Gravitational waves are the last prediction of Einstein's Theory of General Relativity.

And they seem almost impossible to detect.

But have we finally done it?

What is it about Einstein?

Why is he the most famous and best-loved smart person ever?

It has a lot to do with his revolutionary Theory of General Relativity, in which he showed us that the force of gravity is an illusion.

Instead, mass warps the fabric of 4-D spacetime, leading to what we see as gravitational motion.

Check out this playlist for more details.

Now it turns out that General Relativity makes predictions far beyond the familiar gravity.

There's the deflection of light that we see as gravitational lensing.

There's the slowing of time in gravitational fields.

There's the dragging of spacetime by spinning masses.

Einstein is amazing, because every one of these predictions from his beautiful work has been physically tested and verified.

We love Einstein because he's been proven right so many times.

However, there's one last, incredible prediction that has never been directly observed-- gravitational waves.

The idea of gravity not as a force, but as warped spacetime is often depicted in analogy as a flexible rubber sheet being depressed by a heavy ball.

Now, this isn't entirely apt.

However, the analogy can give us a sense of what a gravitational wave really is.

Drop a ball onto a rubber sheet, and a dip forms, which then causes other objects to move differently along the sheet, analogous to gravity.

Move the ball around, and I create a series of ripples that flow outwards on the sheet, similar to the ripples on a pond.

Same deal with gravitational waves.

Accelerate a mass through space in the right way, and you produce gravitational ripples-- an outflowing fluctuation of expanding and contracting spacetime.

So what sort of movement produces g-waves?

Here's a technical term-- you need to change the quadrupole moment of a mass distribution.

That just means any change that isn't spherically or cylindrically symmetric.

So a rotating sphere or a cylinder doesn't make waves.

But two objects orbiting each other, or an asymmetrically spinning or exploding thing, does.

Now, just as the ripples in a rubber sheet propagate at a certain speed determined by the stiffness of the rubber, gravitational waves-- and indeed, gravity itself-- propagate according to the stiffness of spacetime-- in other words, at the speed of light.

Check out this episode for more info.

This speed limit comes from the fact that the speed of light is built into Einstein's field equation, which is necessary for it to be invariant to the Lorentz transformation.

It's worth pointing out that this speed limit is really the speed of causality-- the speed at which spacetime talks to itself.

And all massless things, including g-waves and light, must travel at that speed.

So what on earth do g-waves even look like?

Unlike ripples on a pond or even electromagnetic waves-- which are all simple, up-down, longitudinal waves-- gravitational waves are what we call quadrupole waves.

They propagate as a fluctuation of squeezed and stretched space in a sort of cross-like pattern.

If one passed through your body, you'd become taller and thinner, then shorter and fatter, then taller and thinner, et cetera, until it passed by.

How much would you be stretched?

Well, let's first think about all the sorts of things that might produce detectable gravitational waves.

The most insane gravitational phenomena in the universe-- neutron stars or black holes in-spiraling just before merger, or gravitational catastrophes like supernova explosions or collisions between giant black holes-- these make g-waves that lengthen or contract our space here on Earth by a factor of 10 to the power of minus 21 or less.

That changes your height by less than a millionth of the width of a proton.

And this change is for the most powerful waves that have likely ever passed through you.

Now, this power depends on how far away our catastrophic gravitational event is.

But these things are going to be far, because they're incredibly rare.

They happen in any given galaxy once every several thousand years.

Any g-wave that we're likely to spot is going to come from a distant galaxy, hundreds of millions of light-years away.

Spotting these is a very difficult experiment.

And so it's no wonder that gravitational waves remain the only major prediction of GR without a direct measurement.

Now, I should mention that a Nobel Prize has already gone out in 1993 for indirect detection.

Gravitational waves carry energy.

And so when, say, two massive objects orbit each other close enough to produce a lot of this gravitational radiation, their obits will lose energy and decay, causing them to spiral in towards each other.

This has been seen in binary neutron stars.

And the results agreed exactly with the rates of gravitational radiation predicted by General Relativity.

