E.E. Just Symposium Keynote Address: Dark Energy, Dark Matter and Inflation

Uploaded by Dartmouth on 15.11.2012

[ Silence ]
>> I want to introduce the-- pretty much I will say if there is one person that is, you know,
seriously responsible for my being here, our dean of Sciences, Dave Kotz.
[ Applause ]
>> Thanks Stefan.
It's my great pleasure to be here today and to attend many of these talks.
It's a fantastic symposium.
Thank you all for attending.
And it is my particular pleasure now to introduce the next speaker, David Spergel,
who is the Charles Young Professor of Astronomy at Princeton and also the Chair of the Physics--
of Astrophysical Sciences at Princeton.
And actually, I can relate to this in many ways.
He did his undergraduate degree at Princeton.
I did my undergrad at Dartmouth and then came back on the faculty of Dartmouth
and so I can understand that appeal.
He got his bachelors at Princeton in 1982 and then three years later his PhD at Harvard.
His biography goes on quite extensively, there's a lot of impressive awards
that I won't be able to name all of them.
For example, he had the MacArthur Fellowship in 2001.
He was elected this year to the American Academy of Arts and Sciences.
He is part of the new Princeton Center for Theoretical Science and is a co-founder
of the Institute for Physics and Mathematics of the Universe in Tokyo.
And he currently chairs the National Academy of Sciences Committee on Astronomy
and Astrophysics, so he gets around.
His interests range from the search for planets around nearby stars
to the shape of the universe itself.
And he's working now with others to develop technologies
that may enable the direct imaging of earthlike planets.
We have evidence increasingly-- extensive evidence on the existing--
existence of earthlike planets, but nobody has ever actually seen one.
And so it's pretty cool, I think, that he's working on ways to actually see them.
So with that, it's my pleasure to introduce David Spergel.
>> Thank you.
[ Applause ]
I want to begin by thanking all the folks here
at Dartmouth who'd made this very interesting meeting possible
and by the congratulating Stefan on starting here as the E.E. Just Professor.
I think it's a very exciting opportunity and we're all looking forward to coming back
and visiting you here over the years.
[ Inaudible Remark & Laughter ]
And, you know, you mentioned the American Academy of Arts and Sciences,
one of the neat things is next week I'll be inducted and that two of the people
who are inducted with me include Mel Brooks and Clint Eastwood and I'm bringing a chair.
[ Laughter & Applause ]
So I want to start.
There's a lot of students in the audience.
When you give a talk like this sort of looking at where you field is, it's fun to step back
and say, "Oh, where was this field when I was a student?"
And when I first took my first cosmology course, there were a bunch of questions
that we really didn't know the answer to.
How old is the universe?
It could have been 10 billion years old, 20 billion years old.
We know it was older than the Earth,
but beyond that there was a factor two uncertainty in the age of the universe.
How much stuff is in there in the universe?
There we had more like a factor 10 uncertainty.
You know, where the galaxies come from?
What's the universe shape?
These are all questions we didn't know the answer to.
And to sort of summarize in a sense where we've gone,
these are now questions we can get pretty detailed quantitative answers to.
The universe is now, we believe, 13.7 billion years old since the Big Bang.
And that's something we know to about a percent.
While we don't know what the dark matter and the dark energy is,
we do know how much stuff there is there.
And we had a pretty good idea about the overall geometry or shape of the universe
and a theory for how a galaxy is formed.
And to begin with the conclusion of the talk, today we have a cosmological model
with five basic parameters-- the density of atoms, the density of matter,
how old the universe is, how lumpy it is and how that lumpiness varies with scale.
With those five numbers, each of which I know to a couple percent, we can describe all
of our observations in cosmology and in astronomy,
of large scales and properties of the universe.
That's pretty good, right?
That's what you want to do in science, is to have a simple theory that fits lots of data.
The problem is, while it's simple, it's really strange.
Atoms make up less than 5 percent of the universe.
And we have given names to the other 95 percent.
But just because we gave them a name doesn't mean we know what it is.
In fact, I think one of the great mysteries of physics is trying to figure
out what the dark matter is and what's going on with the dark energy.
And I'll come to that-- back to those subjects as we go on the talk.
In order to describe how we got to this picture, I have to first teach you a few key ideas
in spatial relativity and general relativity.
But don't worry, these are easy.
They'll take less than five minutes.
All right, spatial relativity.
The key idea you need to know for spatial relativity is light travels at a finite speed.
That means when you look out in space, you look back in time.
It takes a few nanoseconds for light to get from Stefan to me.
So I see him as he was a few nanoseconds ago.
I don't see him as he is right now.
Now, Stefan hasn't changed that much over the last few nanoseconds
so he probably hasn't changed that much since.
[Laughter] The further out we go in space, the further back we look in time.
So we sit here and, you know, and hand over and we look out and we look out towards Mars
and it takes like, you know, a finite time to get to Mars so you can [inaudible]
with the Curiosity rover, there's a real delay, right?
You send a signal, the time it comes back
and told you you've done that, it takes like over 20 seconds.
Sorry, over 20 minutes.
20 minutes, sorry.
And that's a big effect.
You actually want to have a lot of autonomy for the robot because lots-- things can happen.
You know, Martians can come over and turn the Curiosity upside
down before we have time to respond.
The further out you go in space, the further back in time.
The nearest stars are about four light years away.
What does that mean?
