00:00:00
1.3 billion years ago
00:00:02
in a galaxy far, far away
00:00:05
two black holes merged
00:00:07
As they violently spiraled into each other
00:00:10
They created traveling distortions in the fabric of space-time
00:00:13
gravitational waves
00:00:15
in the last tenth of a second
00:00:16
the energy released in these waves
00:00:18
was 50 times greater
00:00:20
then the energy being released by everything else
00:00:22
in the observable universe combined
00:00:24
It's like an awe-inspiring kind of energy
00:00:27
after spreading out through the universe at the speed of light for over a billion years
00:00:32
the waves reached earth, where they stretched and squeezed space
00:00:36
such that two light beams traveling in perpendicular pipes were put slightly out of step
00:00:40
allowing humans to detect the existence of
00:00:42
gravitational waves for the first time.
00:00:51
That's a simple enough story to tell but
00:00:53
what I found out when I went to visit
00:00:54
professor Rana Adhikari at Caltech is
00:00:57
that it hides the absurdity of just what
00:01:00
was required to make that detection.
00:01:03
There's a lot of things about
00:01:04
gravitational waves which are absurd
00:01:06
*Humming simulation of gravitational waves
00:01:14
Is that.... is that it? [RA]: That's it, yeah.
00:01:16
The main problem with detecting
00:01:17
gravitational waves is that they're tiny
00:01:20
they stretched and squeezed space by
00:01:22
just one part in 10 to the 21.
00:01:24
That's the equivalent of measuring the
00:01:26
distance between here and Alpha Centauri
00:01:28
and then trying to measure variations in
00:01:31
that distance that are the width of a human hair.
00:01:34
To detect such tiny wiggles
00:01:36
you have to measure over as large
00:01:39
distance as possible, which is why the
00:01:40
arms of the interferometers are four
00:01:43
kilometers. And even with arms this long
00:01:45
gravitational waves vary the length of
00:01:47
the arms by at most 10 to the minus 18 meters
00:01:49
so the detector has to be able to
00:01:51
reliably measure distances just 1/10000
00:01:55
the width of a proton. It's the tiniest
00:01:57
measurement ever made.
00:01:59
So how is it possible to measure that
00:02:01
considering all the other sources of
00:02:02
vibrations and noise in the environment,
00:02:04
like earthquakes, traffic, and electrical storms.
00:02:07
Well for one thing the mirrors are the
00:02:09
smoothest ever created.
00:02:11
They weigh 40 kilograms or 90 pounds and
00:02:13
are suspended by silica threads just
00:02:16
twice the thickness of a hair to isolate
00:02:18
them from their environment
00:02:20
and even then the only way to be certain not to
00:02:23
be tricked by environmental noise was to
00:02:25
build two detectors far apart from each other
00:02:27
in reasonably quiet locations that
00:02:29
allows you to distinguish between local
00:02:31
noise which would appear only one side
00:02:32
and gravitational waves which would pass
00:02:34
through both sides
00:02:35
almost simultaneously
00:02:38
I'm in a building that contains a 1 to 100 scale
00:02:43
of LIGO, the gravitational wave detector.
00:02:46
The next challenge is the laser.
00:02:50
Whoa, whoa.
00:02:52
That's a lot of stuff.
00:02:54
You need a laser that can provide one, and exactly one wavelength.
00:02:58
You can imagine, if your laser
00:03:00
wavelength is changing and you're trying
00:03:01
to use interference of light waves to
00:03:03
make this measurement it's never going
00:03:05
to work because it's something like
00:03:07
trying to measure this distance but your
00:03:08
ruler stick is constantly changing back
00:03:10
and forth you can't tell how many inches this is.
00:03:13
All this equipment, at least
00:03:16
three-quarters of it, all we're trying to
00:03:17
do is make the laser more stable, and by
00:03:20
the end of the day what we've achieved
00:03:21
is something which has a stability of
00:03:24
one part in 10 to the 20. What does that mean...
00:03:28
That's a hundred billionth of a trillion.
00:03:30
That's kind of what we end up with.
00:03:32
The best lasers for this purpose have a
00:03:33
wavelength of 1064 nanometers. That's
00:03:36
infrared light. But this presents a problem.
00:03:39
How can you measure 10 to the minus 18 with 10 to the minus 6
00:03:43
wavelength of light?
00:03:44
Yes, I wish more people would ask this question.
00:03:47
It's great for this animation
00:03:48
to show such a large shift in the
00:03:50
wavelength but the reality is, it's only
00:03:52
one trillionth of a wavelength that the
00:03:54
arms are shifting in length.
00:03:56
It seems obvious that you can measure half a
00:03:59
wavelength because that will cause the
00:04:01
light to interfere with itself.
00:04:02
Yeah, but that's fully. That will go from
00:04:03
completely dark to completely bright
00:04:05
So are you looking at, like, slightly darker
00:04:08
and slightly brighter?
00:04:09
Yeah and the limit here at how good we
00:04:12
can measure this difference between dark
00:04:14
and bright has to do with the, the fact
00:04:17
that the light is discrete. It comes in
00:04:19
discrete chunks which are called photons.
00:04:21
The variation in the number of photons
00:04:23
hitting the mirrors at any instant due
00:04:24
this quantum uncertainty is proportional
00:04:26
to the square root of the total number
00:04:28
of photons. What this means is the more
00:04:31
photons you use, the smaller the
00:04:33
uncertainty gets, that's a fraction of the total.
