Horizon (1964) s54e05 Episode Script
Aftershock - The Hunt for Gravitational Waves
RADIO STATIC NEWSREADER: Scientists in the United States say they have found the first direct evidence of what happened in the first moments of the universe.
RADIO STATIC .
.
But it's being called one of the greatest discoveries in science RADIO STATIC .
.
A faint signal from moments after the universe began When I was a young student, I was fortunate to come across a book called The First Three Minutes.
That book described the hot big bang universe up to the first three minutes or so.
That's a breathtaking leap when you think that the universe we now know is 14 billion years old.
The theories that we're testing with our present-day telescopes are much more audacious than that, of course.
The theory of inflation tries to push the frontiers back to the first trillionth of a trillionth of a trillionth of a second.
The very first instance of time.
The theory of cosmic inflation, which attempts to On March 17th 2014, astronomer John Kovac and his team held a press conference.
They'd been searching the skies for evidence of gravitational waves from the early universe, the fingerprint of creation.
If they were to find it, their discovery would answer the most fundamental question in science - how the universe was born.
This is the inside story of the greatest scientific quest of our time, and of two scientists with one dream .
.
to discover how the universe was born.
John Kovac is an astronomer working at the South Pole.
There you go.
Alan Guth is a theoretical physicist in Boston.
The fate of both scientists depends on the outcome of their seemingly impossible quest .
.
to understand what happened in the universe almost 14 billion years ago, before stars or galaxies before even the first atoms .
.
before light itself could travel through the universe.
WIND HOWLS I've made 24 trips to the South Pole over the years.
I was hooked immediately by the adventure of going to such a unique place, the bottom of the Earth, to peer back to the beginnings of time.
And so the adventure drew me in, and the science has kept me hooked ever since.
The South Pole.
3,000 metres above sea level.
One of the coldest, remotest places on Earth .
.
and one that provides a clear view out to space for the telescopes that probe the earliest moments of the universe.
Well, welcome to the South Pole.
Behind me in these crates are parts of our latest telescope.
We're going to spend the next few weeks putting it together, here at the South Pole, and then it's going to begin scanning the skies, looking at the oldest light in the universe.
The signals that we hope it will be searching for, though, are from gravitational waves that come from a period even earlier, in the first tiny fraction of a second of the universe's history.
The gravitational waves that John Kovac is looking for carry precious information about the very first moments of the universe.
But they're elusive.
No-one has ever detected them before, and the hunt for them has led John's team here to the Amundsen-Scott South Pole Station.
The air temperature, it's about minus 30 centigrade in the summertime.
You step out of the aircraft and the cold hits you like a slap in the face.
You actually have to breathe very carefully to avoid burning your lung.
I kind of think of the Pole as like summer camp for scientists.
You don't have great connections with the outside world, but everything's provided for you and all you have to do is work and, you know, in our case, build our experiment.
So I really enjoy it.
So there's a big science station with a cafeteria, with all the rooms, and a basketball court and a greenhouse.
And then about a mile from there is where our experiments are.
The team use two telescopes - the Keck Array and BICEP.
All year round, they pursue their extraordinary quest, scouring the aurora-filled skies at the bottom of the world for signs of gravitational waves from the early universe.
When John and the team first embarked on their search, nobody was even certain that these gravitational waves existed.
A colleague of mine called this a wild-goose chase.
Wild-goose chase.
A wild-goose chase.
It's better to fail at something important than to succeed at something unimportant, so I think it's with this mind-set that we collectively started doing that back in 2003.
The hunt for this cosmic wild goose pushes both the team and their telescopes to their physical limits.
They're looking for the faintest of signals from the very beginning of time.
The South Pole is a great place for us to put our microwave telescopes, because here at the South Pole, the air is the coldest on Earth.
It's incredibly dry.
There's very little that gets in the way of our microwave telescopes' observations.
It's almost like the telescopes being in space.
It's really important, because the telescopes are searching for exceedingly faint signals, signals that might arise from gravitational waves in the early universe.
Gravitational waves are one of the most mysterious phenomena in the universe.
First predicted by Albert Einstein almost exactly a century ago, they are invisible disturbances in the fabric of space and time itself.
A gravitational wave is really just what it sounds like.
It's a wave of gravity.
A ripple in space time.
Gravitational waves can arise whenever you have a rapid acceleration of mass in the universe, so a classic example is colliding black holes spiralling around each other.
If a gravitational wave much, much larger than any one we could possibly imagine were to pass right through this room, it would look like the room would get squashed and then expanded.
It would pull everything apart, squeeze it all together, pull it apart again with a certain pattern.
So it would distort the space time in the room.
Astronomers have found evidence for gravitational waves from objects within galaxies.
But the gravitational waves that the team are looking for are the oldest in the universe, and they've never been detected before.
The gravitational waves that we're searching for come from the very first moments after the big bang.
And what we're actually doing is we're using gravity as a messenger particle to take us further back than light can, back to potentially the first instance of the evolution of the universe.
When John Kovac and his team embarked on their epic challenge, there was not a shred of hard evidence to back it up.
Not a single observation from any telescope in the world.
What there was, was one man and his theory.
OK, I think we can get started now.
Good morning, everybody.
The man ultimately responsible for the entire wild-goose chase is Professor Alan Guth.
He's one of the world's most eminent cosmologists, and he came up with what scientists consider the leading theory of how the universe was born, what happened in the first fractions of a second after the big bang.
The story begins early in Alan's career, in the 1970s.
Back then, the best theory for explaining how the universe began was the so-called "standard hot big bang model" .
.
the idea that the entire observable universe emerged from a tiny, hot, dense region of space, and has been expanding and cooling ever since.
But the hot big bang model was far from perfect.
The conventional big bang theory described how the universe expanded, how it cooled, how the matter coagulated to form galaxies and structures.
Oddly, though, in spite of its name, it really said nothing about the bang itself.
I like to say that it didn't tell us what banged, why it banged or what happened before it banged.
It's interesting, because we talk about the hot big bang MODEL, and we also talk about the hot big bang EVENT, and these are actually two different things.
In fact, the big bang event, the first moment in the history of the universe, may or may not be real.
We just don't know anything about what happens at time equals zero, whereas the hot big bang model is the theory of what happens AFTER the big bang, after time equals zero.
But the hot big bang model was about to get rewritten.
In December 1979, Alan came up with a revolutionary new idea for what happened just after the big bang, in the first fractions of a second of the universe's history.
He named this theory "the theory of inflation", and it was to have a profound effect on cosmology.
These actually are copies of the notebook pages that I wrote on the night that I came up with the idea that has become inflation.
I went home one night to my rented house in Menlo Park, California, and wrote down the basic equations, and I became very excited about it.
And I even made a comment here, with a double box around it, which is not the sort of thing I did very often.
"Spectacular realization", doubly boxed.
I had only been working on cosmology for about a year or so at this time, so I was very worried that when I started showing it to other people, somebody would point out something that was obviously wrong about it, but I was excited nonetheless, and then the next day I started telling my friends about it.
The equations that Alan had written down that night in 1979 were just the beginning.
Within a few months, Alan went on a tour of universities around America, giving a series of talks to promote his bold new idea of inflation.
I first heard about inflationary theory in the spring of 1980, when Alan Guth came by to give a talk at Harvard University, introducing this new idea of the inflationary universe.
Although only a young postdoc himself, Alan was ripping up conventional ideas about how the universe began and pushing further back in time than anyone had dared to do before.
The title of the talk that I was using at that time was "Ten to the Minus 35 Seconds After the Big Bang".
This was one of the most exciting talks I had ever heard.
In the theory that Alan presented, this period of inflation was a brief burst of extraordinarily rapid accelerated expansion that occurred for a short interval of time instants after the big bang.
Inflation is basically a theory, I like to say, of the bang of the big bang.
It's a theory that describes what propelled the universe into this period of gigantic expansion that we call the big bang.
A plausible number, for example, for the starting time of inflation might be something like ten to the minus 37 seconds after the instant of creation.
So, something like a trillionth trillionth trillionth of a second after the big bang.
An incredibly short length of time.
It took me a long time to convince myself that it made any sense to talk about these things.
What Alan was proposing was mind-boggling - that in the very first tiny fractions of a second, the universe went through a growth spurt on a cosmic scale.
Once it starts, the universe would double in size, over and over again, every trillionth trillionth trillionth of a second.