That's a unit of distance.
A lot of times in science fiction shows, it really bugs me when they say oh
that happened four light years ago.
No. Four light years is distance.
It's how far light goes in a year, is one light year.
So is you're-- we see the nearby stars as they were four years ago.
The further out you go, the further back.
If you were look-- orbiting one of these planets that people are discovering now around, say,
a star that was 20 light years away, you'll be seeing us as we were 20 years ago.
You'd see me with full head of hair.
[Laughter] The further back you go in space, the further back you go n time.
So you'll hear about the Hubble telescope discovering, you know, some galaxy that was,
you know, we see it as it was 13 billion years ago.
We're seeing galaxies so far away that we're seeing them, you know,
relatively close to the beginning of the universe.
When we look out-- the further out we look, the further back we go in time.
All right, so you now know spatial relativity.
Good. Now I want to talk about the Big Bang Theory.
And when I talk about the Big Bang theory, I'm not talking about the TV show.
Though, one of my colleagues likes to claim that its view
of Caltech is not a comedy, it's a documentary.
[ Laughter ]
The Big Bang Theory really rests on two pillars.
One is general relativity, and general relativity,
as my academic great grandfather Johnny Weiler [phonetic] taught, consists of two ideas:
matter tells space how to curve,
and the curvature of space tells matter and light how to move.
So once I specify how matter is distributed, that tells me about the curvature of space.
And that's all people are pretty comfortable with.
The thing that I think people struggle with when they hear about the Big Bang Theory
and the thing that comes out of it which is to say the universe is expanding actually don't
like the name of The Big Bang Theory.
It was invented by Fred Hoyle who was an opponent to the Big Bang Theory.
He likes to study state cosmology.
I think a better name for it is the Expanding Universe Theory
because what it tells us is the universe is expanding, getting less dense with time;
and since it's filled with this hot radiation, getting cooler with time.
And I think of it as describing the history of the universe since the first few moments.
That's a very interesting question, Lee raised some of these issues, Robert raised some
of these issues, and what happens earlier on, but I think if I-- I like to start with the--
what we see today and run it back in time so if the universe is expanding going forward in time,
it's contracting going back in time.
When we hear the universe is expanding, the first question many people ask and some
of you are probably thinking this now, what is it expanding into?
And that's because you're thinking abut space as absolute that it's, you know,
it's like a balloon that's expanding in this room.
It's expanding into the space around it.
What I like to think about relativity is it says space is not absolute, space is relative.
Space is only defined in the relationships between objects.
So we're not expanding into anything.
Things are getting just further apart.
Or if you want to think about it expanding into something,
think about expanding into the future.
And the image I have when I think about the universe's expansion is I can't think
about four dimensions so I think about three.
I ignore-- and I ignore one dimension of space so I imagine that we're all living
on the surface of a big ball and we're ants and all the galaxies are ants in the big ball.
And as the universe expands, it's expanding into the future.
So that balloon is getting bigger and the radial direction is time.
The radius of the balloon is time.
So as the universe expands, the ants on the balloon get further apart
and that's the expansion of the universe, making ants getting further and further apart.
And as we go further into the future, it gets-- the ball gets bigger and bigger.
We'd run that back in time, the universe gets denser and denser, hotter and hotter,
and the moment of the Big Bang is when I shrink that balloon down to zero.
Okay. So here's history of our universe.
Here we are today looking back in time at stars and galaxies forming.
Keep in mind, the age of the Earth if kind of interesting on this, right?
The Earth is about four and a half billion years old, so the Earth--
the age of the Earth is about a third of the age of the universe, so we're relatively recent.
But in the grand schema things, we kind of show--
or when the Earth forms, shows up in this [inaudible].
Our solar system forms.
Go further back in time, a lot of what I'm going to talk about today is the microwave background
which is the leftover heat for the Big Bang.
And most of what we see in the microwave background is what the universe looked
like about 400,000 years after the Big Bang.
And I'll talk about some results from our space observations with the WMAP
and some measurements we'll be making from the ground.
So first let me begin by talking about WMAP that stands
for the Wilkinson Microwave Anisotropy Probe.
Wilkinson's easy, that's my late colleague Dave Wilkinson
who played a pivotal role in experimental cosmology.
And after he died, NASA generously let us name the satellite after him.
Microwave, that's suppose the-- it's the radiation we're looking at.
And it's just like your microwave oven.
It's that wavelength of radiation.
Anisotropy is a word that many people struggle with.
I know when I get introduced, people often can't pronounce it.
And it just means how things vary across the sky.
Isotropic means the same everywhere on the sky.
Anisotropic means it varies from place to place.
So all that says is we're looking at how thing vary.
And this just shows where the satellites goes in order
to make better environment that's very clean to observe the leftover heat from the Big Bang.
We actually go out at a distance about four times the distance of the moon.
And we're out there or-- we were out there in space orbiting,
taking pictures of this leftover heat.
So when we look out in space, we look back in time.
If we look back in time, the universe gets hotter and hotter.
How hot the universe is relative to today?
It's one way of thinking about redshift.
You hear astronomers talk about redshift.
So when we get to redshift of 1,000, that's when the universe was 1,000 hotter than it is today.
10,000 when it's 10,000 much hotter.
So that going this way, the universe is getting hotter and denser
and this plot shows how density behaves with time.
As the universe gets hotter, it goes from being made up of neutral hydrogen to ionized hydrogen.