00:04:36
This is why the laser power in the arms is one megawatt.
00:04:41
That is enough energy to power a thousand homes, in a light beam.
00:04:47
And a megawatt, you know
00:04:48
*snap* boom, they won't even rip your head off.
00:04:50
Just, vaporized be just a smoking stump.
00:04:54
Even with a perfect laser and one
00:04:55
megawatt of power, anything the light
00:04:57
hits would interfere with it,
00:04:59
even the air, so all the air in the arms
00:05:01
of the detectors had to be eliminated
00:05:03
and it took 40 days to pump down to just
00:05:06
a trillionth of atmospheric pressure and
00:05:09
the tubes were heated up to the
00:05:11
temperature of the oven to expel any residual gases.
00:05:14
They pumped out enough
00:05:15
air to fill up two and a half million
00:05:17
footballs, making it the second largest
00:05:19
vacuum in the world after the Large Hadron Collider.
00:05:22
Now here's something most people don't think about
00:05:24
which is that gravitational waves stretch
00:05:26
space-time so light traveling through
00:05:29
that space should be stretched as well.
00:05:31
If everything is stretching how do you
00:05:33
know anything is stretching?
00:05:35
How do you know anything is stretching?
00:05:35
That's the conundrum.
00:05:36
It doesn't make any sense!
00:05:39
This whole thing is bogus shut it down!
00:05:42
I would send a laser beam down this tube
00:05:45
and then wait for it to come back and
00:05:47
then i would say "well nothing happened"
00:05:49
because the space got stretched and the
00:05:51
laser wavelength got stretched
00:05:53
Its...It looks the same if you got it stretched or not stretched.
00:05:55
It doesn't make any sense
00:05:57
well it's sort of a matter of timing is
00:05:59
how it works.
00:06:00
So the amount of time it takes for light
00:06:02
to go down this tube and come back is
00:06:06
very short.
00:06:07
However the wave... the gravitational wave
00:06:09
when it comes through its doing the slow thing like
00:06:11
*low humming*
00:06:13
this noise I made, which is low. Its this
00:06:15
it's this slow stretching, its only a
00:06:17
hundred times per second. And it's true
00:06:19
when the wave comes through the light
00:06:22
which is in there it actually does get stretched.
00:06:25
And... and then that part doesn't...
00:06:27
doesn't do the measurement for us but
00:06:29
um.. now that the space is stretched that
00:06:31
laser light is like come and gone
00:06:33
it's out of the picture. We're constantly shooting
00:06:35
the laser back into the system so the
00:06:37
new fresh light now goes through there
00:06:39
and has to travel a bigger distance than the light before.
00:06:42
And so by looking at how this interference changes with time
00:06:46
and keeping the laser wavelength from
00:06:48
the laser itself fixed, we're able to do the measurement.
00:06:51
So what was needed to detect gravitational waves?
00:06:54
Well, a megawatt of laser power to
00:06:56
minimize shot noise of exactly one wavelength
00:06:59
because we're trying to measure just a trillionth of that wavelength,
00:07:01
continually inserted to replace older light
00:07:04
that's been stretched and squished,
00:07:06
in the world's second-largest vacuum chamber at just a
00:07:08
trillionth of atmospheric pressure,
00:07:10
hitting the smoothest mirrors ever created,
00:07:12
suspended by silica threads,
00:07:14
at two distant sites to eliminate noise,
00:07:16
with four kilometer long arms to
00:07:17
increase the magnitude of gravitational
00:07:19
waves to just a thousand of the width of a proton.
00:07:23
You know what we already do
00:07:25
daily in here is what I would have said
00:07:27
is impossible if you asked me about it 20 years ago.
00:07:30
One of the things that was
00:07:31
most interesting for me to learn was
00:07:33
what is limiting the sensitivity of the
00:07:35
detectors today, and it turns out it is
00:07:37
quantum mechanics and essentially you can think of it
00:07:39
like a Heisenberg uncertainty principle, we've got two
00:07:42
things and together their uncertainty
00:07:44
has to be bigger than a certain value.
00:07:46
Luckily for us we are only trying to
00:07:48
measure one thing here, we're not trying
00:07:50
to measure two things at the same time.
00:07:51
All we want to know is how much more
00:07:54
this arm get stretched from that arm
00:07:56
and that's... that's the key point which people
00:07:58
did not understand until recently.
00:08:00
The way to build these systems is such that
00:08:01
they're extremely good in measuring one
00:08:04
thing and that all of the uncertainty
00:08:06
which comes from quantum mechanics is
00:08:08
completely crammed into this other thing
00:08:10
that we don't care about.
00:08:12
I feel like we're getting down to these levels of
00:08:14
nature where it seems like nature
00:08:16
doesn't want you to go any further.
00:08:18
But, through our ingenuity we're figuring out
00:08:20
ways to engineer quantum noise, I think
00:08:24
that's such a remarkable concept
00:08:25
and I look forward to the results
00:08:27
that it's going to bring.
00:08:28
I think the next logical step is to go from
00:08:31
two signals to detecting all the black holes
00:08:34
in the universe all the time.
00:08:35
It's not like an alien civilization level of
00:08:38
technology it's just... we have to do a lot better
00:08:41
than what we're doing now but it's
00:08:42
I see it sort of within... within our grasp