A typical size for the universe at the beginning.
Something like ten to the minus 24cm across.
Decimal point, 23 zeros, one centimetre.
Unbelievably small.
This is more than a billion times smaller than the size of a single proton.
By the end of the inflationary period, the final size of the universe will be something in the order of maybe one centimetre.
It's still small at that point, but this is still the early universe.
From then on, it goes from the 1cm stage to the universe we observe today.
Alan's theory was a revolutionary new take on the idea that the universe began with a bang.
The original hot big bang model had a number of flaws, not least of which was that what happened in the very first fractions of a second was a complete mystery.
Alan's theory neatly explained the first few moments of the universe, with a new ingredient called inflation, and it was inflation that caused the universe to expand astonishingly fast before suddenly slowing down to a more pedestrian expansion.
It was the closest that a theory had ever got to the actual moment of creation.
Alan had fundamentally rewritten the story of how the universe was born.
I can't imagine what it would have been like to invent inflation myself, because it's just one of those, you know, small number of incredibly influential ideas, that it really came together very, very quickly in the mind of one person.
The theory of inflation is a little bit crazy, but it really does a very natural, nice job at explaining some of the most puzzling features of our universe.
Alan showed in his lecture how inflation, this incredible burst of expansion after the big bang, could very neatly explain some of the great unsolved mysteries of the universe.
Why IS the universe so large, with a geometry that appears to be almost perfectly flat? And why, at the largest scale, is the universe so incredibly uniform? Alan's theory would also predict the existence of gravitational waves.
It was a triumph of science.
But then came the sting in the tail.
Alan's lecture about inflation theory built to what I thought was going to be the climax, and I thought, "This is fantastic.
" And then, Alan, in the last few minutes, explained how this idea fails.
During the period when I was going around the country talking about inflation, I was aware that there was a piece of it that I didn't understand yet, namely the piece about how exactly inflation ends.
Alan's theory of inflation had possibly solved some of the biggest mysteries of the universe, but it had spawned its own conundrum.
How exactly did inflation stop? This became known as the "graceful exit problem", and if it couldn't be solved, then inflation theory was dead in the water.
So I imagined that there might be some way of salvaging it, but I did not have any good ideas about how.
It fell to two other scientists to lead the rescue attempt.
Looking back, I think that this was the reason why I had an ulcer at that time, because it was like emotional disturbance which squeezes you like that, because you have this possibility to explain the origin of the universe, and then demonstrate that impossibility to do so.
It's just so painful.
I thought, "Well, this can't be right.
"There has to be a way to fix this idea.
" The problem was, once it had started, how did inflation come to an end? It couldn't just stop.
There had to be some kind of transition between the universe that was inflating and one that wasn't.
Alan's idea of inflation was to imagine that coming out of the big bang, in addition to the ordinary matter and radiation we know of, there was an additional form of energy that had the property that it was gravitationally self-repulsive.
Not only does it push itself apart, but it causes the expansion of the universe to speed up at an accelerated rate, which is what he wanted to have.
But you don't want to continue for ever, otherwise even today the universe would be expanding at this incredible rate, so you had to have some way that this energy would decay.
His original idea is that it would decay by the formation of bubbles.
These bubbles start growing, and when they grow, the walls of the bubbles, they move and collide with each other, and when they collide, it's like explosive process.
But there was trouble with the bubbles.
The idea just didn't work.
The problem is that the very inflation that was causing the smoothing and flattening of the universe blocked these bubbles from ever being able to come together, because the space between them was stretching faster than the bubbles were expanding.
Andrei Linde and, separately, Paul Steinhardt tried to rethink Alan's idea of how to end inflation.
The question was - if not bubbles, then what? The transition that I was talking about is like boiling water.
But there are other kinds of transformations that we observe in the laboratory.
There's a particular kind of transition that transformed in a smooth, uniform way, more akin to the way that if you were forming Jell-o.
The congealing of Jell-o.
And you began the liquid state, and then you cooled it down, it would uniformly solidify.
A gradual phased transition in which uniformly and largely within a space, the system goes from one phase to the other.
It was another spectacular realisation.
If the transition from an inflating universe to a non-inflating universe occurred less like boiling water and more like congealing jelly, then inflation would end gracefully.
The graceful exit problem had been solved.
At the time, it looked like really good news for inflationary theory.
Whether you finish the job or somebody else finish the job, this is the beauty of science.
When I received a preprint from Andrei Linde, I became very, very excited.
I remember running across the hall and talking to the person across the hall about how excited I was about this.
He probably thought I was crazy.
Alan's inflation theory had been rescued from the brink.
Science now had a plausible explanation for how the universe was born.
In the decades since Alan came up with his theory, new variations of inflation have been proposed, but the essence of the idea has remained the same.
And, crucially, the first real evidence for inflation has begun to emerge.
MISSION CONTROL: 'Lift-off for Delta II' Launched into space, a series of satellites - COBE, WMAP and most recently Planck - have carried out ever more precise measurements of the afterglow of the big bang and revealed that it has features precisely as predicted by inflation.
But there's one thing above all that would put the theory beyond question.
The discovery of its greatest prediction.
Gravitational waves from the beginning of time.
I think inflation is certainly the very leading theory for what happened in those very first moments in the history of the universe, but it's still not settled.
My view about inflation has slowly transformed from one of being proponent to one of being sceptic and looking for alternatives.
Before we just accept this is what happened, it's really, really, really critical that we see the experimental evidence.
That's why we have to go out with our experiments and actually measure something that tells us whether inflation is correct or is just the wrong theory.
OVER RADIO: 'Runway three, zero left' The detection of gravitational waves is often described as the smoking-gun signature of inflation.
It proves that inflation really is responsible for kicking off the start of our universe.
If gravitational waves are detected, that provides very strong evidence to this whole story of inflation.
We're talking about how the universe was born, and it's quite humbling I think, that Or surprising, that we can say anything about that time.
Instead of exploring inflation with our minds, to build a machine that can probe inflation, I mean, I think that's quite fantastic.
The theory of inflation naturally predicts that there will be gravitational waves, but it doesn't tell you how much, how big they'll be.
It just says that, you know, they'll exist.
So when people talk about gravitational waves from the big bang, you might be tempted to think that this is some huge shock wave propagating through the universe.
This isn't true at all.
The secret to gravitational waves from inflation is really quantum mechanics.
It's simply the fact that on very small wavelengths, quantum theory tells us that everything fluctuates, including the gravitational field.
Quantum mechanics says you can't pin down a physical system to something that is in an absolutely precisely defined state.
There will always be some uncertainty, there will always be some quantum mechanical jiggle in the universe.
So what that means is that the gravitational field during inflation is not precisely smooth.
It has these little jiggles.
ALAN GUTH: What inflation does is it stretches out these very small fluctuations to make the wavelength large enough that they can be observed in the early universe.
It's a deceptively simple idea.
Quantum jiggles in gravity were stretched out by inflation to become gravitational waves, ripples in the fabric of space and time.
And if inflation really did produce these gravitational waves, then it should be possible to detect evidence of them using telescopes.
So the way we are searching for this gravitational wave signal is to study the microwave background.
It's the oldest light in the universe and it's a treasure trove of information, but it really gives us a snapshot of the universe as it looked 300,000 years or so after the big bang.
The cosmic microwave background is essentially the afterglow of the big bang, released as the universe cooled down and, for the first time, light could travel across the cosmos.
Theory predicts that gravitational waves - if they exist - would have affected the orientation of the light waves, what's called "polarisation".
So the cosmic microwave background has a polarisation, and polarisation is kind of like a directional thing, so think of a pattern of little headless arrows over the sky.
Gravitational waves stretching and compressing would leave an effect on the polarisation of the cosmic microwave background.
It would produce a particular swirling pattern in that polarisation that we call a B-mode polarisation.
It's kind of a pinwheel pattern of polarisation, or a twisty pattern.
A curly pattern on the sky.
It's the unique signature of gravitational waves.
Gravitational waves stretching and compressing space would have left this pattern imprinted in the cosmic microwave background, the afterglow of the big bang.
If that B-mode signature is there, it is an incredibly powerful messenger coming to us from the first tiny fractions of a second of the universe's history.
The hunt for gravitational waves from inflation had become a hunt for this B-mode.
If inflation theory was correct, then this B-mode was out there, waiting to be found.
