LIGO-VIRGO Detects The First Three-Detector Gravitational Wave

"Einstein's gravitational theory, which is said to be the greatest single achievement of theoretical physics, resulted in beautiful relations connecting gravitational phenomena with the geometry of space; this was an exciting idea." -Richard Feynman

For over a century after the publication of General Relativity, it was uncertain whether gravitational waves were real or not. It wasn’t until their first direct detection less than two years ago, by the LIGO scientific collaboration, that their existence was spectacularly confirmed. With the VIRGO detector in Italy coming online this year to complement the twin LIGO detectors, however, so much more became possible.

Aerial view of the Virgo gravitational-wave detector, situated at Cascina, near Pisa (Italy). Virgo is a giant Michelson laser interferometer with arms that are 3 km long, and complements the twin 4 km LIGO detectors. Image credit: Nicola Baldocchi / Virgo Collaboration.


An actual position in space could be identified for the first time, enabling a possible correlation between the gravitational wave sky and the electromagnetic one. The three-dimensional polarization of a gravitational wave could be measured, and compared with the predictions of Einstein’s theory. And gravitational wave signals can be teased out earlier and measured to smaller amplitudes than ever before. Not only have we just seen our fourth gravitational wave event, we’ve seen it in all three detectors.

This three-dimensional projection of the Milky Way galaxy onto a transparent globe shows the probable locations of the three confirmed black-hole merger events observed by the two LIGO detectors—GW150914 (dark green), GW151226 (blue), GW170104 (magenta)—and a fourth confirmed detection (GW170814, light green, lower-left) that was observed by Virgo and the LIGO detectors. Also shown (in orange) is the lower significance event, LVT151012. Image credit: LIGO/Virgo/Caltech/MIT/Leo Singer (Milky Way image: Axel Mellinger).


This discovery is, indeed, something big, but there’s even bigger science to come in the future!


More like this

"Wormholes are a gravitational phenomena. Or imaginary gravitational phenomena, as the case may be." -Jonathan Nolan Yes, we detected gravitational waves, directly, for the first time! Just days after Advanced LIGO first turned on, a signal of a 36 solar mass black hole merging with a 29 solar mass…
"Einstein's gravitational theory, which is said to be the greatest single achievement of theoretical physics, resulted in beautiful relations connecting gravitational phenomena with the geometry of space; this was an exciting idea." -Richard Feynman There’s no doubt that LIGO has given us one of…
"What's really exciting is what comes next. I think we're opening a window on the universe -- a window of gravitational wave astronomy." -Dave Reitze, executive director of LIGO It revolutionized our view of the Universe when the LIGO annoucements – and we’re up to three, now – came out. They…
"Einstein's gravitational theory, which is said to be the greatest single achievement of theoretical physics, resulted in beautiful relations connecting gravitational phenomena with the geometry of space; this was an exciting idea." -Richard Feynman When we look out into the Universe, we normally…

It really hasn't given us much. I also have not heard any recent explanation about the 'noise' problem of the initial experiment being explained and resolved without frantic hand waving. Rushing on to new plaster new 'discoveries' in the media when you haven't resolved the initial discoveries' authenticity looks like damage control.
It will be nice when they can confirm this 'detection' by some other means than their wiggling laser beams. Until then, they can confirm a million theoretical signals, but until there is actual confirmation by some other's just rinse and repeat of a made up computer generated template they have best fitted a signal to. The signal may nor may not be caused by what they claim...just like polarized dust was being attributed to something they WANTED to be the cause.

@CFT #1: It is hard not to admire someone with such an absolute commitment to permanent, wilfull ignorance, and refusal to even attempt to understand subjects with which they disagree.

When you actually have something intelligent to say, especially some true technical objection to some scientific analysis, it might be worth responding. But your comments here are nothing more than a rinse and repeat of the same mindless spew of antiscience you dump out every day.