Neutral hydrogen is transparent to radiation.
So light can travel freely through it, or microwave light.
Ionized hydrogen, electrons and protons, the radiation interacts with the electrons.
So it acts like a dense cloud.
So over here it's like a-- those of you who are up--
you know, was up running in the morning, it's all foggy.
Dense fog back here.
We can't see through it.
As we look out in space back in time, we see back to the moment or the period of time
when the universe made this transition from being neutral to being ionized.
So our satellite maps what the universe looked like back then.
So this cosmic background radiation, if we had eyes that could see in the microwave,
would look more like this orange picture.
It's actually very uniform at a temperature of about 2.7 degrees.
You have to look with very high sensitivity and this is the image
from the COBE satellite which was WMAP's predecessor.
And it's made the first detections of variations in the microwave background,
variations about 1 part in 100,000.
And our goal was to try to look at that in more detail to be able to determine things
like what's the shape, what's the geometry of the universe.
Since you're all general relativist now, you know that the amount
of matter determines the geometry.
So if you got a lot of matter, it's going to curve space a lot.
Space will be positively curved closed geometry.
That means when we look at the-- left these fluctuations back then,
we actually know how big the fluctuation should be.
The characteristic size is how far light can travel in the 400,000 years.
So nature's is kind of holding up a ruler for us that's 400,000 light years across.
And if the universe was closed, those spots would look very big
because like the path light would take-- would be different.
The density of the universe was low, the spots should be smaller.
If the density of the universe was jus right, so that the total energy of the universe is zero,
that the energy in expansion equals the energy in mutual attraction, which is negative,
then the spots would be about a degree.
So those are the kinds of things we want to measure
by making a precision measurement to the sky.
So we sent the satellite up.
It was really fun.
We went down to the cape.
Put it on a big delta rocket, launched it into space,
and I actually did go to Disney World afterwards.
[ Laughter ]
So here is our satellite.
Oops. There's our satellite.
But I want to see it right, spinning in space.
There you go.
I'm just going to have to hit-- and it's out there spinning in space.
It's measuring these tiny fluctuations.
And the way we try to make these measurements
to higher precision is we make a differential measurement.
A very good experimental trick is if you want to measure something to very high accuracy,
don't make an absolute measurement, do it differentially.
So if I brought two of you down who were both, you know, one is 5'8", one is 5'9"
and I want to measure who's taller.
If I take a ruler and try to measure it with those 12-inch rulers we've got
in elementary school, that's hard.
On the other hand, if I put you back to back and I put a ruler on your head,
it's easy to see if people are back to back who's taller.
That's a differential measurement.
And if you make a careful differential measurement and you do things like we want
to make sure this ruler is not warped so we're going to flip it around lots of times
and that's what we do with the spinning satellite.
We have two different points.
We look at it in any given time.
We measure the temperature difference.
We spin around, a minute later, measure the difference again, reversed.
You can make a very accurate map of the sky.
And that's what the satellite does.
And these are our maps of the sky.
We mapped them at different frequencies.
This is like seeing what the intensity looks like and how much red light there is
in different parts of the sky, how much yellow light,
how much green light except for radial frequency.
So we're seeing how much energy emission are we getting
at 22 gigahertz, 30 gigahertz, and so on.
And this is a map is like looking at a globe where you take the sphere
and project in on the plane, right?
I could-- this is showing the sky around us.
And the first thing you'll notice if you look at this picture is
that big bright red thing across the sky.
That's the galaxy.
That's dust and electrons in our own galaxy, the Milky Way.
That's not what we're interested at.
That's stuff that's been traveling towards us for about 10,000 years.
Boring. You know, we're interested in this blue and yellow stuff on the bottom and the top.
That's the stuff that's been going-- traveling towards us for 13.7 billion years.
So we take data at many different frequencies, five in our case.
As we go up in frequency, the galactic emission gets less
and we know how this galactic emission varies with frequency.
So what we do is we try to make our best map combining all of the frequencies
and try to remove the contamination from our own galaxy and from quasars and nearby galaxies
and this is our picture of what the universe looked like 13.7 billion years ago.
And this is a really handy picture to have because it can tell us a lot
about the properties of the universe.
We can count up and see what's the number of spots we see on the sky.
How big they are?
How the fluctuations vary with scale.
So I can ask, you know, how lumpy is this map at the 1 degree scale,
the five degree scale, the 10 degree scale.
Do I see equal numbers of hot spots and cold spots?
Lee mentioned the possibility of looking for non-Gaussianity.
One form of non-Gaussianity would be we have more really bright hot spots than cold spots.
That would be a non-Gaussian fluctuation.
And that'll be very interesting, had we seen that, that would have told us
that there's something beyond our current simplest theory.
As Lee pointed out the interpretation of that is more difficult, but that's the kind
of thing you can look at that would tell us something new from this data.
Now, when you look at data like this, if you look carefully,
you'll start noticing patterns like this.
Notice the letters SH, I first noticed this-- actually in my colleague, Lyman Page
and I were sitting in a big conference.
Steven Hawking was speaking.
This was at UC Davis in 2003.
Lyman turned to me who did like-- Lyman designed most of the instrument and said does that--
did they put up the letters SH as a joke?
Is this right?
Because of-- it's what the students put there and we looked at our data.
No. It's there on the data.
[Laughter] This maybe telling us something profound.
