What Happens When Black Holes Collide?

It's astonishing to me really that Einstein's theory of relativity formulated from 1905 to 1915-- Yeah.

--not knowing anything about black holes, not knowing anything about relativistic astrophysics, not knowing about what the true nature of galaxies was at that time-- that that theory, as far as we know, completely predicts what these wave forms will look like-- That's incredible.

--for the most relativistic systems ever observed, ever witnessed by human beings, and two black holes revolving around each other at over half the speed of light and smashing together.

So that's in itself stunning.

And we do these measurements to try to understand, is there something missing from our understanding of the waveform that relativity can't do for us because that could be evidence for what's beyond Einstein.

Is there a quantum theory of gravity?

Right.

Right now, we've not been able to see any problems with relativity in finding what these signals are.

Which is amazing.

Amazing and disappointing.

Hey.

I'm Dianna, and you are watching a different kind of Physics Girl episode.

This one is going to be a lot longer.

So this is an interview that I did with Mike Landry, who's the head of LIGO Hanford.

LIGO is the observatory that first discovered gravitational waves, which are one of the most amazing, intriguing phenomena.

We made a video about that last week with our visit to LIGO.

But while making that video, I grilled Dr. Landry for about an hour and a half with all of my questions about gravitational waves, and some questions that you guys asked me on Twitter as well.

And usually when I do an interview like this, I just cut the most interesting 30 seconds.

But what happens if the entire hour and 45 minutes is interesting?

I really wanted to share more of that.

So grab your tea, start your laundry, settle in to hear some of the weirdest questions about the universe for the next half hour.

So we're driving along the X arm of the detector.

Is that right?

That's right.

Yeah, the X arm with the observatory head and outreach coordinator.

This is the VIP tour.

So I have a lot of questions for you.

Excellent.

Yes.

What does it mean to stretch spacetime?

How can you conceive of stretching and squashing space and/or time?

So an analogy is good, right?

It's the medium that we live in, spacetime, that's doing the oscillating, that's getting stretched, and expanded, and contracted.

And the things within it go along for the ride.

And that's what we witness.

So a medium might be a canvas, canvas of a painting.

So if you actually had an invisible canvas and you wanted to register that the medium itself was moving-- if the medium was stretched, you could see the painting get distorted.

Is this the halfway station?

Already.

Yeah, this is the mid-station.

Are we turning around here or are we going to the end?

Oh, you can go right to the end.

Ah, Levi, we're going to the end.

We didn't get to do that last time.

This is uncharted territory.

What would be the feeling, like the overall perception of-- that you said, you could maybe hear that because your eardrum is-- Yeah, I understand your eardrum has an ability to see a pressure change of about one part 10th of a 6, one part per minute-- Oh wow, OK.

If you can survive that, I don't know-- maybe that's a question for a doctor.

Could your cells survive that?

Would it literally feel like your body is stretching apart, like parts of your body are stretching apart from each other?

I don't think people know the effects.

Of what it would feel like?

What it would feel like.

Yeah, right.

[MUSIC PLAYING] In the control room of LIGO, right?

This is where we are?

Mm-hm.

Of LIGO Hanford Observatory.

[LAUGHING] Just due credit to LIGO Livingston Observatory.

Of course.

Of LIGO Hanford Observatory.

Oh, Physics Girl tour.

Sorry, I'm very distracted.

We're on the calendar.

How do we know that the first collision was two black holes?

And then how do we know if it's two neutron stars?

What evidence do we have that it's black holes versus some other object?

Right, so, essentially, the signal tells you what it is.

It kind of cries out to you what it is.

Not if you're not a gravitational wave astronomer.

[LAUGHING] Well, so like if you just listen to the first binary black hole merger, it's low in frequency, and it's really in our detector just for a short time, a quarter second or so, so it goes "whoop" back.

And then if you listen to the binary neutron star event, it's 100 seconds long.

And it chirps up to much higher frequency.