By Michael Kelsey (not verified) on 27 Sep 2017 #permalink

Looking at the picture in the Forbes article (where all three detectors/signals are shown), Livingston signal does look like an actual signal. Hanford looks so/so, but Virgo looks just like noise. Why are they so different? On the other hand, why does a waveform look different in all three detectors if it;s the same signal?

By Sinisa Lazarek (not verified) on 27 Sep 2017 #permalink

Oh boy they keep on coming up with strange imagery to sell their work:

"This three-dimensional projection of the Milky Way galaxy onto a transparent globe shows the probable locations of the three confirmed black-hole merger events observed by the two LIGO detectors"

By Paul Dekous (not verified) on 27 Sep 2017 #permalink


Orientation, but than again for the Whitened strain they suddenly all look the same. : /

By Paul Dekous (not verified) on 27 Sep 2017 #permalink

Ah, the irony of CFT complaining about a noise problem.

By Naked Bunny wi… (not verified) on 27 Sep 2017 #permalink

Boom! ...and THAT is comment of the week.

Starts with a boom now?

By Ragtag Media (not verified) on 27 Sep 2017 #permalink

@Sinisa #3: Paul's response at #5 is part of it: in fact, the different orientations of the detectors, and the consequent different detection amplitudes, were important in this case (see the PRD article) for pinning down the polarization.

Another part is that VIRGO is smaller than either of the LIGO detectors, and so has lower sensitivity (higher noise vs. signal) for a given signal as compared to the LIGOs.

My guess (as a physicist, not a GWD expert) is that this is going to be a pattern in future detections: the VIRGO signal alone will likely always look "worse" in an S/N sense than the two LIGO sites. It would be great if a specialist could correct me about this.

By Michael Kelsey (not verified) on 27 Sep 2017 #permalink

There were earlier rumors about a short GRB being observed simultaneously. Was there any truth to this? Supposedly short GRB are NS mergers, and the BH masses shown for these event(s) are an order of magnitude too large for NS.

By Omega Centauri (not verified) on 27 Sep 2017 #permalink

@Michael Kelsey,
As far as your skills go,
You are far better at insulting than explaining.

I have little respect for people such as yourself that consider themselves above reproach, so informed, yet for some reason have next to no real ability to convey this information without contempt when they are seriously questioned. No professional or educator (of any field) who actually knows their subject responds in this manner unless they are hiding something...or just full of shit.
You , I think are just full of yourself, enough so that you think your position or title actually has value outside of your office. It really doesn't. Only your arguments and reasons do.
Try explaining more, insulting less, speaking less from the self satisfied authority you think you are, and more from the knowledge you claim to possess, Then you might be able to pass yourself off as knowledgeable instead of just being a snob with a piece of framed paper on the wall.

@Sinisa Lazarek #3,
Good question.

@Omega Centauri #10: I thought what happened was that LIGO-VIRGO had sent out "trigger announcements" to telescopes (including Fermi GLAST) that _could_ see a GRB if one happened, which is what led to the rumor/guess that they had seen a neutron star merger.

I did find a Science Daily article ( about some very detailed GRB observations, but that was published in July, and referred to a 2016 GRB. I also found an article in Ap.J. ( discussing the coincidence of GW170104 and GRB170105 (Jaunary 2017 events). But I don't find anything about the 14 August GW.

By Michael Kelsey (not verified) on 28 Sep 2017 #permalink

If you make criticisms in a disrespectful way, is it really a surprise you may get disrespectful responses though?

@Frank #13,

I'm not a cheer leader or a fanboy, I'm a skeptic. I ask questions that challenge assertions, not stroke egos. If I raise a question like "how do you know how your dust was polarized?" that is not a disrespectful question, it's an honest one which i expect an honest answer to. It's the kind of question a real scientist would ask before they wasted millions of dollars of taxpayer money on project that was pointless.
I was treated with contempt for asking it and called 'anti-science' by science groupies such as yourself who put their faith in the opinions of experts. I'm glad to say I don't share in that cargo cult belief anymore.
“Science is the organized skepticism in the reliability of expert opinion.”

― Richard Feynman

@CFT # 14:
So you cannot see any disrespectful attitude in your comment #1 for example?


