrared light hits a beam splitter, an aptly named device that divides the light into two perpendicular beams, each of which races down a four-kilometer-long arm of LIGO. A hair over 13 microseconds later, the beam hits a hefty 40-kilogram mirror and heads back toward the beam splitter.
In a rigid universe of starched and unflappable tapestries, each beam would always travel exactly the same distance out and back. Upon returning, the beam’s peaks and troughs would line up as they had initially, and the recombined beam would have the same profile as the outgoing one. But in the wobbly, dynamic universe that we inhabit, one arm of LIGO might be stretched a bit compared to its perpendicular partner. In this case, a trough from one returning wave might line up with the peak from another. The details of the newly comingled profile would show exactly how much the relative arm lengths changed and when. And that would tell scientists what caused the distortion.
In principle, LIGO is elegantly simple. Demonstrations illustrating the underlying concept have been performed on late-night television and in traveling science shows. Having seen these, you might wonder why it took a century and over a billion US dollars to make LIGO work. Moreover, why do scientists find it necessary to have so many of this sort of observatory? LIGO Livingston is tucked away in Louisiana, but 3,000 kilometers to its northwest in the state of Washington is LIGO Hanford. Across the Atlantic near Pisa, Italy, is Virgo, a right angle with three-kilometer-long arms funded by the European Gravitational Observatory. Just to Virgo’s north is GEO600, a gravitational wave observatory in Germany with 600-meter-long arms. In Japan, there’s the Kamioka Gravitational Wave Detector, deep underground with its three-kilometer-long arms embracing the vastly upgraded neutrino observatory that caught a dozen ghost particles from SN 1987A.
The Hanford and Livingston observatories will eventually be joined by LIGO India, a collaborative effort between the United States and India. On top of that, the next generation of gravitational wave observatories is already in the works. The coming years will see construction in the United States of the Cosmic Explorer, a LIGO-like setup with 40-kilometer-long arms, and the Einstein Telescope, a triangular observatory 10 kilometers on each side, in Europe. The Einstein Telescope and Cosmic Explorer promise to be at least ten times more sensitive than anything currently operating.
But sensitive to what?
Imagine someone playing a slide whistle, starting with a low note that gradually shifts to higher and higher pitches and ends with a quick, high chirp. Now imagine that the person is standing across a crowded room, and you have mild, persistent tinnitus. And there’s a fountain in the corner, a ticking clock on the wall, and an exhaust fan in the adjacent kitchen. Further, the entire building is beneath the flight path of a busy airport.
If only gravitational wave detection were as easy as picking out that whistle.
Our best shot at finding ripples in spacetime lies in objects with the strongest gravity. Neutron stars will work, but black holes work even better. The problem is that physics starts to behave very oddly around things that put such steep dents into spacetime. And it gets even weirder when those things interact with each other, weirder still when they collide.
Einstein’s theory of general relativity, which describes the unintuitive relationship between matter and space and time, was just the starting point for researchers like Susan Scott, a chief investigator for the Australian Research Council Centre of Excellence for Gravitational Wave Discovery. One of the things that she and her colleagues have wrestled with for decades is what kinds of ripples to expect in an encounter between two massive, compact objects.
“We hit a bit of a roadblock in the numerical relativity side of things,” Scott recalled of her work in the early 2000s. The behavior of the messy, dynamic spacetime near a black hole binary could not be determined by solving a few equations and confidently circling the answer. The shape of spacetime is continually and radically changing as the system loses energy to gravitational waves and as the steep dents in spacetime warp each other. These are not easy things to figure out, even with the help of supercomputers.
She continued, “While the instruments were being built from the 1990s onwards and being made more sensitive, the binary black hole system had a number of modeling problems.”
In other words, while one part of LIGO was steadily creating something capable of detecting the cat hiding in the data, another part hadn’t quite worked out that they should even be looking for a cat. Perhaps a dog template was called for.
“There was a big breakthrough around 2005 by a number of groups around the world,” Scott said. “That enabled us to have really accurate predictive wave forms.”
You see, it isn’t enough to imagine the sound of a slide whistle. Gravitational wave researchers need to know what the exact notes are. How long does it take to slide to the higher pitches? How loud is the whistle? Everything about the evolution of the frequency and loudness is crucial because it reveals something about the source. More massive binaries like black holes are analogous to louder, shorter-duration whistles. Lower-mass objects—binary neutron stars perhaps—are lingering, quieter whistles as their shallower dents in spacetime spiral more gradually toward merging into a single pit.