[Laughter] My own theory is that the galactic emission
over here to the right is a bit stronger.
Perhaps we'd lost some information when we removed it.
We're missing the letters I and T.
[ Laughter ]
When the initial conditions were created,
they realized there was a mistake made and [inaudible].
More likely, I think, this actually represents evolution.
We've been selected to find patterns and it's-- those of us whose ancestors, you know,
they'd look in the woods and nine times out of ten
when they jump was they thought there was a tiger, they were wrong.
But one time out of ten, they were right.
So it's better to error on jumping when you see something.
And the people who didn't jump at all when they saw something
in the woods, they're not our ancestors.
[Laughter] They were ET.
[ Laughter ]
So how do we take-- what do we do with data like this?
Well, we try to quantify what's going on with the fluctuations.
We measure how lumpy as a function of scale.
And this plot here shows the points,
the amplitude of the fluctuations as a function of angle.
And the conventional way we do it, spherical harmonics for those of you who are more expert.
[Inaudible] the units we use are 180 over the angle.
So that at ten, we're measuring how lumpy things are smooth on the 18 degree scale.
The 1 degree scale, at about 200.
The tenth of the degree scale, when we get out [inaudible] of 2,000.
So just looking at the lumpiness as a function of scale,
and you'll notice there's a really regular pattern you see.
And that regular pattern is set by the fact that we're looking
at are sound waves on the early universe.
And that the amplitude of the signal depends on how the sound waves behave.
So the spacing between the peaks depends on what universe is made of.
You have more atoms that changes the spacing between peaks.
We measure that.
The position of the peaks depends on the distance from here to the surface.
That's how we know that the universe is 13.7 billion years old.
If the universe was older, the peaks would shift to the right.
Younger, they'd shift to the left.
The relative amplitudes of peaks depends on the composition
of the universe because gravity matters.
The amount of matter in the universe determines, particularly dark matter,
the amplitude of that third peak in particular.
So by seeing how-- if there was more matter on the universe, that peak would be higher.
So you can look at this data and it's all just sound waves in the early universe.
It really is fundamentally, you know, it's freshman physics.
So it would take one or two fresh lectures for a freshman class glass to get it.
But we can work all that through and see what we predict for a simplest model.
And what we find with the simple model with five parameters, we predict that red curve
and that's a pretty good fit to the data.
And when you look at the properties of the fluctuations,
the universe turns out to be remarkably simple.
That whole math, I only need five numbers to describe it,
the fluctuations that we call Gaussian.
So this is the number of hot spots versus temperature as a smooth on the 4 degree scale,
the 1 degree scale, and the quarter degree scale.
The black histogram is our data.
The red curve is the Gaussian curve and it has one parameter, the width.
So once I specify that one parameter, that red curve is completely predictive.
And you can see it's a very good fit to that-- those-- that black data.
We don't have more hot spots than cold spots.
They're really basically equal numbers.
The universe turns out to be remarkably simple.
We don't see any evidence of interaction.
We don't see any evidence that the universe is finite.
We look for these things and so far it seems very simple.
So we now have the universe's baby picture.
We know what it looks like 400,000 years after the Big Bang.
So we can then evolve it forward in time.
And for that purpose, I've selected three randomly chosen babies [laughter]
and evolved them forward in time to about-- this is about eight years ago.
And then that's what the universe would look like, you know, about when the Earth formed.
And we can then compare them to our observations then.
And we could then evolve it even further forward in time-- oops, you don't really care,
but that's a picture of my kids skiing.
[Laughter] And that's further forward in time.
And this is what we do with our models.
We take the initial conditions, we put them in a big computer simulation,
we evolve that computer simulation forward in time starting with the initial conditions based
on what we see in the microwave background to go to, at least talk,
this is sort of classical physics in the sense.
You're specifying things not in the very first moments, but 400,000 years after the Big Bang.
That sets our initial conditions.
The laws of physics we use here is general relativity.
You're all relativist, it's straightforward to do.
And we evolve it forward, you see structure emerge.
And we then compare the structure we see how lumpy that is today in the models
or how lumpy it was when you-- at the time the Earth formed and we compare it
to our observations and see what we see.
And to do that, you need to see what the galaxies are.
And one of the efforts to do this is the Sloan Digital Sky Survey.
This is my colleague Jim Gunn who led the survey and here is some work by one
of my thesis students, Beth Reed, where she took some of the data that was collected
and measured the lumpiness versus scale.
And black curve is an extrapolation from our data and those are the points.
In this plot, don't worry about the axis, but just lumpiness versus scale
and it just shows we can match the properties of today's universe running it forward in time.
That says that our model has the right ingredients.
And in fact, as we walk at all the different ways we can look at the universe.
So, in cosmology, how fast the universe is expanding?
We do that by measuring the Hubble Constant, it's one of the things the Hubble telescope did.
The age of the universe, we can look at how old stars are.
The properties of clusters, and I'll show some cluster data later.
The abundances of the deuterium and helium produced in the universe.
How much lensing we see of distant galaxies?
The properties of distant gas cloud seem toward quasars.
Supernovas, which are these powerful explosions we see
through the universe and this is the supernova data.
For example, this platter is a function of red shift.
Remember, that's how hot the universe is, how far away we look
versus the brightness of the Supernova.
If there was no dark energy, the curve will look like the dash line.
At best, what we project from our data is the solid line.
