Written by Samantha Usman, who is currently pursuing an MPhil at Cardiff University, UK under the supervision of Prof. Stephen Fairhurst. She graduated in May 2016 with a BS in Mathematics and Physics at Syracuse University. While at Syracuse, Usman worked with Prof. Duncan Brown on improving LIGO’s sensitivity to gravitational waves from binary star systems. In her spare time, Usman trains in Brazilian jiu jitsu and Muay Thai kickboxing and enjoys walks with her Australian Shepherd, Marble.
The discovery of gravitational waves from an undergraduate’s perspective
The first time I learned LIGO might have detected a gravitational wave, I was listening in on a conference call on September 16, 2015. Two days earlier, ripples in the fabric of space from massive black holes crashing into each other at half the speed of light had passed through the Earth. The LIGO detectors picked up these faint changes in the length of space, but they pick up all sorts of extra noise that you’d never expect; how could we be sure this was really a gravitational wave?
On September 16th, I was an undergraduate starting my senior year at Syracuse University. I’d been doing LIGO research with my advisor, Prof. Duncan Brown, for almost two and a half years. Since LIGO had yet to start an observing run, my research had been focused on testing improvements to the codes that we use to search for gravitational waves. I’d been told in those two and a half years that it would take a few years to get our detectors to design sensitivity and not to expect a detection until I was well into graduate school.
So when I sat in my boss’ office listening to a colleague in Germany say he thought we’d really seen something, I rolled my eyes and muttered, “There’s no way it’s real.” I was convinced people were jumping the gun and getting excited about a chance coincidence between noise transients in the two LIGO detectors. Maybe someone had put a hardware injection, a fake signal made by shaking the detectors’ mirrors to mimic a gravitational wave. I’d heard stories of the infamous “Big Dog”: the event that’d passed all of the tests five years earlier. The event started up a flurry of activity, everyone scrambling to run checks and write the detection paper, only for it to be revealed that it was fake signal put in by the leaders of LIGO.
We went forward with the analyses. Since I had worked on the search codes for the past few years, I set up and ran the analysis using the offline pipeline I’d been working on; PyCBC. Getting the right result was important, so a colleague of mine at the Albert Einstein Institute in Hannover, Germany also ran an identical analysis for redundancy. So, make a bank of possible gravitational waves, compare it to every second of data, do statistical tests to check if they fit the wave, measure the background noise by making hundreds of thousands of years’ worth data by shifting the data in time so signals aren’t coincident…
While the analysis was running, my colleagues in the LIGO Scientific Collaboration were hard at working checking everything they could to see if something else could be picked up. Was there a lot of noise in the detectors? No. Were the fast-response searches running properly? Yes. As a generally pessimistic person, I still remained skeptical despite each passed check. I only began to believe GW150914 was a real signal when it was revealed that new code to perform hardware injections in the upgraded detectors was still being tested and was not ready to make fake signals.
My analysis finished running, I recall staying at work until 11 PM on a Thursday with my boss and two other students just to get a chance to look at the background noise as soon as possible. It looked clean, not too much noise. Another check passed. But we couldn’t open the foreground results without approval from the collaboration: we had to sit on the analysis, all complete and ready to tell us if we actually detected a gravitational wave, over an entire weekend. Finally, on Monday, impatiently sitting on another telecon with a few hundred other researchers, we agreed the background of the searches looked clean and no more changes would be made to the data or the analyses once we looked at the results. We were hoping for a signal-to-noise ratio (read: signal loudness) of at least 15. That would be enough to claim a detection, more than 99.9% probability that we saw a gravitational wave, more than 5 sigma from what would be expected if gravitational waves didn’t exist.
We opened the box: the chance of it being noise was much less than one in a million. I forget that I’m an adult and bounce around the room like a small kid to the amusement of my colleagues. Then we all head to the closest bar for a celebratory lunch and drinks.
But this was by no means the end of the work around GW150914: since we looked at the results in October, it took another few months of careful scrutiny and review, of paper writing and of tight-lipped conversations with anyone outside of the collaboration before we could announce our discovery. On February 11th, 2016, announcements were planned across the world. At Syracuse University, we took over one of the bigger auditoriums decorated with gravitational-wave themed displays and space-patterned sugar cookies.
We streamed the announcement live. I’ll admit I even got a bit teary-eyed when a giant 10-foot-tall projection of David Reitze, the LIGO lab director, proudly announced, “We did it!” I can’t even imagine what it’d be like for my mentors, Prof. Brown and Prof. Peter Saulson, who’d been working on this for decades. We switched over to our own question and answer session, where my colleagues and I answered questions about the detection. While that day and the months leading up to it will forever be in my memory, I know that this is just the first step, and the era of gravitational-wave astronomy has only just begun.
To learn more about my research on PyCBC, the program to detect gravitational waves, read our paper in CQG: The PyCBC search for gravitational waves from compact binary coalescence.
Read the full article in Classical and Quantum Gravity:
The PyCBC search for gravitational waves from compact binary coalescence
Samantha A Usman et al 2016 Class. Quantum Grav. 33 215004
Sign up to follow CQG+ by entering your email address in the ‘follow’ box at the foot of this page for instant notification of new entries.
CQG papers are selected for promotion based on the content of the referee reports. The papers you read about on CQG+ have been rated ‘high quality’ by your peers.
This work is licensed under a Creative Commons Attribution 3.0 Unported License.