The 2005 breakthrough allowed scientists to model exactly what gravitational wave observatories would see if, say, a black hole with 5 times the mass of the Sun crashed into another with 20 times the mass of the Sun 2 billion light-years away, or if binary neutron stars collided 300 million light-years away. They could even disentangle information about the tilt of the binary orbits relative to Earth. Are we seeing the system edge-on? Or are we looking down at the system from some angle?
“We were ready,” Scott said triumphantly. “We had the templates, and so when a signal came in, we would just measure it up against all these templates and get the one that fit.”
By 2005, LIGO Livingston and LIGO Hanford were mostly up and running after a decade of site clearing, construction, testing, upgrades, more testing, and more upgrades. And now they had the templates in hand. Or, rather, in computer. In 2006, they began the search for gravitational waves in earnest.
Five years came and went, and no signal was definitively detected. While this was disappointing, it was not ultimately surprising.
“The thing is,” explained Scott, “gravity is the weakest force. And that weakness affects everything related to gravitational wave science. The [waves] sail through the material universe, and their effect on things isn’t as great as you might imagine.”
Scientists were looking for evidence that the arms of LIGO were stretching and contracting and that they were doing so in a way that betrayed the profile of a gravitational wave signal. But the amount of stretching and contracting expected from such an event is ridiculously small. Stupendously small. Unimaginably small. Smaller still.
Plenty of things that aren’t gravitational waves can make one of the two laser beams show up a little earlier or later than the other one. Waves from the Gulf of Mexico crashing into the shoreline of coastal Louisiana or tiny seismic shifts in the state of Washington can do the trick. Then there’s the force of the infrared photons hitting the 40-kilogram mirror and the fact that the structure itself is built from atoms that are not perfectly still. Any stray gas particles left in the four-kilometer-long vacuum tubes can mask the results, as can a car driving on the interstate miles away.
Gravitational wave scientists are listening for a faint slide whistle in a very noisy room indeed.*
All these complications and more mean that there are certain gravitational wave frequencies that LIGO is best suited to detect. Its sweet spot lies between about 10 and 2,000 hertz,* but only if the gravitational waves stretch and compress the arms enough for the signal to be disentangled from the noise. For LIGO’s first few observing runs, “enough” meant that the length of its arms would have to change by one part in 10 billion trillion as a gravitational wave signal swept through. That would add or subtract a mere millionth of a billionth of a millimeter from the four-kilometer-long arms. Spanning an impressive thousandth of a billionth of a millimeter in diameter, a single proton, the minuscule particle residing in atomic nuclei, is gargantuan by comparison.
And yet, by 2009 researchers had seen no hint of a ripple this significant in the fabric of spacetime. A sensitivity of one part in 10 billion trillion simply wasn’t good enough to spot anything in the great cosmic demolition derby.
Astronomers had much higher hopes for the next five years. Although much of 2010–2014 was spent upgrading the observatories to Advanced LIGO rather than searching for gravitational waves, there were still terabytes of data from the previous years that promised to harbor something unambiguous. But again, there were zero confirmed detections. A string of papers on the unsuccessful searches placed plenty of limits on the types of gravitational wave sources that were detectable with LIGO, but it had clearly been a disappointing half decade.
After 2015, despite the promise of Advanced LIGO’s greatly improved sensitivity, many senior astronomers remained cautious. Some had gambled on LIGO—and lost—before. Others felt that the first detections were practically inevitable once observations resumed in earnest.
“Personally, I never doubted that if the theory was correct, we would eventually detect them,” Scott recalled of these uncertain days. “I knew we were going to have the sensitivity in time.”
“I see your Skype icon is green,” LIGO member Eric Thrane typed to his colleague Paul Lasky. “Are you awake?”
Even though it was almost 11 p.m. in Melbourne, Lasky was still up, getting ready to teach a first-year class on quantum physics the next morning.
“How excited should I be?” Lasky responded.
Very, as it turns out.
It was 14 September 2015, and the new and improved Advanced LIGO was still technically a few days from beginning its first observing run. But all the equipment was functioning as scientists and engineers took it through its final dress rehearsal. They had no idea that an entirely new window to the universe was about to be flung wide open.