So, behind me, you see one of our telescopes.
It's actually a very simple design.
It consists of a number of small two-lens refracting telescopes, each of which has an entrance aperture of only about this big, 30cm.
It's scanning back and forth on a relatively small patch of sky, relentlessly collecting microwave photons, and it needs to do that because the B-mode polarisation signals that it's looking for are exceedingly faint.
John and his team have chosen a particular patch of sky to search for evidence of gravitational waves.
It's known as the Southern Hole.
So as far as we can tell, the cosmic microwave background looks much the same anywhere that we can observe it on the sky, and so the best place to observe it, to look for very faint signals, is where our own galaxy and the emission from our galaxy is the faintest.
So we pick a patch of sky called the Southern Hole.
It's about 1,000 square degrees.
The moon is about a quarter square degree, so it'sit's fairly large.
It's about 2% of the whole sky.
It's visible from the South Pole 24 hours a day, 365 days a year, and so our telescope can train itself on this small patch of sky and scan back and forth collecting data almost non-stop, almost as if it were in space and able to observe the sky unobstructed.
Doing astronomy at the bottom of the Earth brings with it its own unique challenges.
The team work on their telescopes during the few months of the Antarctic summer.
But it is in the Antarctic winter when most of the observations are done.
South Pole Station is only accessible for about three months out of each year.
The temperatures are only warm enough to fly planes in and out for that period of time.
So, when we take one of these telescopes, like BICEP1 or BICEP2, to the South Pole, we come in with a team and we work furiously for three months to try to get everything to work, to put it all together, to calibrate it, to tune it up, to get it in pristine condition and then all of us get on an airplane and leave - except for one guy.
He watches the plane go and knows there isn't going to be another one for about nine months, and during those nine months he'll watch the sun get lower and lower on the horizon and then disappear, and then six months of darkness.
WIND HOWLS And during that six-month night, the temperatures can get down to minus 80C.
The skies can be lit up with the most beautiful southern lights, aurora.
You get an amazingly unique experience when you spend a winter at South Pole.
It's akin to being in space.
In January 2006, the team's first telescope, BICEP1, began its observations.
Other telescopes, like the Keck Array, would follow.
For three years, BICEP1 scanned the Southern Hole, trying to detect the faint signal left by gravitational waves.
But at the end of the three years, the team had found nothing.
There was no glimpse yet of any B-modes.
The wild-goose chase may have been just that.
We knew that searching for gravitational waves would be a challenge far greater than any that we had taken on before.
BICEP1 was ultimately limited by the number of detectors that it had.
It was fewer than 100 detectors.
It just wasn't enough to make maps sensitive enough to tease out very faint B-mode signals.
In November 2009, John and the team were back at the South Pole to install a new telescope - BICEP2.
Even before the three-year observation of BICEP1 ends, we already knew we want something more sensitive.
You can't just go straight in and build the ultimate experiment.
The way that you get higher sensitivity is by learning from the previous experiment and developing the technology using the previous experiment.
So that's why we've had this succession of increasingly sensitive telescopes.
In BICEP1, we had nearly 50 detector pairs.
In BICEP2, we had about 250, and each one is slightly more sensitive than in BICEP1.
So in the end we get almost a factor of ten improvement in sensitivity.
So this is the detector technology in BICEP2, and what you see here is effectively a printed camera.
Each of these pixels is basically a camera, in the sense that it's not just the detector, it's actually the lens and the filter as well.
Like its predecessor, the BICEP2 telescope was in operation across three Antarctic winters, continuing the hunt for gravitational waves.
It was a quantum leap in sensitivity.
We were able to map the sky ten times faster with BICEP2 than we could with BICEP1, and achieve much more sensitive maps of the polarisation in the same patch of sky that we had observed before, but much deeper now.
This time, a tantalising signal began to emerge.
So we started seeing something interesting in, I think, I would say, December 2012.
There were hints of signal, I think, you know, before BICEP2 had stopped running.
Different people on our team have different memories for when they first started to suspect there was a signal in the BICEP2 data.
We analysed the data as it came in through 2010 and through 2011, the first two seasons of operation, but we quickly ran into a problem.
We did all these tests with BICEP2 and we couldn't get rid of this signal.
The signal that had been picked up by the BICEP2 telescope displayed all of the characteristics of gravitational waves.
It had the distinctive swirling pattern, the B-mode that John and the team had been looking for.
Yet it was less faint than they'd been expecting.
Something did not seem quite right.
It was at a much higher level than we were expecting, either from emission from our own galaxy or really from what we thought were favoured models of inflationary gravitational waves.
It was higher than either of those and so we thought, "Ah, well, this signal can't be real.
"It must be a problem with our instrument, "there must be some kind of subtle effect that we haven't yet "controlled in the experiment that's producing a false B-mode signal.
" Yet despite the team's initial doubts, tests confirmed that the B-mode signal was coming from the sky.
The signal was real.
Once you've established that there's a signal, then the next question is, what is the signal? You can't just assume right off the bat that you're looking at the signature of primordial gravitational waves.
It would be nice if you could, but unfortunately, although most of space is remarkably empty and therefore we can make a lot of measurements of the microwave background, we live in a galaxy, and within our own galaxy, there are a number of sources that can create a polarisation that have B-mode patterns.
So the two ones that we worry about the most are something called "synchrotron radiation" and something called "dust emission".
a type of light that is produced in the galaxy when its magnetic field sends tiny electrically charged particles whizzing around very fast in spirals.
This can mimic the B-mode pattern of gravitational waves, but using data from a satellite, the BICEP team were able to show that this effect was too small to produce the signal that they had detected.
The other big potential contaminant is dust, and this dust isn't so different from the dust in your living room that you see when the sun pours through the windows.
It's made up of carbon and silicates, just like little rock, bits of rock coated in water ice, and these little dust grains can line up in the magnetic fields, and then when light shines through them, they can create a little bit of polarisation.
Dust was harder for the team to discount.
There was less data available, but the models that did exist suggested that dust seemed unlikely to produce such a large signal.
Dust was also ruled out.
Once you're really convinced that you're seeing signal that comes from the early universe, then you can say you've detected primordial gravitational waves.
And the team now felt that this was the most likely conclusion.
After four years of analysis, everything pointed to the signal coming from gravitational waves.
So, as the final tests of the BICEP2 data set were completed by our analysis team, we called a collaboration-wide meeting and about half of us were calling in from the South Pole, and half of us from locations around North America.
We had looked at the data set in every way that we knew how to, scrutinised it from every different angle and convinced ourselves that there were no other tests that we could perform.
We were feeling at that stage that it was our obligation to share our results with the community.
In fact, that it was overdue that we do so, because we had had that data for so long at that point.
Until this moment, the team had kept their discovery under wraps, but now, finally, they felt compelled to go public with their news.
We knew that it was time to show the B-Mode map that BICEP2 had made to the world and not keep it a secret any longer.
In December 2013, John Kovac returned from the South Pole.
As the team prepared to announce their discovery, there was one man above all with whom John wanted to share the news, the man whose theory had predicted the existence of gravitational waves, the wild goose that John and the team now had evidence for.
John initially sent me an e-mail saying that he would like to talk to me urgently about an issue "that's very important "to your research and mine", in quotation marks.
A meeting was hastily arranged at MIT in Boston.
I got in a taxi cab and drove over here on the evening of Monday March 10th of this year, 2014, walked down the corridor behind me and carried the draft of the paper that we had been preparing.
We arranged for him to come through the back door of the Center for Theoretical Physics, so he would be unlikely to be noticed by the other people in the centre.
I was greeted by Alan very cordially, very calmly.
In fact, we both sat down and he knew immediately what this would be about.
This was the news that Alan had been waiting for more than 30 years to hear - definitive proof that his inflation theory was right.
Well, he told me that his group had been looking for years to examine the possibility of gravitational radiation from the early universe.
He told me that he was initially sceptical that it could be found.
He told me that once they had a signal, he was very anxious to make sure that the signal passed all possible tests, to be sure that it was real, and that gradually he and the rest of the group became convinced that the signal was real, and that now they were ready to make a public announcement about that.
John's team had potentially made a Nobel Prize-winning discovery and Alan's theory had perhaps finally been vindicated.
I think we were both very excited.
My reaction at the time was amazement.
I was astounded to suddenly have a group come forward and say that they had a measurement with unbelievably high statistical significance.