So it's kind of like "waaaa," for 100 seconds-- "waahh!"

So the high frequency is the frequency of those objects orbiting, spinning-- Exactly, so what you're hearing is the modulations of space and time that are transduced, changed to-- spacetime changes to an electrical signal in our detector.

So ultimately, to analyze these signals, we use general relativity.

So if you solve Einstein's equations for this in-spiral phase, we make approximations on the in-spiral phase and on the out phase, but in the in between, at the merger, you're using supercomputers and numerical relativity, and stitching it all together to give you what the waveform is, and extract what are the masses, what are the spins, what's the distance to those stars, what's the orbital inclination?

All of these different parameters from relativity are teased out of the data.

Is it possible that it's a different object, that it's something else that has the mass that you're measuring?

Or are there other things besides mass that give a signature of this being a black hole or a neutron star?

Well, there's kind of two different types of answers to that question.

That's a really probing question.

So on one hand, you're just saying, like, What if it's something else, not gravitational waves at all?

And there's different ways in which you show it's not something else.

One of the things is we register about 300,000 channels per interferometer at the two sites.

So it's 600,000 channels overall.

And at a given site, 30,000 of those channels are environmental sensors, things like wind speed and tilt meters and seismometers and voltage meters and RF monitors and magnetometers, cosmic ray detectors, all in order so that you know what the environment that the detector lives in is doing.

So if there's a lightning strike in Burkina Faso, could that somehow electromagnetically couple into both detectors at the speed of light?

And so you have to know that.

And we can show that-- nope, that didn't happen.

And so on one hand, you're just eliminating alternatives.

OK. And then-- OK, so this is the question of whether-- how we know it is a gravitational wave in the first place.

In that case, we know it's not something-- we may not know if it's a gravitational wave yet, because it could be the-- OK. --detector itself that has its own noise sources.

But in those cases, you're trying to exclude the environment.

And lots of scientists asked about various types of couplings after the first detection.

How do you know?

You gotta have-- extraordinary claims need extraordinary proof-- Right.

--so you can make sure that you can exclude all of these different possibilities.

But then, secondly, you want to know, well, what if the detectors just somehow managed to glitch like this on their own?

They vibrated a bit, and maybe it's something that you can't detect with your sensors.

It's in the photo detectors.

How do you know?

So then you statistically also assess the data.

And so you basically artificially time-slide the data between the two detectors and analyze for statistical fluctuations in the detectors to eliminate that possibility.

OK. And you can pretty quickly show that any statistical fluctuation in the detectors would take more than hundreds of thousands of years of running to actually make such an event, and then we don't even make such an event.

So you're just putting a bound.

You'd have to run the detectors for hundreds of thousands of years to even have a chance to have this kind of fluctuation.

And ultimately, you use general relativity, and see that the time frequency evolution, the way the signal changes is a function of frequency over time, behaves in both detectors in the same way and as predicted.

OK. Can gravitational waves hypothetically interfere in the way that light waves can, like, get constructive in-- Superpose.

Yeah.

Yeah, they can.

Yeah, but in principle, in the far-field regime, they'll add-- they'll superpose-- Yeah.

--half linear superposition the way waves do.

Close to those black holes, that's not the case.

General relativity is this wildly non-linear theory.

So-- OK, wait, wait, wait.

--close-- Hold on.

[LAUGHTER] Close, that's not true.

OK. Go on.

[LAUGHTER] Sorry, I just needed a moment to let that sink in.

It's the case-- OK, so you're saying, like, the kind of linear superposition of waves that you typically get-- You don't get that.

--from electromagnetic waves, you don't get that in-- Oh, my goodness.

Yeah.

That's so-called far-field regime.

That's mind-blowing to me, but it also has no implications for everyday life.

No, that's right.

[LAUGHTER] That's one of those things that's, like, we know the speed of gravity.

Gravitational waves don't combine linearly close to the collision of two black holes, like, things that I'm like, what?