Those points are the points from the supernova and you can see our data,
the mile that fits our data fits that quite well.
In fact, you know, go and-- I've given versions of this talk in Stockholm.
And it's one of the things that can help convince the folks in Stockholm
to give the Nobel Prize this past year to Riess, Schmidt, and Perlmutter, shown here smiling.
And what they all did, and this is the strongest evidence we really have that the universe--
we have this dark energy, the universe is accelerating, they looked at distant galaxies.
They then came back a bit later.
Notice when they took data in those distant galaxies,
these bright arrows appeared [laughter] and the bright arrows pointed to these exploding stars
that were so powerful that they're as bright as the galaxy.
We know how bright those stars are intrinsically,
these explosions called supernova.
We looked at how bright they are.
We looked at that brightness versus distance, and that's another way
of measuring the basic properties of the universe.
And all these pieces fit together.
Here's another way of seeing this.
The same sound waves, this is a big Sloan--
recent data from the big more recent Sloan survey, version of the Sloan survey.
This is actually led by a guy Dan Eisenstein.
Dan took freshmen-- since junior year took cosmology from me.
And it's kind of cool when there are people who you--
first taught them as undergraduates are now leading major projects that are finding evidence
for things like the existence of dark energy.
And you can see at this plot below, those same sound waves that we saw in the early universe,
that made those fluctuation patterns, Dan and his colleagues are finding
in a large scale distribution of galaxies in your body.
We're seeing that same imprint.
And the position and height of those peaks are bang on what the simple model predicts.
So you take what fits the data we have from space,
predict what Dan's group should see and it fits very nice.
Let me just skip through clusters.
This is fun.
There's a list of all the parameters measuring all the different ways.
That's a real eye test and you should just see the numbers, all look about the same.
So, how do we push the model further?
You want to test it.
So this is a plot of how lumpy it is versus scale.
The black points of the data, but our theory predicts what you should see
on even smaller scales.
So you want to go test that.
So we go and build a small-- even more sensitive small scale experiment for--
in our case, we went to Chile to the Atacama Desert which is very cold and dry.
Our colleagues went to the South Pole and there's a very similar experiment
at the South Pole, also mapping the sky.
And we look at this microwave background on smaller patches in the whole sky,
but at much higher sensitivity and higher resolution.
And here's a Goggle Sky image of our telescope.
Because it's inside a ground screen,
we can't see what it looks like from above, but Goggle can.
Goggle Sky images are pretty impressive, you can look at your--
I know we can look at our house and I can count the number of sky lights in my house.
I can just-- Goggle Sky knows there's a sky light over our bathroom.
[Laughter] You guys know everything.
And so we're zooming in and looking at these fluctuations on smaller scales.
This is what our data actually look like.
Actually, we turned down the light just a little bit.
You can actually see-- this is sort of nice
because you see a bit of the messiness of real data.
You'll see these bright little sources, that little dot, that's a big radio source.
We have to mask that.
And towards the bottom and the top, you see stripes.
That stripes, that's the Earth's atmosphere.
We-- those are regions where we have it masked-- mapped often enough because we'd map it more
in the center, and we're left with noise from the atmosphere,
at least [inaudible] streaky, we have to filter that out.
You're trying to measure things that are fluctuations
at level of a few millions of a degree.
The Earth's atmosphere is a lot hotter than that and you have to deal with that.
And we focus, you know, and we see actually this whole black spots are clusters casting shadows.
And one of the fun things, I'll show you one or two images of the clusters we discovered.
But most importantly, we looked at these smaller-scale fluctuations.
Here's the clusters and we go, we see these clusters as shadows;
and what's fun is you then go to your optical telescope, you know where the shadow is,
you discover new clusters of galaxies.
And we've discovered hundreds of new clusters.
And the numbers of clusters we see,
and their properties are again consistent with our basic cosmology.
And remember I had this prediction on what the model should look like.
How do we do?
We do remarkably well.
So before, our curve only went out to here.
We saw those first three peaks in the sound waves.
Now, we see about eight picks and the amplitude of those sound waves, how they're propagating,
how the universe behaves when you hit it and bang it
and let it shake a bit, just cons-- is very consistent.
So this is, you know, kind of classically how we do science, right?
You have a model, make it fit the data, predict more sensitive--
make predictions, you go test it, it's holding up really well.
And these are two independent experiments.
The red points are from measurements made in the group in the South Pole.
The blue points are our measurements, completely independent, consistent story.
So, this is great.
The good news is we have a standard model.
All the pieces fit together.
We know what's going on.
Well, it's not quite true.
The glass is either half full or half empty.
Atoms make up only five percent of the universe.
We don't know what that dark matter is, we don't know what the dark energy is,
we don't really understand where the fluctuations came from.
We have interesting ideas, but we don't believe them.
So, let me end by asking three questions.
What's the dark matter?
Is it new particles from supersymmetry?
Those of you who are lucky-- fortunate enough to hear Jim's talk yesterday,
you heard a bit about supersymmetry.
One of its really neat predictions is it offers a way of explaining what the dark matter is,
is that the lightest particle predicted
by supersymmetry could be stable, it could be the dark matter.
And what's very exciting about that possibility is it's one
that we might know what the answer to soon.
This big Collider in Geneva could see it.
We might see it underground searches looking for dark matter.
We might see its astrophysical signature.
So we actually might know that the dark matter is soon.
Or it could be something very different.
It could be black holes.