It really was a shock to me and of course a very pleasant shock because it would be very strong evidence for inflation, if it was real.
It happened in less than a trillionth of a second after the big bang.
FRENCH NEWS REPOR They detected gravitational waves or ripples in what they believe is the oldest light in the sky.
The news made headlines around the world but had actually started off rather low key.
On March 17th 2014, the team had held a press conference in Harvard to announce their discovery.
So when we announced BICEP2's B-Mode findings, we knew that that would generate some news, but we were not expecting anything like the attention that the release got, or the excitement that it produced.
Honestly, I expected the reaction to be fairly quiet.
The title of Detection Of B-Mode Polarisation In Cosmic Microwave Background, with that title, you wouldn't think it would generate much interest, but it did.
I first got wind of the BICEP2 signal about a week beforehand.
I can tell you exactly when I first heard about the BICEP2 signal because it was on Facebook.
The BICEP2 announcement was incredibly exciting.
This would be one of the final confirmations not only of inflationary theory, but also with general theory of relativity.
It appeared the smoking gun of inflation theory had been found.
'This is BBC Radio 4.
'Scientists in the United States say they've found 'the first direct evidence of what happened in the first moments 'of the universe, having detected gravitational waves 'or ripple patterns, in the oldest light in the sky.
'It's being called one of the greatest discoveries in science, 'and astronomers say what they saw confirms that what Alan Guth 'theorised in 1979 looks right.
'Other experiments are hot on the heels of the announcement 'so it won't be long before scientists find out 'whether their expanding model of the early universe 'is just a lot of hot air.
'Jeff Broomfield, MPR News.
' So after the initial announcement and the news, I think most people in the community, in the broader cosmology and physics community, accepted it at face value.
But as happens in every kind of claim discovery or scientific claim, then people begin to look more closely and examine whether or not that claim is really justified.
So since the March announcement of the BICEP2 detection of gravitational waves, there's been a flurry of activity trying to determine whether or not the source of the detected B-Modes is actually the gravitational waves from the early universe, or if it could be something more mundane, like, in particular, people are wondering if it could be dust in our own galaxy.
The possibility that dust in the Milky Way might produce a B-Mode signal had been considered by John and the BICEP team.
They had chosen the patch of sky, the Southern Hole, precisely because it was relatively clear of galactic dust.
Yet the team's work came under increasing scrutiny.
When we announced this result, the most important fact for us was that it was real.
It wasn't produced by the instrument, and that was the hard part, as far as we were concerned.
That's what we had spent the last 14 years doing.
At that time, based on the information that was available, these uncertainties were rather hard to quantify.
It looked very much like a large fraction, you know, 90% of the signal, was gravitational waves.
The uncertainty over the real identity of the BICEP2 signal was to grow, as in Europe, a separate team of scientists began to release new data about galactic dust.
The world's best data on polarised dust is available from the Planck satellite.
So what they knew about dust emission in our particular field in March, you would have to ask them.
Yes, we had already sufficient information to be quite sure that the optimistic modelling of BICEP was not proper.
The Planck satellite was a state-of-the-art telescope that, for several years, had been surveying the skies from its vantage point in space.
Its principal mission had been to make the most detailed map ever of the cosmic microwave background, the afterglow of the big bang.
But at the same time, Planck had also carried out the best measurements ever made of galactic dust.
The Planck satellite has started to release results on polarised emission from our own galaxy, from data that they've taken at much higher frequencies, where that polarised emission from dust is quite a bit brighter, and as they've mapped this out in detail across the whole sky, we've seen, as they've released their results, that that polarised emission from dust is actually brighter than the typical models indicated before the Planck data came in.
The new data from Planck showed that the models of dust that the BICEP team had relied on were wrong.
The dust was brighter than expected.
It brought the entire discovery of gravitational waves into question.
In our review, without having access to BICEP data, it was consistent with between 50% and 100% of the BICEP signal being from dust.
In other words, that it was possible that maybe half of it would be dust or maybe 100% would be dust.
That's what led us to sort of agree to collaborate with the BICEP2 team.
There was a great confluence of interest from the Planck team and from our own team, a realisation that there was a critical, exciting scientific question at stake here, and the best way to answer it was to use all the best data in the world, altogether, in one analysis.
So, in a sense, we set up a memory and an understanding between the two teams and went on with it.
In July 2014, the two teams began sharing their data, using the measurements of galactic dust emission from the Planck satellite to find out where in the universe the BICEP2 signal might have come from.
Precisely how much of the signal was actually from gravitational waves? Let's make an analogy.
I'm listening to some headphones.
Imagine that there are two tunes that are playing and I'm trying to hear one tune, which is the gravitational wave signal that we're after, but there's also another tune that's playing that's potentially a bit louder, right, and that's the galactic dust emission.
And so whether or not I can hear the tune that I'm after versus the tune that I'm not interested in depends on whether I can push down that distracting tune, the one I don't want to hear.
So, by using the data from Planck, the higher-frequency data from Planck, we can, actually, essentially, make that distracting tune quieter and improve our ability to listen for that gravitational wave signal.
We're at a fork in the road.
If this is positive, this is just amazing.
I mean, you know, it means, well, we have discovered primordial gravitational waves and we have actually discovered them jointly.
On the other hand, if it was a false hope, well, let's know it.
Whether the B-Mode patterns that we've seen so far do carry evidence of gravitational waves from the first instance of time, or whether, so far, all we've seen are swirls in the sky that come from galactic dust.
And on January 30th 2015, the two teams released the results of the joint analysis.
What BICEP has done is a tour de force.
It's a magnificent measurement.
It will be tantalising, exciting.
But we confirmed that this is not a discovery.
So the results of the joint analysis of BICEP2 with Planck is that the most likely answer is that 75% of the signal that we're seeing is due to galactic dust, but - and this is a very important but - there's a big uncertainty on that 75%, right? That uncertainty spans the range of half to 100%, right? I have a piece of coal and I'm not able to tell you how much gold is in there.
Did I detect something? Do you think you are rich? You just don't know.
So, I mean, I'm just telling you that what we have established is that there is no more than sort of half of the initial claim in gravitational waves, and possibly none at all.
The analysis had shown that the signal was most likely three-quarters dust and a quarter from gravitational waves but, crucially, the possibility that the signal was entirely from dust could not be ruled out.
The claim to have discovered gravitational waves could no longer be made.
So, at this point in time, the proper scientific statement is that there's no evidence of gravitational wave, B-Modes, and that if they exist, they're less than about half of the signal that we're seeing.
It's a disappointing result for the BICEP team.
Yet this is how science works, and it means that the wild goose of inflation is still waiting to be found.
The thrill of the chase is back on.
This wild-goose chase searching for B-Mode polarisation from inflationary gravitational waves is still ongoing.
We're certainly not at the end of this story yet.
Around the world, teams of scientists are embarked on the hunt for gravitational waves, from the Atacama Desert of northern Chile to the wilderness of Antarctica.
The quest to confirm how the universe was born, to find the fingerprint of creation, continues.
Science doesn't deal in hope, but, you know, we're human.
Humans do science and if we didn't have some some hope, then we wouldn't continue to do the work that we do.
One has to put aside one's personal bias here.
We have to look and see what the data tell us and go with whatever the answer is.
That's the cruel reality of being experimentalist, and that's the point of it.
Sometimes convergence to a scientific result doesn't always follow a straight line.
We were, like, here in March, you know, we were kind of bouncing back and forth, but we're just trying to make some solid measurements and I'm confident it'll converge.
The theory that Alan Guth came up with more than 35 years ago still awaits its final confirmation, but one early champion of the theory is now its harshest critic.
So where we stand today with inflationary theory is that what seemed like a very sweet idea at the beginning has some very serious flaws.
Inflation is in exactly the same state that it was in before BICEP2 came along - that is to say it's still the best theory we have.
Discovery of gravitational waves through inflation, it's a very big smoking cannon but we have quite a lot of other smoking guns which we already have discovered.
As far as I'm concerned, inflation is still in very strong shape but whether or not gravitational waves will be added to our list of pieces of evidence for inflation, I don't really know at this point.
When I was a young student, I was fortunate to come across a book called The First Three Minutes.
That book described the hot big bang universe up to the first three minutes or so.
The theories that we're testing with our present-day telescopes are much more audacious than that, of course.