[MUSIC PLAYING] So what are some of the newer discoveries or, like, the physics we learned from these gravitational wave observations?

Yeah, OK, let's start there.

OK, for this neutron star collision in particular?

The very first detection was of two black hole mergers-- That's right.

--two black holes merging.

That's right.

But the paper that you wrote was about this neutron star merger?

Yeah, that was the-- --this capstone paper or something that described the entire detection pathway from gravitational waves to, ultimately, X-rays and radio waves-- OK. --for binary neutron stars.

Gotcha.

So if we start there, I mean-- it's just one event, but a huge amount of physics and astronomy was learned, and astrophysics was learned, from that one object.

OK. For instance, prior to that, people weren't quite sure what caused short gamma ray bursts, which have been seen since the '70s, and are some of the most violent events in the universe.

People see them with gamma rays.

Every couple of days, space-based detectors see them.

And short gamma ray bursts, some of them are caused by neutron star collisions.

Now, we understood that within a few minutes of the call, that rapid response team call, where people get on the phone when they get an alert on their cell phone saying, Hey, we just had a gravitational wave trigger.

Let's go look and see if it's real or if it could be something that's terrestrial.

And so one of the things that we learned or maybe knew, but at some point, the idea was supported that the speed of light, as far as we know, is the same as the speed of gravity.

Yeah, there's several different scientific results that came out of these two neutron stars, and one was the first direct measurement of the speed of gravity.

So we know when Einstein laid down his field equations for general relativity in 1915 and solved them for gravitational waves, the prediction that spills out is that they travel at the speed of light.

And so that had been indirectly measured in the past-- OK. --and limits set on whether or not gravitational waves travel at the speed of light or not.

But you had a direct race in this measurement of binary neutron stars from 2017.

So 130 million years ago, two neutron stars collided together in a galaxy called NGC 4993, discovered by Herschel in the late 1700s.

And for 130 million years, gravitational waves and light had been coming at the Earth.

And we first measured gravitational waves, and then 1.7 seconds later, NASA's Fermi space-based detector saw a burst of gamma rays.

So last time, I sort of came to this realization, like, holy cow.

If the sun disappeared-- I thought about this before, where if the sun just, like, poof, disappeared.

So it would take you about eight minutes to know-- to see that the sun had disappeared-- Absolutely.

--because that's the speed of light.

So you would get the knowledge, you'd get the information, the sun is gone, eight minutes later.

Yes.

But I didn't realize that the speed of gravity is the same, so therefore-- correct me if I'm wrong-- but therefore, the Earth would continue to orbit as if the sun was still there for eight more minutes.

Absolutely.

That's right.

Yeah.

Yeah, that's the crux of gravity having a finite speed-- Yeah.

--is that that information isn't conveyed to the Earth that you somehow vanished the sun.

Yeah.

So it would continue on an arc and then it would go on a tangent.

Yep.

Whereas in Newton's theory of gravity, where gravity is instantly conveyed, it's just a 1 over r squared, tells you nothing about the propagator and-- Yeah.

--how gravity moves around the universe.

Right.

The moment you vanish the sun, the Earth would go off at a tangent.

Yep.

Here in relativity, it continues on that arc for eight minutes and then goes off on a tangent.

Which is so cool.

And also, we had talked about, like, if you had an observer that was halfway between the sun and the Earth's orbit, then the sun disappears, you see that after four minutes.

And then another four minutes go by, then the Earth goes off.

But you still have to wait for that light to get back to you.

So it would be eight minutes, right, until you saw the Earth continue on.

Because you're always looking in the past.

Yeah, yeah, exactly.

It's astronomical scale.

Right.

So I guess you would still see-- you would see the Earth, like, continuing on, not knowing what's coming until suddenly you would get the light from the Earth.

Well, then you wouldn't be able to see it, though, most likely, because there would be no sunlight shining.

