How can LIGO still detect the gravitational waves?

I was watching this video by Veritasium (note: I don't have much physics knowledge). As I understand, at LIGO they detect the gravitational waves that were generated by the collision of the two black holes. How can they still measure these waves if the energy that they measure was released was only the last 10th of the seconds of the merging black holes (as I understand from the video)? As far as I understand, that would mean that there is only one peak that they can measure, which is that 10th of a second, but their experiment seems to be going on many years and they have made many measurements. How is this possible if the final collision was so short? What do they really measure then?

Edit

Basically my question comes down to: was that a "once in a lifetime chance" of measuring the waves? Have they been sitting there waiting for the exact moment and then do a measurement? It isn't something they can measure everyday?

• They are seeing multiple BH collisions. Jan 10, 2017 at 16:37
• You're correct; the goal of detecting such events is to observe a ~0.1 second long signal every few months. The real challenge is having the equipment to be able to detect any event, however. In physics terms, 0.1s is an eternity.
– user12029
Jan 10, 2017 at 16:42
• You write "their experiment seems to be going on many years". The point here is that the detectors were improved in 2015, so that only now are we sensitive to these signals which have been passing the Earth undetected. In fact, the first detection happened soon after the upgrade, on 14 September 2015.
– J-T
Jan 10, 2017 at 17:49
• and an increase in sensitivity corresponding to a increase in distance of a factor x will give them access to a volume larger by x^3. Assuming events leading to a detectable emission of gravitational waves are evenly distributed in space, this corresponds to an increase of the 'chances' of seeing gravitational events by the same factor x^3 . Jan 10, 2017 at 21:57
• "Have they been sitting there waiting for the exact moment and then do a measurement?" - no, you've got it the wrong way round. They don't have astronomers tell them when a collision will occur, then switch on their equipment and try to measure it. (The astronomers won't know it before they can watch it either). Instead, they have the measuring instrument running for months, and then they look at the data to find a pattern: Oh, this one tiny oscillation looks like there has been a black hole collision. Jan 10, 2017 at 22:56

This is the data recorded from the first black hole merger:

The figure is from this paper by the LIGO collaboration. A PDF of the paper is available here.

The detectable signal lasted around 0.1 of a second, but the black holes were orbiting each other so fast that they completed about ten orbits during that time. Basically each oscillation in the data is one orbit.

The data immediately gives the rate of decay of the orbit as the black holes merge and the amplitude with which the gravitational waves are emitted, plus lots of other information hidden away in the detail. This is easily enough to confirm that this was a black hole merger and to measure the masses of the black holes involved.

Each pair of black holes only merge once, so this was the first and last signal detected from that particular pair of black holes. However the universe is a big place and there are lots of black hole binaries in it, so we expect black hole mergers to take place regularly. LIGO has already detected three mergers. The first (shown above) on 14th September 2015, then a second possible detection (at low confidence) in October 2015 and then a third firm detection on 26th December 2015.

LIGO took a pause to upgrade its sensitivity, but is now working again. As a rough estimate we expect it to detect a merger around one a month, that is roughly once a month a black hole binary will merge somewhere in the region of the universe that lies within LIGO's detection limits.

We don't know in advance where an when a merger will occur, so it's just a matter of waiting until one happens near enough to be detected.

• So as I understand, this was really sort of a "once in a lifetime" measurement? There was only one try they could do, and if let's say there was a system failure, it was all for nothing? Jan 10, 2017 at 17:16
• @Derp: obviously we won't get any more info from the two black holes that have now merged, But there are lots of other merging black holes in the universe. Since the first detection there have since been two more black hole merger detections though one was borderline. When LIGO is running at its maximum performance we expect around one detection per month (very rough estimate). Jan 10, 2017 at 17:24
• @Derp There are more than two black holes in the universe. Jan 11, 2017 at 3:27
• @JohnRennie: I think your comment was really the answer to this question, and your answer was really a comment to this question... Jan 11, 2017 at 12:01
• @Mehrdad: I see Derp has edited the question so I have extended my answer to respond. Jan 11, 2017 at 12:29

You are right in that they are observing once in a life time (black hole's life time) event. But there are plenty of black holes, so few of such events may be detected within a year by LIGO.

The first announced observation was on September 14, 2015, while the second observation was on December 26, 2015.

• As Lawrence Krauss put it "the universe is big and old and, as a result, rare events happen all the time." Jan 10, 2017 at 23:14
• If I was able to mark two answers as accepted I would mark yours too since you answered one of my questions as well. Jan 11, 2017 at 7:01
• Well, once in the lifetime of a particular pair of black holes. The merged hole could then merge with another hole, and then another, etc. Jan 11, 2017 at 19:24

As far as I understand, that would mean that there is only one peak that they can measure, which is that 10th of a second

No, there are between 4 and 5 peaks in that 0.1 second time span between 0.3 and 0.4s. The black holes were orbiting their center of gravity several dozen times per second just before the merge. (I believe it is half the frequency of the wave in the diagrams shown by John Rennie; please correct me if I am wrong.)

• Yes, GW frequency is twice the orbital frequency.
– OTH
Nov 25, 2017 at 1:36