The theory of inflation tries to push the frontiers back to the first trillionth of a trillionth of a trillionth of a second.
The very first instance of time.
RADIO STATIC .
.
But it's being called one of the greatest discoveries in science RADIO STATIC .
.
A faint signal from moments after the universe began When I was a young student, I was fortunate to come across a book called The First Three Minutes.
That book described the hot big bang universe up to the first three minutes or so.
That's a breathtaking leap when you think that the universe we now know is 14 billion years old.
The theories that we're testing with our present-day telescopes are much more audacious than that, of course.
The theory of inflation tries to push the frontiers back to the first trillionth of a trillionth of a trillionth of a second.
The very first instance of time.
The theory of cosmic inflation, which attempts to On March 17th 2014, astronomer John Kovac and his team held a press conference.
They'd been searching the skies for evidence of gravitational waves from the early universe, the fingerprint of creation.
If they were to find it, their discovery would answer the most fundamental question in science - how the universe was born.
This is the inside story of the greatest scientific quest of our time, and of two scientists with one dream .
.
to discover how the universe was born.
John Kovac is an astronomer working at the South Pole.
There you go.
Alan Guth is a theoretical physicist in Boston.
The fate of both scientists depends on the outcome of their seemingly impossible quest .
.
to understand what happened in the universe almost 14 billion years ago, before stars or galaxies before even the first atoms .
.
before light itself could travel through the universe.
WIND HOWLS I've made 24 trips to the South Pole over the years.
I was hooked immediately by the adventure of going to such a unique place, the bottom of the Earth, to peer back to the beginnings of time.
And so the adventure drew me in, and the science has kept me hooked ever since.
The South Pole.
3,000 metres above sea level.
One of the coldest, remotest places on Earth .
.
and one that provides a clear view out to space for the telescopes that probe the earliest moments of the universe.
Well, welcome to the South Pole.
Behind me in these crates are parts of our latest telescope.
We're going to spend the next few weeks putting it together, here at the South Pole, and then it's going to begin scanning the skies, looking at the oldest light in the universe.
The signals that we hope it will be searching for, though, are from gravitational waves that come from a period even earlier, in the first tiny fraction of a second of the universe's history.
The gravitational waves that John Kovac is looking for carry precious information about the very first moments of the universe.
But they're elusive.
No-one has ever detected them before, and the hunt for them has led John's team here to the Amundsen-Scott South Pole Station.
The air temperature, it's about minus 30 centigrade in the summertime.
You step out of the aircraft and the cold hits you like a slap in the face.
You actually have to breathe very carefully to avoid burning your lung.
I kind of think of the Pole as like summer camp for scientists.
You don't have great connections with the outside world, but everything's provided for you and all you have to do is work and, you know, in our case, build our experiment.
So I really enjoy it.
So there's a big science station with a cafeteria, with all the rooms, and a basketball court and a greenhouse.
And then about a mile from there is where our experiments are.
The team use two telescopes - the Keck Array and BICEP.
All year round, they pursue their extraordinary quest, scouring the aurora-filled skies at the bottom of the world for signs of gravitational waves from the early universe.
When John and the team first embarked on their search, nobody was even certain that these gravitational waves existed.
A colleague of mine called this a wild-goose chase.
Wild-goose chase.
A wild-goose chase.
It's better to fail at something important than to succeed at something unimportant, so I think it's with this mind-set that we collectively started doing that back in 2003.
The hunt for this cosmic wild goose pushes both the team and their telescopes to their physical limits.
They're looking for the faintest of signals from the very beginning of time.
The South Pole is a great place for us to put our microwave telescopes, because here at the South Pole, the air is the coldest on Earth.
It's incredibly dry.
There's very little that gets in the way of our microwave telescopes' observations.
It's almost like the telescopes being in space.
It's really important, because the telescopes are searching for exceedingly faint signals, signals that might arise from gravitational waves in the early universe.
Gravitational waves are one of the most mysterious phenomena in the universe.
First predicted by Albert Einstein almost exactly a century ago, they are invisible disturbances in the fabric of space and time itself.
A gravitational wave is really just what it sounds like.
It's a wave of gravity.
A ripple in space time.
Gravitational waves can arise whenever you have a rapid acceleration of mass in the universe, so a classic example is colliding black holes spiralling around each other.
If a gravitational wave much, much larger than any one we could possibly imagine were to pass right through this room, it would look like the room would get squashed and then expanded.
It would pull everything apart, squeeze it all together, pull it apart again with a certain pattern.
So it would distort the space time in the room.
Astronomers have found evidence for gravitational waves from objects within galaxies.
But the gravitational waves that the team are looking for are the oldest in the universe, and they've never been detected before.
The gravitational waves that we're searching for come from the very first moments after the big bang.
And what we're actually doing is we're using gravity as a messenger particle to take us further back than light can, back to potentially the first instance of the evolution of the universe.
When John Kovac and his team embarked on their epic challenge, there was not a shred of hard evidence to back it up.
Not a single observation from any telescope in the world.
What there was, was one man and his theory.
OK, I think we can get started now.
Good morning, everybody.
The man ultimately responsible for the entire wild-goose chase is Professor Alan Guth.
He's one of the world's most eminent cosmologists, and he came up with what scientists consider the leading theory of how the universe was born, what happened in the first fractions of a second after the big bang.
The story begins early in Alan's career, in the 1970s.
Back then, the best theory for explaining how the universe began was the so-called "standard hot big bang model" .
.
the idea that the entire observable universe emerged from a tiny, hot, dense region of space, and has been expanding and cooling ever since.
But the hot big bang model was far from perfect.
The conventional big bang theory described how the universe expanded, how it cooled, how the matter coagulated to form galaxies and structures.
Oddly, though, in spite of its name, it really said nothing about the bang itself.
I like to say that it didn't tell us what banged, why it banged or what happened before it banged.
It's interesting, because we talk about the hot big bang MODEL, and we also talk about the hot big bang EVENT, and these are actually two different things.
In fact, the big bang event, the first moment in the history of the universe, may or may not be real.
We just don't know anything about what happens at time equals zero, whereas the hot big bang model is the theory of what happens AFTER the big bang, after time equals zero.
But the hot big bang model was about to get rewritten.
In December 1979, Alan came up with a revolutionary new idea for what happened just after the big bang, in the first fractions of a second of the universe's history.
He named this theory "the theory of inflation", and it was to have a profound effect on cosmology.
These actually are copies of the notebook pages that I wrote on the night that I came up with the idea that has become inflation.
I went home one night to my rented house in Menlo Park, California, and wrote down the basic equations, and I became very excited about it.
And I even made a comment here, with a double box around it, which is not the sort of thing I did very often.
"Spectacular realization", doubly boxed.
I had only been working on cosmology for about a year or so at this time, so I was very worried that when I started showing it to other people, somebody would point out something that was obviously wrong about it, but I was excited nonetheless, and then the next day I started telling my friends about it.
The equations that Alan had written down that night in 1979 were just the beginning.
Within a few months, Alan went on a tour of universities around America, giving a series of talks to promote his bold new idea of inflation.
I first heard about inflationary theory in the spring of 1980, when Alan Guth came by to give a talk at Harvard University, introducing this new idea of the inflationary universe.
Although only a young postdoc himself, Alan was ripping up conventional ideas about how the universe began and pushing further back in time than anyone had dared to do before.
The title of the talk that I was using at that time was "Ten to the Minus 35 Seconds After the Big Bang".
This was one of the most exciting talks I had ever heard.
In the theory that Alan presented, this period of inflation was a brief burst of extraordinarily rapid accelerated expansion that occurred for a short interval of time instants after the big bang.
Inflation is basically a theory, I like to say, of the bang of the big bang.
It's a theory that describes what propelled the universe into this period of gigantic expansion that we call the big bang.
A plausible number, for example, for the starting time of inflation might be something like ten to the minus 37 seconds after the instant of creation.
So, something like a trillionth trillionth trillionth of a second after the big bang.
An incredibly short length of time.
It took me a long time to convince myself that it made any sense to talk about these things.
What Alan was proposing was mind-boggling - that in the very first tiny fractions of a second, the universe went through a growth spurt on a cosmic scale.
Once it starts, the universe would double in size, over and over again, every trillionth trillionth trillionth of a second.
A typical size for the universe at the beginning.
Something like ten to the minus 24cm across.