If you bounce radar off.

Right, right, right, yeah.

Yeah, of course.

We have radar at our observatory halfway between-- Exactly.

--the center of the Earth, yeah, this is great.

Is there a word for the collision of two neutron stars?

So there's the kilonova, is that what you're thinking?

Yeah.

Yeah, that's it.

Yeah, so the kilonova describes the object that forms after the collision.

Oh!

And it's thought to be the site of the formation of much of the heavy elements in the universe.

This is a kilonova.

Is a kilonova, yeah.

So does it explode in the way that it-- in an analogous way to the way a supernova explodes, or does it just collide?

Because neutron stars, aren't they the leftover cores of supernovas?

They are.

They're one of the corpses-- Right.

--of supernovas.

Right.

So you can have-- there's all these different types in supernovas.

It's pretty complicated.

Yeah.

So there's different ways in which you can get a supernova.

So a kilonova's a very different object.

Yeah.

It's these two neutron stars that were-- they're probably born in a binary pair.

Yeah.

But half the stars in the night sky are binaries.

And one will go supernova, and the other one will go supernova later, and form two neutron stars, and they revolve around each other.

And people have seen those-- Oh, OK. --in our galaxy-- Yeah.

--neutron stars locked in orbit.

There's something like 10 pairs of neutron stars known in our galaxy.

The reason why people want to study gravitational waves is because they're excellent probes of what's dark in the universe.

You can't probe the interior of a supernova collapse with light, because the light scatters on the way out, whereas gravitational waves will stream freely out of the core of a supernova, and inform you, and tell you about that.

Similarly, light from the Big Bang occurs 300,000 years after the Big Bang.

That's when the universe has cooled enough such that the cosmic microwave background can stream freely from matter.

And so there's lots of physics before that.

Right.

If you actually detect the gravitational wave analog of that, you'll sample the universe at 10 to the minus 30 seconds after the Big Bang.

Oh!

So that's why people look for gravitational waves.

The sun gets its energy from thermonuclear fusion at the core of the sun, right?

And so that's generating photons.

And those photons take about a million years to get to the limb of the sun.

So it takes a million years because the light is scattering around, because the mean free path for a photon is so small-- I knew that it was long.

--it bangs around in there.

I didn't know it was a million years.

About a million years.

And then it takes eight minutes for that light to get out.

And so that is an example of why you want to know about gravitational waves for these dense objects, dense stars.

My imagination is running wild and thinking if we had dark matter and dark energy collisions, or violent dark matter and dark energy events, that gravitational waves could potentially detect those, or we could potentially detect them with gravitational waves.

Well, I'm not saying this right.

We could potentially detect the gravitational waves that come off of that of those-- So-- --if that happened.

It depends on what those things are.

We call it dark energy and dark matter-- But we don't know what they are.

--because we don't know what it is.

Exactly.

So that's our placeholder saying, I don't know yet.

Yeah.

But it's true that at some point, people will find some gravitational wave signature, which is completely unexpected from something nobody theorized.

That's a hugely exciting thing.

And it's happened before.

When people turn on new detectors in different sectors-- not gravitational waves, but in different sectors like Jocelyn Bell and Antony Hewish putting on a radio telescope that could detect short timescale changes.

Bell showed these-- showed these were pulsars, right-- Yes.

--detected neutron stars.

So people find things unexpectedly.

Which is very exciting.

I mean, to think that we've turned on this new detector that detects a different kind-- it's a kind of technology, and we may be able to see things we never knew of, objects we've never heard of, never even theorized.

First detection was like that, right, that heavy stellar mass black holes.

So the mass of-- Oh, so that was-- --the black holes that we discovered, that was a surprise at some level.

Interesting.

[MUSIC PLAYING] You mentioned space.

You mentioned in a talk that space is very-- is not very stretchy; it's stiff.

It's more-- I think you said it's more like steel than like rubber or something.

What does that mean?

What does it mean for space to be stretchy?