It could be a new type-- another type of particle collapsing on--
or maybe something we haven't figured out yet.
[ Pause ]
I mentioned the universe seems to be accelerating today.
It's actually very strange.
The universe is behaving and that's what the supernova data is showing,
that's what our data is showing.
As if you threw a ball up instead of gravity making it fall back down, it kept accelerating--
it started accelerating and started moving faster.
So this expansion of the universe, you'd expect to slow down from gravity.
But instead of slowing down, it's accelerating.
It's very weird.
We don't know whether this is due to the fact that General Relativity, which is we're applying
on these large scales is breaking down.
Or whether it's being driven by energy associated
with empty space that we call vacuum energy.
Or whether it's being driven by some new physics entirely.
And, Stefan, I will get to that paper draft on our ideas on this new physics.
But we'll get-- life is busy, Too many things to do.
One of the intriguing things we're seeing is we see this pattern of fluctuations.
And this pattern of fluctuations, we see it consistent with this idea
of inflation that Robert talked about.
And this inflationary model has some very attractive features in it.
If you go back and look at this inflationary model and what are predicted in the 1980's,
it predicted those fluctuation should be Gaussian.
It predicted the amplitude should be about the same on all scales.
It predicted that they should be able to call Adiabatic, which means when I have more atoms,
I have more dark matter, I have more radiation.
And when I have less, I have less of everything.
It predicted the basic properties of those fluctuations.
It-- you know, it's the-- one of the basis we use
for computing those beautiful curves that fit the data so nicely.
So observationally, it seems to be doing really well.
The problem is, as we started thinking
about its theoretical underpinnings, there's a lot of problems.
We don't understand where the initial conditions come from in the model.
We don't understand-- or to make inflation work, you have to tune all these things
in the model very, very accurately.
We don't know why we have to do that.
We know it's a-- we can make it work by tuning every thing,
but it seems like an artificial construction in some ways.
There's lots of problems.
The model tends like to run off and it tends somewhat to predict the universe
that keeps inflating forever, it doesn't stop.
So, my own feeling is this, you know, we do seem to see evidence
that the universe accelerated in the past.
It's happening now, it probably happened then.
I think our theoretical understanding of it are-- is that an early stage, perhaps where--
like the theory quantum mechanics where there was an old quantum mechanics
to explain the hydrogen atom that Niels Bohr developed
and it was a good start, but we needed something deeper.
And one of the things that you've got a taste of and some of the early--
the talks today and you'll hear some more tomorrow, is different ways we're struggling
with trying to understand these kinds of questions.
And as you know, alternative versions we can come out,
and we have to get the final revision done.
And while this is an area where we're asking very profound questions
that cross the border sometimes between physics and philosophy,
it's also in area where there are significant-- there are real observations that we can make.
Some we're making now, some we hope to make in the next few years
that can test some-- a lot of these ideas.
So, one of the things that is predicted, for example, by law,
these inflationary models is a pattern of gravitational waves,
ripples and space time itself that are actually predicted
to leave a distinctive signature on the microwave background.
And one of the kind of next generation experiments that we're thinking about
and working on is trying to detect these gravitational waves.
So we're not, you know, these are models that make real predictions and will be able
to separate or at least falsify some of the models by these future observations.
So let me conclude.
And this picture of a turtle and a hare, I tool in my front yard.
I don't know what it symbolizes, but I was just so cool to see a turtle and a rabbit lined
up looking like they're about ready to race.
But I thought I had to use it in the talk.
[Laughter] But I tried to convey in my talk sort of two-halves of the story.
On one hand, we've made significant progress in understanding--
and one of the properties the universe, it's composition, it's age, how lumpy it was,
where galaxies come from, and we have a pretty simple model
in some ways that fits a host of data.
And the data over the past decade has improved dramatically and will continue to improve.
And as it continues to improve, well there's two possibilities.
It will either keep fitting this model or it will point to something beyond that model.
And I think we sort of know there's got to be more stuff out there that we haven't understood.
We have a model in which atoms make up 4 percent of the universe and-- or 4.5 percent.
We-- and we don't know what the rest is.
We're missing some important things.
You know, and we don't understand why there's more atom--
more, you know, we heard about matter and antimatter, why are there more electrons
in the universe than positrons, antielectrons?
We have ideas for this, but we don't know what the right theory is.
So this question of, you know, why is there the stuff that makes us up?
We don't understand the atoms that well either.
And while it's very exciting to me that we've seen what the universe looked like 400,000 years
after the Big Bang and can extrapolate back to the first moments after the Big Bang,
we're still struggling to understand what that data is telling us
about the basic properties of the universe.
So, you know, for those of you who are students thinking about which way to go
and this is an area where those of us working on the field, we've solved some amazing problems.
And there's some really hard interesting ones for you to solve and, you know,
you're probably going to be the ones who'd do it.
So that's a great-- it's a great and very exciting time.
So let me stop there.
[ Applause ]
>> All right, so the-- we actually have of time for discussion, 10 minutes.
But people have to-- we then-- we are going to head over to 104--
>> 105.
>> Sorry.
>> Kemeny.
>> 105 Kemeny.
[Inaudible] to myself.
Working with where we were before.
>> Why not?
>> But I'm serious.
[Inaudible] And the next talk is also going to be pretty cool and exciting.
No panel discussions and [inaudible], but-- [inaudible].