Decimal point, 23 zeros, one centimetre.
Unbelievably small.
This is more than a billion times smaller than the size of a single proton.
By the end of the inflationary period, the final size of the universe will be something in the order of maybe one centimetre.
It's still small at that point, but this is still the early universe.
From then on, it goes from the 1cm stage to the universe we observe today.
Alan's theory was a revolutionary new take on the idea that the universe began with a bang.
The original hot big bang model had a number of flaws, not least of which was that what happened in the very first fractions of a second was a complete mystery.
Alan's theory neatly explained the first few moments of the universe, with a new ingredient called inflation, and it was inflation that caused the universe to expand astonishingly fast before suddenly slowing down to a more pedestrian expansion.
It was the closest that a theory had ever got to the actual moment of creation.
Alan had fundamentally rewritten the story of how the universe was born.
I can't imagine what it would have been like to invent inflation myself, because it's just one of those, you know, small number of incredibly influential ideas, that it really came together very, very quickly in the mind of one person.
The theory of inflation is a little bit crazy, but it really does a very natural, nice job at explaining some of the most puzzling features of our universe.
Alan showed in his lecture how inflation, this incredible burst of expansion after the big bang, could very neatly explain some of the great unsolved mysteries of the universe.
Why IS the universe so large, with a geometry that appears to be almost perfectly flat? And why, at the largest scale, is the universe so incredibly uniform? Alan's theory would also predict the existence of gravitational waves.
It was a triumph of science.
But then came the sting in the tail.
Alan's lecture about inflation theory built to what I thought was going to be the climax, and I thought, "This is fantastic.
" And then, Alan, in the last few minutes, explained how this idea fails.
During the period when I was going around the country talking about inflation, I was aware that there was a piece of it that I didn't understand yet, namely the piece about how exactly inflation ends.
Alan's theory of inflation had possibly solved some of the biggest mysteries of the universe, but it had spawned its own conundrum.
How exactly did inflation stop? This became known as the "graceful exit problem", and if it couldn't be solved, then inflation theory was dead in the water.
So I imagined that there might be some way of salvaging it, but I did not have any good ideas about how.
It fell to two other scientists to lead the rescue attempt.
Looking back, I think that this was the reason why I had an ulcer at that time, because it was like emotional disturbance which squeezes you like that, because you have this possibility to explain the origin of the universe, and then demonstrate that impossibility to do so.
It's just so painful.
I thought, "Well, this can't be right.
"There has to be a way to fix this idea.
" The problem was, once it had started, how did inflation come to an end? It couldn't just stop.
There had to be some kind of transition between the universe that was inflating and one that wasn't.
Alan's idea of inflation was to imagine that coming out of the big bang, in addition to the ordinary matter and radiation we know of, there was an additional form of energy that had the property that it was gravitationally self-repulsive.
Not only does it push itself apart, but it causes the expansion of the universe to speed up at an accelerated rate, which is what he wanted to have.
But you don't want to continue for ever, otherwise even today the universe would be expanding at this incredible rate, so you had to have some way that this energy would decay.
His original idea is that it would decay by the formation of bubbles.
These bubbles start growing, and when they grow, the walls of the bubbles, they move and collide with each other, and when they collide, it's like explosive process.
But there was trouble with the bubbles.
The idea just didn't work.
The problem is that the very inflation that was causing the smoothing and flattening of the universe blocked these bubbles from ever being able to come together, because the space between them was stretching faster than the bubbles were expanding.
Andrei Linde and, separately, Paul Steinhardt tried to rethink Alan's idea of how to end inflation.
The question was - if not bubbles, then what? The transition that I was talking about is like boiling water.
But there are other kinds of transformations that we observe in the laboratory.
There's a particular kind of transition that transformed in a smooth, uniform way, more akin to the way that if you were forming Jell-o.
The congealing of Jell-o.
And you began the liquid state, and then you cooled it down, it would uniformly solidify.
A gradual phased transition in which uniformly and largely within a space, the system goes from one phase to the other.
It was another spectacular realisation.
If the transition from an inflating universe to a non-inflating universe occurred less like boiling water and more like congealing jelly, then inflation would end gracefully.
The graceful exit problem had been solved.
At the time, it looked like really good news for inflationary theory.
Whether you finish the job or somebody else finish the job, this is the beauty of science.
When I received a preprint from Andrei Linde, I became very, very excited.
I remember running across the hall and talking to the person across the hall about how excited I was about this.
He probably thought I was crazy.
Alan's inflation theory had been rescued from the brink.
Science now had a plausible explanation for how the universe was born.
In the decades since Alan came up with his theory, new variations of inflation have been proposed, but the essence of the idea has remained the same.
And, crucially, the first real evidence for inflation has begun to emerge.
MISSION CONTROL: 'Lift-off for Delta II' Launched into space, a series of satellites - COBE, WMAP and most recently Planck - have carried out ever more precise measurements of the afterglow of the big bang and revealed that it has features precisely as predicted by inflation.
But there's one thing above all that would put the theory beyond question.
The discovery of its greatest prediction.
Gravitational waves from the beginning of time.
I think inflation is certainly the very leading theory for what happened in those very first moments in the history of the universe, but it's still not settled.
My view about inflation has slowly transformed from one of being proponent to one of being sceptic and looking for alternatives.
Before we just accept this is what happened, it's really, really, really critical that we see the experimental evidence.
That's why we have to go out with our experiments and actually measure something that tells us whether inflation is correct or is just the wrong theory.
OVER RADIO: 'Runway three, zero left' The detection of gravitational waves is often described as the smoking-gun signature of inflation.
It proves that inflation really is responsible for kicking off the start of our universe.
If gravitational waves are detected, that provides very strong evidence to this whole story of inflation.
We're talking about how the universe was born, and it's quite humbling I think, that Or surprising, that we can say anything about that time.
Instead of exploring inflation with our minds, to build a machine that can probe inflation, I mean, I think that's quite fantastic.
The theory of inflation naturally predicts that there will be gravitational waves, but it doesn't tell you how much, how big they'll be.
It just says that, you know, they'll exist.
So when people talk about gravitational waves from the big bang, you might be tempted to think that this is some huge shock wave propagating through the universe.
This isn't true at all.
The secret to gravitational waves from inflation is really quantum mechanics.
It's simply the fact that on very small wavelengths, quantum theory tells us that everything fluctuates, including the gravitational field.
Quantum mechanics says you can't pin down a physical system to something that is in an absolutely precisely defined state.
There will always be some uncertainty, there will always be some quantum mechanical jiggle in the universe.
So what that means is that the gravitational field during inflation is not precisely smooth.
It has these little jiggles.
ALAN GUTH: What inflation does is it stretches out these very small fluctuations to make the wavelength large enough that they can be observed in the early universe.
It's a deceptively simple idea.
Quantum jiggles in gravity were stretched out by inflation to become gravitational waves, ripples in the fabric of space and time.
And if inflation really did produce these gravitational waves, then it should be possible to detect evidence of them using telescopes.
So the way we are searching for this gravitational wave signal is to study the microwave background.
It's the oldest light in the universe and it's a treasure trove of information, but it really gives us a snapshot of the universe as it looked 300,000 years or so after the big bang.
The cosmic microwave background is essentially the afterglow of the big bang, released as the universe cooled down and, for the first time, light could travel across the cosmos.
Theory predicts that gravitational waves - if they exist - would have affected the orientation of the light waves, what's called "polarisation".
So the cosmic microwave background has a polarisation, and polarisation is kind of like a directional thing, so think of a pattern of little headless arrows over the sky.
Gravitational waves stretching and compressing would leave an effect on the polarisation of the cosmic microwave background.
It would produce a particular swirling pattern in that polarisation that we call a B-mode polarisation.
It's kind of a pinwheel pattern of polarisation, or a twisty pattern.
A curly pattern on the sky.
It's the unique signature of gravitational waves.
Gravitational waves stretching and compressing space would have left this pattern imprinted in the cosmic microwave background, the afterglow of the big bang.
If that B-mode signature is there, it is an incredibly powerful messenger coming to us from the first tiny fractions of a second of the universe's history.
The hunt for gravitational waves from inflation had become a hunt for this B-mode.
If inflation theory was correct, then this B-mode was out there, waiting to be found.
So, behind me, you see one of our telescopes.
It's actually a very simple design.