We typically think of something material as being stretchy, something that we can touch and pull apart?

How stretchy is space?

Not very, it turns out.

So what does that mean, though?

How do you quantify stretchiness of spacetime?

Well-- I never thought I'd ask that question.

Why are you laughing?

MAN: It's a good question.

OK. Well, so you can just actually look at Einstein's field equations and look for what is effectively the Young's modulus, which tells you how stretchy, how expandy or compressy-- Compressy.

So don't show that.

Please don't show that.

I think that's legit-- Compressible.

I think that's something I would have said-- compressy.

Compressy.

If you look at what is the-- in general relativity is effectively the Young's modulus of spacetime, which in materials, Young's modulus tells you how compressible something is.

You could say, oh, here's Young's modulus of spacetime, and let's compare that to steel.

And that's why we need black holes, and neutron stars, stars exploding, the birth of the universe.

These are the kind of things that are actually going to create vibrations in space that will be large enough in amplitude for us to see.

When people ask, well, can you make an experiment on Earth to make the waves and detect it, the answer's no, just because these numbers come out to be really, really large.

It's certainly the kind of thing people wanted to do.

Could you make a rotor and-- say, for a hypothetical detector, even ignoring what the detector might be, could you make a bar and spin it around fast enough, a big enough bar, in order to detect gravitational waves in the lab next door, which is sort of the way electromagnetic waves were discovered.

You in one half of the lab make a spark-gap sort of emission of electromagnetic waves.

And on the other side, you have the radio receiver to detect them, and turn it on, turn it off, chop it on, chop it off.

And you can see these radio waves.

If you look at how fast you have to spin the bar that exceeds the tensile strength of steel and the bar rips apart.

The atoms literally rip apart.

So there's just no way to do that here in a lab given how stiff space is.

Right.

Yeah, so we talked about how we cannot create gravitational waves, not even with geological activity, which is violent on Earth as far as Earth terms go.

Violent but, yeah, not even close.

Many people ask, can you set off a nuclear bomb and measure the gravitational waves there?

Still a paltry amount compared to what you need to actually see them.

Good segue because-- so a nuclear bomb converts matter into energy.

So in the collision of two large objects, like two black holes, a large amount of mass is converted into energy in the form of gravitational-wave energy.

How does that work?

Basically, again, we appeal to relativity here.

So you have the waveform from your data, and you've used Einstein's theory of relativity to calculate what the waveform should be for these two black holes.

Take the first event-- 29-solar-mass black hole, 36-solar-mass black hole merging together.

And so you have the inspiral phase.

Then there's a merger phase where the things combined, and it's some object that looks like a black hole.

But it's got lobes on it, and it's spinning around very briefly.

And then finally you have the ringdown phase of that object as it smooths out and becomes a spinning black hole.

Oh, wow.

Now, what's left over?

36 and 29, that's 65 solar masses.

You can tell from the waves on the ringdown that this thing was 62 solar masses.

So in that process, three times the mass of the sun were converted into gravitational-wave energy.

Right?

So three times-- and that's what's coming out from the system is these gravitational waves, the churning up of.

So there's this conversion of those three times the mass of the sun into gravitational waves.

None of it's coming out light.

That makes it this super exotic object.

And at the time, it was the most powerful astronomical event known to human beings because of that conversion-- three times the mass of the sun in a quarter second.

That outshines all the stars in the known universe for that quarter second.

None of it's coming out in light.

It's all coming out in gravitational waves.

So Twitter has a bunch of questions, like a rapid fire, just some Twitter questions.

These are not quick questions.

What was I thinking?

The questions are like, why is mass the cause of gravity, and why is gravity only attractive?

Why is gravity attractive?

Well, one, it's an observation in the universe, but it's also not absolute.

There's at least one example that we know where gravity appears to be repulsive, and that's in dark energy.

And so what the right solution of that is, is still not well understood.

