MALE SPEAKER: Our guest today
is a renowned astronomer, and a member of the National
Academy of Sciences, with award-winning
accomplishments documented in about 700 research papers. As the Richard and Rhoda
Goldman Distinguished Professor in Physics and Physical
Sciences at UC Berkeley, he's been voted the
best professor on campus a record nine times, and
was named the Carnegie-CASE National Professor of the Year
among doctoral institutions in 2006. In 2011, the Nobel
Prize in Physics was awarded to the
leaders of two teams, for the discovery
of the accelerating expansion of the universe. Those two teams had one member
in common, and here he is. Please welcome Dr.
Alex Filippenko. DR. ALEX FILIPPENKO: So, I'm
very pleased to be here today. This is my first time
at the Googleplex, and it's been fun
wandering around and seeing your neck of the woods here. What I'd like to describe
today is the discovery of the accelerating
expansion of the universe. And the reason that
I wanted my slides, I mean I can tell the 5
or 10 or 15 minute version without any props whatsoever
to various people, but since this is
a Google audience, I thought I'd
actually use graphs, which is something you can't
do for the general public, of course. But I wanted to give you some
of the inside scoop of how the work was done. So if we could have
the next slide, please. This is unfortunate because I've
got 100 slides or something, so you're going to have
to listen to my cues. I'll say Next, all right? Go ahead, next. This is going to
happen really a lot, so you're not going to be
able to read the newspaper or something while I'm
giving my talk, OK? Thank you so much. Anyway, in 2011 the
Nobel Prize in physics was given to these
three gentlemen. The Nobel Prize went to the
leaders of the two teams that discovered the accelerating
expansion of the universe. Saul Perlmutter, at the Lawrence
Berkeley Lab and the Department of Physics at UC Berkeley,
was the head of the Supernova Cosmology Project, and I was
associated with that team initially. And then there was Brian
Schmidt's High Redshift Supernova Search Team, with
which I was later associated. And then my
post-doc, Adam Riess, who did a lot of the work
in the mid- to late 1990s, he actually made the
measurements of the data when he was a
post-doc in my group. And he was the first author
on the Schmidt team paper, and I'm glad that he was
awarded the prize as well. It's unfortunate
that the prize can go to only three
people, at most. This is unlike the
Peace Prize, which is handled in a
very different way. That's the rule, or at
least the tradition. But these gentlemen understood
that without the other 48 of us working in the trenches,
the discovery would never have been made. So they spent a good
fraction of their prize money flying the rest of us out to
Stockholm in December of 2011 to participate in Nobel Week. And that was just an amazing set
of celebrations and parties-- everyone just sort of
cheering for science. It was like the
Swedish Super Bowl. Anyway, if you could go
to the next slide please, you can see the High Redshift
Supernova Search Team right after the awarding
of the gold medal to Adam Riess and Brian Schmidt. And the rest of us like to
say that besides the handshake from the King of
Sweden, the gold medal, and a share of the $1.5 million,
we experienced everything else that Nobel Week was all about. So we got most of what
there was to offer. But it was really a lot of fun. If we could go to the
next slide, please. The story starts of
course with Edwin Hubble, who discovered the
expansion of the universe. Next, please. He examined galaxies, which
are these giant collections of hundreds of billions of
stars, gravitationally bound together. And he devised a way to
determine their distances. I'll discuss that
method in a few minutes. He also looked at their spectra. Next, please. And the spectra were
all red shifted. That discovery was actually
made by Vesto Slipher some one decade earlier. So these various spectral
lines, due for example to neutral sodium and hydrogen
and singly ionized calcium, the whole pattern was
shifted to the red. And he found-- next,
please-- that the relatively nearby galaxies had small
red shifts for their spectra. And-- next slide, please-- the
more distant galaxies, which typically appear smaller
and fainter in the sky had bigger red shifts. And this can be interpreted
in terms of a motion. And you've all heard the audible
Doppler effect, of course. But in fact, we
now know-- and this is something that Hubble
himself resisted for awhile-- we now know-- next
slide, please-- I'm just showing you a graph
of a low red shift galaxy, in fact one that's very near us. It's moving away very little--