It consists of a number of small two-lens refracting telescopes, each of which has an entrance aperture of only about this big, 30cm.
It's scanning back and forth on a relatively small patch of sky, relentlessly collecting microwave photons, and it needs to do that because the B-mode polarisation signals that it's looking for are exceedingly faint.
John and his team have chosen a particular patch of sky to search for evidence of gravitational waves.
It's known as the Southern Hole.
So as far as we can tell, the cosmic microwave background looks much the same anywhere that we can observe it on the sky, and so the best place to observe it, to look for very faint signals, is where our own galaxy and the emission from our galaxy is the faintest.
So we pick a patch of sky called the Southern Hole.
It's about 1,000 square degrees.
The moon is about a quarter square degree, so it'sit's fairly large.
It's about 2% of the whole sky.
It's visible from the South Pole 24 hours a day, 365 days a year, and so our telescope can train itself on this small patch of sky and scan back and forth collecting data almost non-stop, almost as if it were in space and able to observe the sky unobstructed.
Doing astronomy at the bottom of the Earth brings with it its own unique challenges.
The team work on their telescopes during the few months of the Antarctic summer.
But it is in the Antarctic winter when most of the observations are done.
South Pole Station is only accessible for about three months out of each year.
The temperatures are only warm enough to fly planes in and out for that period of time.
So, when we take one of these telescopes, like BICEP1 or BICEP2, to the South Pole, we come in with a team and we work furiously for three months to try to get everything to work, to put it all together, to calibrate it, to tune it up, to get it in pristine condition and then all of us get on an airplane and leave - except for one guy.
He watches the plane go and knows there isn't going to be another one for about nine months, and during those nine months he'll watch the sun get lower and lower on the horizon and then disappear, and then six months of darkness.
WIND HOWLS And during that six-month night, the temperatures can get down to minus 80C.
The skies can be lit up with the most beautiful southern lights, aurora.
You get an amazingly unique experience when you spend a winter at South Pole.
It's akin to being in space.
In January 2006, the team's first telescope, BICEP1, began its observations.
Other telescopes, like the Keck Array, would follow.
For three years, BICEP1 scanned the Southern Hole, trying to detect the faint signal left by gravitational waves.
But at the end of the three years, the team had found nothing.
There was no glimpse yet of any B-modes.
The wild-goose chase may have been just that.
We knew that searching for gravitational waves would be a challenge far greater than any that we had taken on before.
BICEP1 was ultimately limited by the number of detectors that it had.
It was fewer than 100 detectors.
It just wasn't enough to make maps sensitive enough to tease out very faint B-mode signals.
In November 2009, John and the team were back at the South Pole to install a new telescope - BICEP2.
Even before the three-year observation of BICEP1 ends, we already knew we want something more sensitive.
You can't just go straight in and build the ultimate experiment.
The way that you get higher sensitivity is by learning from the previous experiment and developing the technology using the previous experiment.
So that's why we've had this succession of increasingly sensitive telescopes.
In BICEP1, we had nearly 50 detector pairs.
In BICEP2, we had about 250, and each one is slightly more sensitive than in BICEP1.
So in the end we get almost a factor of ten improvement in sensitivity.
So this is the detector technology in BICEP2, and what you see here is effectively a printed camera.
Each of these pixels is basically a camera, in the sense that it's not just the detector, it's actually the lens and the filter as well.
Like its predecessor, the BICEP2 telescope was in operation across three Antarctic winters, continuing the hunt for gravitational waves.
It was a quantum leap in sensitivity.
We were able to map the sky ten times faster with BICEP2 than we could with BICEP1, and achieve much more sensitive maps of the polarisation in the same patch of sky that we had observed before, but much deeper now.
This time, a tantalising signal began to emerge.
So we started seeing something interesting in, I think, I would say, December 2012.
There were hints of signal, I think, you know, before BICEP2 had stopped running.
Different people on our team have different memories for when they first started to suspect there was a signal in the BICEP2 data.
We analysed the data as it came in through 2010 and through 2011, the first two seasons of operation, but we quickly ran into a problem.
We did all these tests with BICEP2 and we couldn't get rid of this signal.
The signal that had been picked up by the BICEP2 telescope displayed all of the characteristics of gravitational waves.
It had the distinctive swirling pattern, the B-mode that John and the team had been looking for.
Yet it was less faint than they'd been expecting.
Something did not seem quite right.
It was at a much higher level than we were expecting, either from emission from our own galaxy or really from what we thought were favoured models of inflationary gravitational waves.
It was higher than either of those and so we thought, "Ah, well, this signal can't be real.
"It must be a problem with our instrument, "there must be some kind of subtle effect that we haven't yet "controlled in the experiment that's producing a false B-mode signal.
" Yet despite the team's initial doubts, tests confirmed that the B-mode signal was coming from the sky.
The signal was real.
Once you've established that there's a signal, then the next question is, what is the signal? You can't just assume right off the bat that you're looking at the signature of primordial gravitational waves.
It would be nice if you could, but unfortunately, although most of space is remarkably empty and therefore we can make a lot of measurements of the microwave background, we live in a galaxy, and within our own galaxy, there are a number of sources that can create a polarisation that have B-mode patterns.
So the two ones that we worry about the most are something called "synchrotron radiation" and something called "dust emission".
a type of light that is produced in the galaxy when its magnetic field sends tiny electrically charged particles whizzing around very fast in spirals.
This can mimic the B-mode pattern of gravitational waves, but using data from a satellite, the BICEP team were able to show that this effect was too small to produce the signal that they had detected.
The other big potential contaminant is dust, and this dust isn't so different from the dust in your living room that you see when the sun pours through the windows.
It's made up of carbon and silicates, just like little rock, bits of rock coated in water ice, and these little dust grains can line up in the magnetic fields, and then when light shines through them, they can create a little bit of polarisation.
Dust was harder for the team to discount.
There was less data available, but the models that did exist suggested that dust seemed unlikely to produce such a large signal.
Dust was also ruled out.
Once you're really convinced that you're seeing signal that comes from the early universe, then you can say you've detected primordial gravitational waves.
And the team now felt that this was the most likely conclusion.
After four years of analysis, everything pointed to the signal coming from gravitational waves.
So, as the final tests of the BICEP2 data set were completed by our analysis team, we called a collaboration-wide meeting and about half of us were calling in from the South Pole, and half of us from locations around North America.
We had looked at the data set in every way that we knew how to, scrutinised it from every different angle and convinced ourselves that there were no other tests that we could perform.
We were feeling at that stage that it was our obligation to share our results with the community.
In fact, that it was overdue that we do so, because we had had that data for so long at that point.
Until this moment, the team had kept their discovery under wraps, but now, finally, they felt compelled to go public with their news.
We knew that it was time to show the B-Mode map that BICEP2 had made to the world and not keep it a secret any longer.
In December 2013, John Kovac returned from the South Pole.
As the team prepared to announce their discovery, there was one man above all with whom John wanted to share the news, the man whose theory had predicted the existence of gravitational waves, the wild goose that John and the team now had evidence for.
John initially sent me an e-mail saying that he would like to talk to me urgently about an issue "that's very important "to your research and mine", in quotation marks.
A meeting was hastily arranged at MIT in Boston.
I got in a taxi cab and drove over here on the evening of Monday March 10th of this year, 2014, walked down the corridor behind me and carried the draft of the paper that we had been preparing.
We arranged for him to come through the back door of the Center for Theoretical Physics, so he would be unlikely to be noticed by the other people in the centre.
I was greeted by Alan very cordially, very calmly.
In fact, we both sat down and he knew immediately what this would be about.
This was the news that Alan had been waiting for more than 30 years to hear - definitive proof that his inflation theory was right.
Well, he told me that his group had been looking for years to examine the possibility of gravitational radiation from the early universe.
He told me that he was initially sceptical that it could be found.
He told me that once they had a signal, he was very anxious to make sure that the signal passed all possible tests, to be sure that it was real, and that gradually he and the rest of the group became convinced that the signal was real, and that now they were ready to make a public announcement about that.
John's team had potentially made a Nobel Prize-winning discovery and Alan's theory had perhaps finally been vindicated.
I think we were both very excited.
My reaction at the time was amazement.
I was astounded to suddenly have a group come forward and say that they had a measurement with unbelievably high statistical significance.
It really was a shock to me and of course a very pleasant shock because it would be very strong evidence for inflation, if it was real.
