Marco Drago saw the first gravitational wave on 14 September 2015.

Marco Drago saw the first gravitational wave on 14 September 2015.

Marco Drago

Here’s the first person to spot those gravitational waves

Today, physicists working with the Laser Interferometer Gravitational-Wave Observatory (LIGO) announced that after decades of effort they had detected gravitational waves—ripples in spacetime itself—set off by the explosive collision of two massive black holes. But which of the 1000 scientists who work on LIGO, a pair of gargantuan instruments in Livingston, Louisiana, and Hanford, Washington, was the first to see the long-awaited signal?

The honor fell to a soft-spoken postdoc who plays classical piano and has published two fantasy novels. His tale shows how elaborate plans devised to keep LIGO team members guessing whether a signal is real or a purposefully planted fake broke down, leaving one lucky physicist and, soon, the entire LIGO collaboration sitting on a thrilling secret.

Marco Drago wasn’t in Louisiana or Washington, or even the United States. Instead, the 33-year-old postdoc from Padua, Italy, was at his office at the Max Planck Institute for Gravitational Physics in Hanover, Germany, where members of the LIGO team work on data analysis. There, Drago oversees one of four data “pipelines,” automated computer systems that comb through the raw data coming out of the two detectors looking for potentially interesting signals. On 14 September 2015, while Drago was on the phone with a LIGO colleague in Italy, his pipeline sent him an email alert—of which he receives about one each day—telling him that both LIGO detectors had registered an “event” (a nonroutine reading) 3 minutes earlier, at 11:50:45 a.m. local time. It was a big one. “The signal-to-noise ratio was quite high—24 as opposed to [the more typical] 10,” he says.

In fact, the signal was so strong that Drago didn’t believe it was real—and with good reason. A gravitational wave from a distance source stretches space by an infinitesimal amount, and to detect that rhythmic stretching LIGO employs two gigantic optical devices called interferometers, which essentially act as gigantic rulers. To test the incredibly complicated devices, LIGO physicists have developed mechanical systems to give them a shake and “inject” a fake signal. The signal Drago saw was so perfect it seemed too good to be true, he says. “No one was expecting something so huge, so I was assuming that it was an injection.”

Here’s where the story departs from the previously prepared script. Injections can be done in two ways: out in the open when researchers are tuning up the machines and secretly when they are taking data. Those latter “blind injections” are meant to keep researchers on their toes. Only four LIGO leaders know when such injections are made, and that information is supposed to be revealed only after a potential signal has been thoroughly scrutinized and written up for publication. That’s how things unfolded in 2010, when LIGO researchers learned at the last minute that a possible signal was in fact a blind injection. So if all had gone as anticipated, Drago might have simply noted the alert and carried on as usual, assuming the truth would come out in the end.

This time, however, the tale of the detection took a different turn. That’s because on 14 September 2015, LIGO physicists were still tuning up their machines after a 5-year, $205 million upgrade. Researchers had intended to start their first data run with the new rigs on that day, but several systems—including the injection system—were not ready to go, says David Reitze, a physicist and executive director of LIGO at the California Institute of Technology in Pasadena. So instead, LIGO leaders opted to continue a shakedown test known as an engineering run for another week. Thus, when the signal came, Drago knew that the injection system was not supposed to be working. He immediately set out to verify that and ended up alerting the entire collaboration to the signal.

Drago’s first step was to ask his fellow postdoc in Hanover, Andrew Lundgren, whether anybody in the collaboration had performed an open injection that hadn’t been properly logged. Lundgren could find no evidence of such an event. Next, Drago and Lundgren called the control rooms in Livingston, where it was just after 5:00 a.m., and Hanford, where it was just after 2:00 a.m., to see whether anybody had monkeyed with the detectors or noticed any peculiar behavior. He reached only one of the facilities—“Livingston, I think,” he says—but was told all was normal.

I’m going to call up my family and say, ‘You know, I was the first to see this.’

Marco Drago

Finally, about an hour after receiving the alert, Drago sent a broadcast email to the entire LIGO collaboration. “I asked if anybody was aware of anything that could be an injection,” Drago says. Nobody said yes. But that email alerted everyone that the instruments had seen a whopping big signal that, at least on the face of it, could not be a blind injection—exactly what they were not supposed to know. A few days later, collaboration leaders sent around a notice saying that the signal probably was not an injection. Within LIGO, the cat was out of the bag.

Once researchers realized they had a potential discovery on their hands, they ended the engineering run and switched to data-taking mode, in which the devices are supposed to operate without any adjustments. Officially, that 18 September 2015 change marked the beginning of revamped LIGO’s search for sources. But actually, scientists were trying primarily to collect enough data to validate the already observed signal, says Gabriela González, a physicist at Louisiana State University, Baton Rouge, and spokesperson for the LIGO scientific collaboration. To measure the background noise and estimate the statistical significance of the signal, researchers needed 5 days’ worth of data from both detectors running simultaneously. By 5 October 2015, they had it.

The team spent much longer than that making sure they had not somehow been fooled by a signal that had been injected either inadvertently or even as part of some elaborate prank. But such scenarios proved untenable, Reitze says. “You’d need a whole team of insiders with a wide variety of technical abilities,” he says.

As the analysis continued and the results were written up, LIGO leaders struggled to keep a lid on their news. “It was a challenge,” González says, “but from the beginning we explained [to the collaboration] the need for confidentiality” while the analysis was ongoing. “You cannot make an announcement and then say, ‘Oops! Sorry, we were wrong!’” Still, word of the discovery slowly leaked out, especially through social media and on theorists’ blogs. González says she found it stressful addressing—or ignoring—inquiries, especially from journalists. “It’s been a lot of pressure, answering to people, not answering to people,” she says. “But we never lied.”

But now that the blockbuster news is out, LIGO researchers are free to talk. And Drago says he’s glad about that: “I’m going to call up my family and say, ‘You know, I was the first to see this.’”

Follow News from Science

A 3D plot from a model of the Ebola risk faced at different West African regions over time.
dancing shoes