[ Inaudible Remark ]
Now, I'm going to put on medical hats on, okay?
>> Bur not yet, I still have 8 minutes.
>> Yes, 8 minutes.
So yes, the question, and wait for the mic to get to you.
>> I came in late and I wonder if you'll address-- oh.
I came in late and I wonder-- and did you, you know, earlier address the issue
of whether there's something drastically wrong with General Relativity and all this?
>> Well, I'd put that not-- I didn't address it.
I raised it as a question [inaudible] sense.
The fact that we see, in my conclusion, that universe is accelerating,
perhaps it's telling us that what's going wrong is the way we're interpreting the observations,
what's going wrong is the way we're applying General--
is General Relativity is invalid on those scales.
Now, we can test this in different ways.
General Relativity makes a prediction on how distance should behave with red shift,
so we can look at the supernova observations and see how they-- what--
whether the supernova data fits the model for how distance behaves with red shift.
Then we could look at how structure grows, you saw that computer simulation
of how the universe got lumpier with time.
So we compare the lumpiness we see with time versus the theoretical prediction
and see if structure is growing the way we think it grows.
There are modifications of General Relativity that people have come up with
that make different predictions for the relationship between distance and red shift
and structure growth and red shift.
So far what we see is consistent with General Relativity,
but one of the things we're interested in doing is improving our data and seeing if we can test
that the universe is behaving as the theory predicts on those scales.
And, you know, we don't have-- you know, people have many alternative theories of gravity
that replace General Relativity to explain the cosmic acceleration.
I think the theories at this point, I think of them as baby aardvarks.
They're beautiful to their parents, [laughter]
but no one else finds them theoretically compelling.
With that said, some of the most interesting ones have the advantage of, you know,
not requiring dark energy and making new predictions.
And one of the things we do is we try to make ever better measurements
and measure the microwave background fluctuations,
the properties of galaxies in-- with better-- more detail.
And here we're helped by, basically, our engineering colleagues
who build more sensitive detectors, bigger cameras, developed the technology to do that
and then we can make more precise measurements and test the theories.
I mean an interesting kind of philosophical discussion is does science drive engineering,
or do advances in technology drive science.
And my own view is yes.
[Laughter] Yeah?
[ Pause ]
>> I have a very elementary question and it's about the observational method.
So, this notion of the background radiation and that it would be comprised if microwaves
and those are the things that you would measure with-- when you got to out to outer space,
where did those notions come from, why would select microwaves as the primary signatures
of the supposed origin of the universe.
>> So we stumbled on to it, that's the history of it.
So, Arnold Penzias and Bob Wilson are two astronomers at Bell Labs.
The '60s, Bell Labs was supporting a lot of fundamental research and Bell--
but stuff that was useful for the company and one of the things they wanted
to do is a maybe we'll use microwaves for phone communication.
We should study what the universe-- what things looked like in microwaves and map the sky.
So Penzias and Wilson built a really sensitive microwave experiment, and they went out
and they looked and they worked, you know, for the company, you know, I'm sure when they went
and did the presentation for their company Vice President they said, "We're doing this so--
when AT&T uses microwaves, we'll be ready."
And-- but as a scientist they want to see what was there.
And they saw our galaxy coming up and down that pattern--
remember that pattern, you saw a red, they found that.
And then they saw this uniform stuff everywhere they looked.
And one of the things that made them
such great scientist was they didn't ignore something they didn't understand.
And they had this leftover signal and they worked for a year trying
to find the source of that leftover noise.
They replaced detectors, they, you know, famous story there was a pigeon boosting
in their telescope and they all thought that maybe it was heat from the pigeon crap.
So they went, you know, well sometimes part of science is going in there
and clearing out the pigeon shit.
And you go out, you clear that out, sometimes it's metaphorical
and sometimes it's real pigeon shit.
And they did that and result stayed.
And at the same time, some of my colleagues, Bob Dickey,
John Pebbles were actually independently rediscovering some work done in the '50s
by Alpher and Herman and Gamow, realized that if you start
out with the hot Big Bang you could explain the abundance of helium in the universe
and predict this leftover radiation, and the temperature would turn
out to be such, that'll be in the microwave.
So that's how we found out there was a hot Big Bang.
And then once you saw this uniform radiation,
the next step was trying to find the fluctuations.
And that led to a series of experiments of increasing sensitivity,
is people looked for the leftover fluctuations.
And, you know, initially the models predicted the fluctuation should be much brighter
because the models were based on the universe made of atoms and radiation
and electrons, you know, and protons.
It didn't assume any dark manner and said that the fluctuation should be
about ten times brighter than they should be.
And in fact, you know, I remember as a student, people hadn't discover those fluctuations yet
and it was-- you know, I go talk about subject, they were boring because they're just limits
on what you saw and, you know, we slowly move forward
until the Colbert Result came out I only saw the fluctuations.
And then, you know, the ideas here have come at-- if you look at their history,
they kind of made different ways.
So the notion of dark matter goes back to work from Zwicky.
When he looked at these clusters of galaxies, he noticed the galaxies themself were moving
around so rapidly that there had to be more stuff holding it there.
So he said there had to be extra matter and he posted that in the '30s
and there was work done looking at the Andromeda galaxy
in the early '30s that showed that it had it.
And it was really only in the '70s when that idea got revived as people trying
to understand galaxy properties realizing you need a dark matter.