I disagree that including more gravity wave detectors gives little improvement. The spatial resolution of the latest GW source is much better than the preceding sources (although I wouldn't want to try to defend "pinpointing" it), and this third detector enabled an additional type of GW measurement. It looks to me to be another corroboration of GR theory.

As for seeking an alternative line of evidence to validate or confirm the measurements made by the LIGO and VIRGO detectors, if (in future recordings) there are some associated photonic measurements (e.g. from intersecting accretion disks) perhaps then your comfort level will go up.

While the LIGO "noise" critique should be resolved, I don’t think it warranted closing down the apparatus pending resolution. Further, as additional, independent devices come on line, any particular source or form of noise in one device can be assessed by comparing its measurements to those from the other devices.(e.g. LIGO compared to VIRGO, GEO 600, and Kagra).

According to the paper the Redshift is about 10%, how can you know something like that from having measured one single pulse? For a Galaxy we know what the frequency of light is at the point of origin, but the frequency of a gravitational wave how can you know the amount of Redshift without know the origin?

By Elle H.C. (not verified) on 28 Sep 2017 #permalink

@Elle H.C. #18: I went back to LIGO's first detection paper ( where they cover these issues in more detail.

What they get from their signal is the luminosity distance: the source amplitude is a known quantity (computable directly from the masses and period). The detected amplitude is measured, and the relationship between the two tells you the distance.

Using "standard cosmology" (i.e. the Hubble relationship, valid for z < 1) you can convert the distance into equivalent redshift for the convenience of optical astronomers.

This is is exactly the same thing we do, in reverse, when we make statements about the distances of galaxies! What we measure is a redshift. From the Hubble relation, we can convert that to a distance. Since the relationship is monotonic, it's invertible.

By Michael Kelsey (not verified) on 29 Sep 2017 #permalink

@Michael Kelsey,

"What they get from their signal is the luminosity distance: the source amplitude is a known quantity (computable directly from the masses and period)"

It's all based on precalculated templates. The problem I see here is that there is no reference frame for these templates. For light you have plenty of references to calibrate your instruments, but what reference do we have for GWs?

By Elle H.C. (not verified) on 29 Sep 2017 #permalink

@Frank #16,
No it was not disrespectful. It was truthful. I didn't call anyone a name or make up something that wasn't true, I pointed out they that if they didn't have their signal/noise separation problem resolved, doing the experiment a million more times and expecting different results is pointless. They didn't succeed the first time and have not been forthcoming about it. If you consider that 'disrespectful, you have obviously never had a job performance review before (it's something people in the real world outside of academia have to do every year). People screw up. Sometimes the truth is harsh. If a sports team does badly, like say the Redskins, and they flub their passes and make obvious mistakes in their plays as a professional team shouldn't, you just say so, and people do. This is expected. You don't make excuses for them, "Aw, don't be mean! Football is hard!!! Maybe they had food poisoning, or the flu bug!" they are in the big leagues, get compensated for it, and should know their business.
When you mess up, take your lumps, and whatever you do, don't try to pretend it didn't happen, or rewrite history by spinning it as a success story. It inspires you to do better next time, or gets you fired if you can't. Either scenario is acceptable in endeavors where actual accomplishment is valued.

@Elle H.C. #20,
What you are asking is the exact point I made when LIGO first announced their discovery. They weren't looking for a an actual needle (something known, observed and understood)in a haystack, they were looking for what they imagined a heretofore never before seen needle to look and act like in a haystack. Gravity waves are purely theoretical, still, and if they do exist (for the sake of argument), they might not have the properties assumed, such as what speed they would propagate at (the assumption is c, which is a purely cherry picked choice "Propagation speed of Einstein's gravitational waves is arbitrary, because it is coordinate dependent." ). Since they have no 'real' baseline signal, just an imagined signal based on certain assumptions, and the orbiting blackholes in question were not then observed by some other means for confirmation, it was an unverified claim.
For the sake of illustration,
If a person claims to have used a new untested (fill in the blank) technology to locate something no one else can find, or has ever observed, you certainly don't just say 'ok, you're right', you go where they say the object in question is and look for confirmation. If you find what they claimed would be there, you have something. If you can't find anything, you don't have something. But you have to have that confirmation, the claim is not enough by itself.