But it does seem like on the largest scales and on these time scales that there's a component of gravity which is repulsive.

So what does that mean?

It looks to us like there's some sort of negative energy density or something that's driving this expansion of the universe.

Is a repulsive form of gravity a thing that is taken into account in a theory of gravitational waves?

No, not at all.

In order to understand these waves, we were using classical general relativity, and so you don't have to fold in any knowledge of dark energy or quantum gravity-- nothing.

So it's the same theory that Einstein laid down in 1915, but people know a lot more about it since by doing more investigations, exploiting the theory, finding solutions for different scenarios, and applying numerical relativity.

So we absolutely don't need it in terms of gravitational waves.

[MUSIC PLAYING] I think it tried to ask you this last time, but I don't know how to ask this question.

And I feel like it's fundamental to what it really means to-- fundamental to how this observatory and how this detector works.

Why is it that when you stretch the distance of the arm, you're not stretching the light as well?

Or I don't know-- I don't even know-- do you know the right question that I'm trying to ask?

WOMAN: I think what you said is perfect.

Yeah, that's right, That's a good way to put it.

It's sort of like, how do you make a measurement with a ruler if your ruler's changing length as well?

Exactly.

Yes.

So the light does change.

It has to change just the same way light gets red shifted when the universe expands.

Exactly.

Same thing.

Same question I'm trying to ask.

Yes.

So, yeah, the light is affected, but it's only affected for a small time in the detector relative to the gravitational-wave signal.

So yes, it's in effect, but it's a small effect for us.

It's about a 1% per kilohertz effect, meaning our detector is sensitive to many different frequencies.

It's a broadband instrument.

So the higher you go in frequency, the more of an effect happens.

And we know it happens, and it gets extracted in the calibration.

Interesting.

Let me see if I-- I feel like I'm starting-- I maybe am coming to an understanding.

But is it that you're sending out light and you're measuring how long it takes the light to go that distance?

And then maybe a half a second later, the detector has stretched, and you're sending a new light, and seeing how long that takes to go whatever distance the detector arm is?

Something more like that rather than you've got this light in the detector the entire time the detector is being stretched?

So light builds in the arms and resonates in the arms.

And on average, a photon has about 300 bounces.

Some get out earlier.

Some take longer to get out.

It's a continuous process.

But the time that those photons spend is the time in which they could be expanded or contracted due to the gravitational wave.

That time, in general, is less than the time that the gravitational wave spends jockeying the arms around.

And so it's the combination of those two effects that matter.

If we made a measurement at 36 kilohertz, then it's possible that if you had a source right on the zenith directly above you or directly below that the effect of the gravitational waves on the arms-- these arms, which are really sensitive to differential motion, would be completely undone by the effect on the light.

And you're at a null.

You're at a null point in the instrument.

And then you wouldn't see anything despite having this perfectly sensitive, wonderfully sensitive differential measurement tool.

Why is that-- hold on.

So first of all, we've made it to the end of the detector arm.

But I was very enthralled and almost maybe cutting to a point of understanding this question I asked that we-- I got distracted.

But we're here.

We're at the end of-- so this is where one of the mirrors is?

Yeah, one of the end mirrors.

One of the end mirrors.

--two end mirrors that's in a chamber connected to the vacuum envelope, which goes all the way back.

The vacuum envelope.

That's the tube-- yeah, the entire set of tubes.

[MUSIC PLAYING] I think I'm done.

I'm not done, but I think I'm done.

I have to let Mike Landry go eventually.

This is the best day ever, maybe.

Wow.

You got a lot of videos.

That's great.

It's not that-- it's just not that often that I get to just sit and ask people questions about things about the universe that are so fascinating, and constantly learn new things, and get my mind blown about every five minutes.

That's great.

Thank you-- Thank you.

--so much.

Thank you so much.

Yeah.

Hoo.

Happy physics-ing.

Happy physics-ing.

[MUSIC PLAYING]