And this idea of a cosmological cause or a vacuum energy, well the mathematics
of it goes back to Einstein and Einstein put the [inaudible] term in.
Einstein actually cleared in for the completely wrong reason.
Einstein looked at his theory of General Relativity,
realized it predicted the universe must be expanding or protracted.
As far as Einstein knew, the universe was static.
He didn't know about the work that Hubble was doing.
And so he added a term to make the universe static,
it's something he called his greatest mistake,
that this theory actually predicted the expanding universe, but then he screwed up
and he took-- you know, then he learned about Hubble, then he threw the term away.
But that term kept coming back.
So in the '50s, there were some deviations.
People postulated and then by the '90s, there was kind of growing theoret--
observational evidence that would be better understood by adding a constant--
constant logical constant and the theory it made that fits the data better.
But it was the supernova data in the late '90s that was really the turning point and it was,
yeah, it wasn't that it was a surprise, that data, but it was strong enough evidence
that really swung the pendulum of intellectual opinion.
And it's the point where it really-- the case became strong for dark energy.
Yeah, [inaudible] right about there?
[Inaudible] do you want to just pass the mic up or?
>> Thank you.
I sort of had a question that might be a little bit of a philosophical question
but in the last talk by Dr. Smolen [phonetic], we were discussing the possibility that the laws
of the nature might have evolved overtime because they can't be completely extrinsic
to the system that they're part of.
And so, if the standard mile that you're presenting can predict accurately
and stay consistent with, you know, the cosmic microwave background
and it's also relevant today, does that sort of preclude the possibility
that the laws of nature evolved over time or?
>> No. And in fact, the question where the laws of nature could evolve with time is one we try
to ask experimentally or observationally extent we can.
So, you know, what we can do is we can ask 400,000 years after the Big Bang to the point
of which were observed in the microwave background,
was the mass of the electron same as it is today.
Was the strength of the attraction between electrons and protons the same as it is today?
And we make an estimate of what will do.
And it turns out what we can say is the laws of physics, those laws, haven't changed to a part
in the 10 to 4 over that period of time.
Now, then the maj-- the basic properties of the universe over that period of time,
the universe since then has been relatively well behaved
and it's been conditions like we see in the laboratory.
You know, I think the kind of things that Lee thinks about when he's thinking
about the law is changing, he is imagining happening earlier.
He's imagining happening in the Big Bang itself and before and before.
So you're looking at, you know-- so yeah, we do ask, but you can imagine the theory
where it's changing then might predict some later changes,
but doesn't have to, so we look for it.
And if we saw evidence for it, that would be for the interesting.
And we hope to, you know, one of the things we will do, you know,
and motivated by just these kinds of questions, is go and look and say,
"All right as our data improves, would we see--
do we see evidence for the laws of nature changing in time?"
So, one of the things I find so exciting about being a cosmologist is you can take a lot
of these questions that are, you know, really very philosophical
and about our beginnings and things.
And some of them you can work them to a point where there are predictions that you can test
with data and go out and measure it, see what's there.
Yeah? Okay.
[ Inaudible Remark ]
Or, right or most immediately as we walk over you could--
I will be walking after I answer these question and happy to answer them as I walk and talk.
I can do both at the same time.
>> I just wanted to pick at your data a little bit.
>> Okay.
>> The power spectrum data that you had with the really high multiple moments up the 3,000?
>> Yup.
>> So it looked like the Antarctic data,
the red dots were slightly higher than your other comma--
>> Yeah.
>> -- points.
Is there a reason behind that?
>> Yeah. There's a reason behind it and a bit of a story I didn't get into,
which is when you look at the data, you have-- in addition to the fluctuations we saw,
you have this radial sources, you have those bright spots
and they're producing additional fluctuations on small scales.
Now the brightest ones, we just removed.
But there's always dimmer ones that we can't see with our telescope.
And those dimmer objects contribute fluctuations.
They don't have the same-- they actually have nothing to do with the microwave background.
But with our data we can't distinguish between a few radial sources and--
>> So would you say that you're data is clearer than the--
>> No, it's just the aversion of the data.
Their data is very high quality.
So, what we can do-- oh.
[Inaudible Remark] So, what I didn't show-- I have it here.
I can show you a more complete version of the data.
That might be here.
[ Pause ]
Well, this is the version of this talk with equations.
[Laughter] If I find a better version, a more complete version of data.
Ah, no [inaudible].
All right, I'll just graph what our data looks like.
Amplitude of fluctuations versus-- equal L, so it's 180 degrees over angle
and our data looks like-- or at least our theory looks like that.
And the database simply follows that.
You'll have a really high L, you get a contribution that looks like that from sources.
And to make life easy, I stop the graph there, but we actually have data going up here.
And we spend a lot of time trying to figure
out for what we measure here what the contamination is here.
And another thing we do, so I showed on frequency, we measured that the sky
at two frequencies, the experiment in South Pole measures at three frequencies.
How does the amplitude depends on frequency?
We're taking data with European satellite called Herschel.
We'll look at the same region, we're looking at that data, we're looking at some data.
The same dusty galaxies, they're kind of interesting in--
you know, in science, one man's annoying source is another woman's object of study.
These dusty galaxies or where most--
the galaxy formation or star information in the universe are taking place,
so they're really pretty interesting in and of themselves.
So they become kind of really interesting objects to study and we try to characterize that
and understand that and then see what's going on here.
[ Applause ]