And these are how you should comment all the time.
Keep up the good work! :-)

Regarding gravitational waves, we should bear in mind there are various meanings of the expression gravitational waves, and those detected by LIGO experiment are not the cause of gravity force. However, they are most probably related to the so called expansion of the Universe.

@Elle H.C. #20: That's not an unreasonable question, if you don't already know the answer. I highly recommend reading up on the underlying astrophysics behind the Hulse-Taylor binary pulsar (PSR 1913+16). This was the first of multiple pulsar-neutron star binary systems discovered, and it was the first indirect evidence for gravitational waves.

The system consists of a visible pulsar orbiting another neutron star (which might be a pulsar as well, but one not pointing in our direction). The beauty of such a system is that we can measure the orbital period to extremely high precision (milliseconds), and monitor it over long periods of time (decades!). When that is done, you observe that the orbital period is very slowly increasing (the orbit is contracting), with a quadratic dependence on time.

So what's the point? Well, from observations of the orbit at some particular epoch, you can derive (from Kepler's laws a.k.a. simple Newtonian gravity) the masses of the two objects and their separation (semimajor axis). Now you can take those three parameters, put them into a full general-relativistic treatment of the orbit, and it will *predict* -- with no free parameters, no fitting, no nothing -- the energy loss due to gravitational radiation. Conservation of energy turns that directly into a prediction -- again, with no free parameters for fitting -- for the rate at which the orbital period should change with time.

The result (which you can see on the Wikipedia apage) is a stunningly exact match to the observations taken over 37 years. And again, this match is NOT A FIT. There are no free parameters to get that curve to go through the data points. The masses of the system are derived independently (basically, from the detailed observations underlying any single one of the data points), and do not need to be derived using GR at all. When you put those masses into GR, you get out the energy loss curve.

The relevance of this to the gravitational wave observations should be obvious. If you have the masses of two bodies, and their separation, you can compute their orbit absolutely. With GR, you can also compute, absolutely (with NO FREE PARAMETERS) the amount of energy that system radiates away in gravitational waves, and you can also compute (with NO FREE PARAMETERS) the frequency and amplitude of that radiation.

As you might guess, these calculations do require a pretty fair amount of computing power (a nice server farm with a few thousand cores), especially if you want to repeat that calculation for a whole lot of different mass combinations and a lot of different semimajor axes. Also, in addition to the masses and orbits, the spins of the objects also come into play in determining some of the detailed shapes (harmonics, cusps, etc.) of the time-varying gravitational wave spectra. So there's an awful lot of detailed computation involved.

All that stuff I just described? Those are the "templates" which LIGO-VIRGO uses to compare with their candidate signals. The templates which the willfully ignorant CFT describes so disrespectfully as "made up", because he's too ill-informed (by intention) to understand the mathematics behind them.

There's no "fitting" in the sense of how we "fit a curve to the data" (varying free parameters and minimizing a difference function). Instead, you're trying to identify which, out of a large family of precomputed predictions, best matches an observed signal.

By doing that matching quantitatively (e.g., by computing the chi-squared or Kolomogorov-Smirnov distance between each tempate and the signal), you can quote a quantitative uncertainty on the parameter values (e.g., object masses) which were the input to each of the templates.

By Michael Kelsey (not verified) on 02 Oct 2017 #permalink

@Michael Kelsey,

Great reply, thanks!

Still some questions and doubts though … but I'll let it rest and see what the results over the coming decade will tell us when more sensors are up and running, and more data is collected and we have better insights.

IMO the Nobel prize has been given a bit too early, but it's up to the committee to take that risk and responsibility, might be a bit political motivated. With a Nobel prize it's easier to get funding and continue this project.

By Elle H.C. (not verified) on 03 Oct 2017 #permalink