It happened in less than a trillionth of a second after the big bang.
FRENCH NEWS REPOR They detected gravitational waves or ripples in what they believe is the oldest light in the sky.
The news made headlines around the world but had actually started off rather low key.
On March 17th 2014, the team had held a press conference in Harvard to announce their discovery.
So when we announced BICEP2's B-Mode findings, we knew that that would generate some news, but we were not expecting anything like the attention that the release got, or the excitement that it produced.
Honestly, I expected the reaction to be fairly quiet.
The title of Detection Of B-Mode Polarisation In Cosmic Microwave Background, with that title, you wouldn't think it would generate much interest, but it did.
I first got wind of the BICEP2 signal about a week beforehand.
I can tell you exactly when I first heard about the BICEP2 signal because it was on Facebook.
The BICEP2 announcement was incredibly exciting.
This would be one of the final confirmations not only of inflationary theory, but also with general theory of relativity.
It appeared the smoking gun of inflation theory had been found.
'This is BBC Radio 4.
'Scientists in the United States say they've found 'the first direct evidence of what happened in the first moments 'of the universe, having detected gravitational waves 'or ripple patterns, in the oldest light in the sky.
'It's being called one of the greatest discoveries in science, 'and astronomers say what they saw confirms that what Alan Guth 'theorised in 1979 looks right.
'Other experiments are hot on the heels of the announcement 'so it won't be long before scientists find out 'whether their expanding model of the early universe 'is just a lot of hot air.
'Jeff Broomfield, MPR News.
' So after the initial announcement and the news, I think most people in the community, in the broader cosmology and physics community, accepted it at face value.
But as happens in every kind of claim discovery or scientific claim, then people begin to look more closely and examine whether or not that claim is really justified.
So since the March announcement of the BICEP2 detection of gravitational waves, there's been a flurry of activity trying to determine whether or not the source of the detected B-Modes is actually the gravitational waves from the early universe, or if it could be something more mundane, like, in particular, people are wondering if it could be dust in our own galaxy.
The possibility that dust in the Milky Way might produce a B-Mode signal had been considered by John and the BICEP team.
They had chosen the patch of sky, the Southern Hole, precisely because it was relatively clear of galactic dust.
Yet the team's work came under increasing scrutiny.
When we announced this result, the most important fact for us was that it was real.
It wasn't produced by the instrument, and that was the hard part, as far as we were concerned.
That's what we had spent the last 14 years doing.
At that time, based on the information that was available, these uncertainties were rather hard to quantify.
It looked very much like a large fraction, you know, 90% of the signal, was gravitational waves.
The uncertainty over the real identity of the BICEP2 signal was to grow, as in Europe, a separate team of scientists began to release new data about galactic dust.
The world's best data on polarised dust is available from the Planck satellite.
So what they knew about dust emission in our particular field in March, you would have to ask them.
Yes, we had already sufficient information to be quite sure that the optimistic modelling of BICEP was not proper.
The Planck satellite was a state-of-the-art telescope that, for several years, had been surveying the skies from its vantage point in space.
Its principal mission had been to make the most detailed map ever of the cosmic microwave background, the afterglow of the big bang.
But at the same time, Planck had also carried out the best measurements ever made of galactic dust.
The Planck satellite has started to release results on polarised emission from our own galaxy, from data that they've taken at much higher frequencies, where that polarised emission from dust is quite a bit brighter, and as they've mapped this out in detail across the whole sky, we've seen, as they've released their results, that that polarised emission from dust is actually brighter than the typical models indicated before the Planck data came in.
The new data from Planck showed that the models of dust that the BICEP team had relied on were wrong.
The dust was brighter than expected.
It brought the entire discovery of gravitational waves into question.
In our review, without having access to BICEP data, it was consistent with between 50% and 100% of the BICEP signal being from dust.
In other words, that it was possible that maybe half of it would be dust or maybe 100% would be dust.
That's what led us to sort of agree to collaborate with the BICEP2 team.
There was a great confluence of interest from the Planck team and from our own team, a realisation that there was a critical, exciting scientific question at stake here, and the best way to answer it was to use all the best data in the world, altogether, in one analysis.
So, in a sense, we set up a memory and an understanding between the two teams and went on with it.
In July 2014, the two teams began sharing their data, using the measurements of galactic dust emission from the Planck satellite to find out where in the universe the BICEP2 signal might have come from.
Precisely how much of the signal was actually from gravitational waves? Let's make an analogy.
I'm listening to some headphones.
Imagine that there are two tunes that are playing and I'm trying to hear one tune, which is the gravitational wave signal that we're after, but there's also another tune that's playing that's potentially a bit louder, right, and that's the galactic dust emission.
And so whether or not I can hear the tune that I'm after versus the tune that I'm not interested in depends on whether I can push down that distracting tune, the one I don't want to hear.
So, by using the data from Planck, the higher-frequency data from Planck, we can, actually, essentially, make that distracting tune quieter and improve our ability to listen for that gravitational wave signal.
We're at a fork in the road.
If this is positive, this is just amazing.
I mean, you know, it means, well, we have discovered primordial gravitational waves and we have actually discovered them jointly.
On the other hand, if it was a false hope, well, let's know it.
Whether the B-Mode patterns that we've seen so far do carry evidence of gravitational waves from the first instance of time, or whether, so far, all we've seen are swirls in the sky that come from galactic dust.
And on January 30th 2015, the two teams released the results of the joint analysis.
What BICEP has done is a tour de force.
It's a magnificent measurement.
It will be tantalising, exciting.
But we confirmed that this is not a discovery.
So the results of the joint analysis of BICEP2 with Planck is that the most likely answer is that 75% of the signal that we're seeing is due to galactic dust, but - and this is a very important but - there's a big uncertainty on that 75%, right? That uncertainty spans the range of half to 100%, right? I have a piece of coal and I'm not able to tell you how much gold is in there.
Did I detect something? Do you think you are rich? You just don't know.
So, I mean, I'm just telling you that what we have established is that there is no more than sort of half of the initial claim in gravitational waves, and possibly none at all.
The analysis had shown that the signal was most likely three-quarters dust and a quarter from gravitational waves but, crucially, the possibility that the signal was entirely from dust could not be ruled out.
The claim to have discovered gravitational waves could no longer be made.
So, at this point in time, the proper scientific statement is that there's no evidence of gravitational wave, B-Modes, and that if they exist, they're less than about half of the signal that we're seeing.
It's a disappointing result for the BICEP team.
Yet this is how science works, and it means that the wild goose of inflation is still waiting to be found.
The thrill of the chase is back on.
This wild-goose chase searching for B-Mode polarisation from inflationary gravitational waves is still ongoing.
We're certainly not at the end of this story yet.
Around the world, teams of scientists are embarked on the hunt for gravitational waves, from the Atacama Desert of northern Chile to the wilderness of Antarctica.
The quest to confirm how the universe was born, to find the fingerprint of creation, continues.
Science doesn't deal in hope, but, you know, we're human.
Humans do science and if we didn't have some some hope, then we wouldn't continue to do the work that we do.
One has to put aside one's personal bias here.
We have to look and see what the data tell us and go with whatever the answer is.
That's the cruel reality of being experimentalist, and that's the point of it.
Sometimes convergence to a scientific result doesn't always follow a straight line.
We were, like, here in March, you know, we were kind of bouncing back and forth, but we're just trying to make some solid measurements and I'm confident it'll converge.
The theory that Alan Guth came up with more than 35 years ago still awaits its final confirmation, but one early champion of the theory is now its harshest critic.
So where we stand today with inflationary theory is that what seemed like a very sweet idea at the beginning has some very serious flaws.
Inflation is in exactly the same state that it was in before BICEP2 came along - that is to say it's still the best theory we have.
Discovery of gravitational waves through inflation, it's a very big smoking cannon but we have quite a lot of other smoking guns which we already have discovered.
As far as I'm concerned, inflation is still in very strong shape but whether or not gravitational waves will be added to our list of pieces of evidence for inflation, I don't really know at this point.
When I was a young student, I was fortunate to come across a book called The First Three Minutes.
That book described the hot big bang universe up to the first three minutes or so.
The theories that we're testing with our present-day telescopes are much more audacious than that, of course.
The theory of inflation tries to push the frontiers back to the first trillionth of a trillionth of a trillionth of a second.
The very first instance of time.