But if we could actually see g-waves, we'd be able to study black holes, neutron stars, even the extremely early universe in ways never before possible.

It would be a monumental scientific discovery.

But how do you detect a 100 billion billionth of a difference in length?

With lasers, of course-- in fact, with a giant Michelson interferometer.

This is the LIGO experiment, and it goes something like this-- shoot a laser beam.

Split it in two, and then send the twin beams at right angles down four-kilometer long vacuum tubes.

Bounce them off mirrors back and forth 400 times before bringing the beams back together.

Now, if we get the length of those paths just right, we can make the peaks of one of those electromagnetic waves line up with the valleys of the other, causing them to completely cancel out-- destructive interference.

No signal is seen, but if a gravitational wave passes by, it will shrink one of those paths and lengthen the other, and then vice versa, oscillating with time.

The returning beam won't cancel out perfectly.

And you'll get these little blips of signal.

Now, the original LIGO was able to spot changes in the length of its four-kilometer arms of around 1/1,000 of the diameter of a proton.

Nice, except for the fact that anything can cause such tiny changes in path lengths-- extremely weak seismic activity, a car driving miles away, a bird flying nearby.

Even quantum fluctuations in the photon rate causes noise.

So how do we tell that it's a gravitational wave?

Well, a g-wave leaves a very distinct signature, first contracting one arm while stretching the other, and then oscillating over time.

It's even possible to get a direction for the wave by measuring the relative path lengths.

But to be extra sure, you want to get the detection in multiple sites.

And there are two LIGO sites-- one in Washington, and one in Louisiana, plus a collaborative facility, VIRGO, in Italy.

So how many g-m waves did LIGO find?

Well, between 2002 and 2010 when it ran, it found zero-- no gravitational waves at all.

Now, this isn't necessarily so surprising.

LIGO really just scratched the minimum sensitivity needed to spot merging neutron stars and black holes in relatively nearby galaxies.

Now, based on astrophysical estimates of the number of these events, it was calculated that every year you should get somewhere between 1 and 1/10,000 of one event.

So best case scenario, we see eight events total.

Worst case, we have to wait 10,000 years to see one.

I guess it was somewhere in between, which still turned out to be zero.

After seeing nothing for a long time, LIGO shut down so it could level up into advanced LIGO.

Some pretty insane engineering upgrades make it 10 times more sensitive, which actually means it sees 1,000 times more volume of the universe-- much more chance of spotting crazy gravitational phenomena.

Given that, we might be expected to see many events per year.

So where are they?

See, even though advanced LIGO only started running a few weeks ago on September 18, some predictions tell us that we should have seen something already.

Are they holding out on us?

Does this mean that g-waves aren't even real?

Well, it's important to keep in mind that the LIGO team is extremely cautious.

Even if they spotted a wave, they'd keep it super-secret until they quadruple-checked results, which could take months.

They wouldn't announce until at least the end of the year.

They are so cautious that the team deliberately injects false signals into the system to check the verification process, and make 100% sure that the first actual g-wave detection is the real thing.

In fact, only three team members know whether a given signal is real or fake.

So every time this has happened in the past, the team has been told at the very end, sorry, just a drill.

Now, even the fake signals are meant to be secret.

But hundreds of scientists work on LIGO.

Wouldn't you expect someone to say something to a boyfriend, their mother, the postman, Twitter?

Maybe.

A rumor emerged in late September about a detection.

It made some Twitter noise, and "Nature" picked it up.

Link's in the description.

A little birdie told me that it's the signal of two black holes in-spiralling towards each other.

But is it just a drill?

Maybe.

Except another rumor is the signal was actually in the engineering data taken before the official turn-on date.

Still a fake?

Now LIGO is being definitely quiet on this, and will remain so until the end of the year.

But if this is real, then Einstein's last great prediction will have been directly verified.

And even if not, the new, advanced LIGO still has a real shot at this.

If we hear any more, we'll definitely tell you on a future episode of "Space Time."