Finally, the squishing and stretching of the LIGO arms was extreme enough to detect.
The ripple makers were two black holes—one weighing in at 29 times the mass of our Sun and the other packing 36 times the Sun’s mass—that had spiraled toward each other in a frantic, ever-tightening, ever-accelerating orbit. As they did so, they sent off gravitational waves of increasing frequency and amplitude until they finally slammed together to create a mammoth black hole of 62 solar masses, give or take. The entire process took only a fraction of a second, and the black holes reached speeds of nearly two-thirds the speed of light in their death spiral.
The collision had happened over a billion years before the detection, which was made months before the announcement. This was nothing like spotting the obvious bright dot of a supernova. Peeling the signal from the noise required ten supercomputers to perform an analysis that would have taken a regular desktop computer 5,000 years to complete.
As is usually the case with groundbreaking scientific discoveries, the official proclamation had to wait for checking, double-checking, and triple-checking.
“I think there was still a significant amount of disbelief,” Lasky recalled. “I absolutely understood the consequences if the event was real. Over the next few days, it started to sink in that this event wasn’t going away, and it could not be explained any other way.”
The team had to write the requisite journal article, await peer review, and ultimately, in February 2016, publish the results. During this time, the LIGO researchers effectively demonstrated that scientists should never be charged with keeping secrets. By the official end of the “embargo,” the strict time frame during which results are not to be made public, everyone in the astronomical community seemed to know everything.
A professional astronomers’ Facebook group had been unofficially discussing the implications for weeks, and at one point, Lasky revealed in a presentation about the discovery, a letter circulated that read in part: “Hi, all, the LIGO rumour seems real and will apparently come out in Nature, Feb 11. . . . spies who have seen the paper say they have seen gravitational waves from a binary black hole merger. . . . the bh masses were 36 and 29 solar masses . . . apparently the signal is spectacular.”
This rumor was grossly inaccurate, joked Lasky. The journal wasn’t Nature. It was Physical Review Letters.
For those familiar with the world of science, particularly regarding scientific findings that are novel or even controversial, this tale is nothing out of the ordinary. When your research team has a thousand members, there is virtually no way that any kind of secret can be kept for long. Scientists know this to be true, but the general public doesn’t. That’s why there are so many people willing to believe that the entirety of NASA, an organization with over 18,000 employees on its payroll, is capable of a large-scale, decades-long conspiracy to hide the “truth,” whether about the Moon landings or visiting aliens or the shape of the Earth.
International news outlets exploded with the story of the discovery. LIGO scientists the world over were asked to demonstrate the slide whistle chirp that characterized the wave form. The internet brimmed with animations of enormous swells rippling through our neighborhood, each of which made it seem as if we were looking at Earth in a funhouse mirror. But GW150914 (GW for “gravitational wave” and 150914 for 14 September 2015) had done nothing that obvious. Earth itself, had it been a perfect sphere, would have been distorted by just a few trillionths of a millimeter. A still bathtub has more impressive ripples.
This isn’t to say that GW150914 was a dud. During the collision, it blasted out over 5,000 foes of energy—the energy equivalent of three solar masses—as gravitational waves. In other words, in a fraction of a second, this single event expelled more energy in gravitational waves than the total emitted energy of
Every
Single
Star
In
The
Observable
Universe.
Astronomers have seen individual stars explode ten times farther away than GW150914, and those supernovae released far, far less energy but were much, much more obvious.
But this event created no light, no neutrinos, no cosmic rays. Just gravitational waves straining against the obstinacy of spacetime. Had you been only 10,000 kilometers from this cosmic showdown, you would have been stretched and squashed by a scant millimeter as the gravitational waves zipped through you at light speed.
It is a powerful testament to the cleverness of humankind that (1) we calculated what kinds of ripples in the invisible fabric of spacetime would arise from a collision like this; (2) we figured out in principle how to measure these ripples; and (3) we actually pulled it off.
“That was an amazing sensation,” Lasky said animatedly, “to realize that we were witnessing and being part of history!”
“We were so lucky with that first detection. It was golden,” Scott recalled fondly. “The number of firsts in that event was just astonishing, actually. First detection of gravitational waves. First direct observation of a black hole. First real confirmation that we had binary black hole systems at all in the universe.”
Moreover, this discovery meant that gravitational wave astronomy had at last become an observational science. Now it was time to make it a statistical one.