zero kilometers per second is very little. And then one that's
moving away at a tenth of the speed of light. And you can see the same pattern
of lines, but shifted over. OK, next please. We now know that this red shift,
the cosmological red shift, is caused by the
expansion of space itself. So it's not as
though the galaxies are moving through some
pre-existing space, like bullets flying
around outside. Rather, it's a stretching
of space itself. And we can tell by, for
example, looking at the surface brightness of galaxies, looking
at the brightness per unit area, it turns out that the
brightness per unit area, as a function of
red shift, differs if you have a truly expanding
universe versus a universe through which
objects are moving. And the data agree
with the expansion of space interpretation for
the cosmological red shift. Next slide, please. So from our perspective,
here we are in the middle and all these other galaxies
are moving away from us with a speed that's
proportional to their distance, because the stretching of
space is greater for the more distant ones than
for the nearby ones. And there's something a bit
strange about this diagram as I've drawn it. What's weird about it? Yeah, we're in the middle. Why would that be? Do these other
galaxies not like us? Do we smell? Is it something we said? Or are all these other
galaxies lactose-intolerant? Get it? Milky Way galaxy,
lactose-intolerant? Anyway. When I tell my students
at Cal about cosmology and the expanding universe,
I say, what is it? Are we from Stanford
or something? With due apologies to the
many of you who I'm sure are Stanford alumni, or even
on the faculty like Peter. It's truly an outstanding
institution, it really is. Just not quite as
outstanding as Cal. But anyway, we don't think
we're in any central location. Next slide, please. This is a property of any
uniformly homogeneously expanding universe,
such that any galaxy, any raisin thinks
it's at the center because the dough is
uniformly filled with yeast, and you let it bake for an hour. And let's say it doubles
in size in that time. You see all these other
raisins going away, but your friend
sees the same thing. And by the way, the
universe is either infinite or it wraps around
itself, so don't worry about the
finite size here. So there is no center, at
least not in the dimensions that we can physically probe. There may be a
mathematically valid center. For example, here's a
two dimensional universe. It's closed. The laws of physics in
this hypothetical universe only act within
the rubber itself. And the creatures
in this universe could figure out the
equation for their universe. R is a constant. Theta and phi go through 0 to 2
pi and 0 to pi, so R theta phi. But they only live on a
two-dimensional surface. Or x squared plus y
squared plus z squared is a constant square
of the radius. Again, three variables, three
spatial dimensions, but they only live in two of them. The center of this
balloon is the center of this expanding universe. That's in a mathematically
describable, but physically inaccessible dimension. So we may live in a three
dimensional version of this. We actually don't know the
true shape of our universe, so we're not sure. Next slide, please. In any case, we're not
in any central location. So here's what I never show
to the general public-- a graph of Hubble's law,
the recession speed. Oh, you've got one. Great. Thank you so much. Let's see if it goes backwards. Ooh, it does go backwards. Goes forward, too. Thank you so much. Thank you. It's good-- there's a
time symmetry in physics. At the microscopic level, if
you reverse the arrow of time, everything looks
the same, which is why entropy and the arrow of
time is such a bizarre thing, right? But anyway, here's Hubble's law. The speed of a galaxy caused
by the expansion of space at any given time is
proportional to its distance, with the constant
of proportionality being given by
Hubble's constant. And it's a constant
only in the sense that it's the same everywhere
in the universe at a given time. The value actually
changes with time, so the knot here just means the
constant at the present time. And you get the
recession speed just by multiplying the speed
of light by the red shift. OK, so that's Hubble's law. And with today's
telescopes, we've measured the current rate
of expansion pretty well. It's, for those who care, of
order 70 kilometers per second, per megaparsec. Megaparsec is about a 3 and
1/4 million light-years. But we expect the expansion
rate to change with time. And this is quite obvious. The universe has galaxies in it. All of them are going to
be pulling on each other. That should slow down
the rate at which the universe is expanding. This is a clear prediction
in general relativity. It's true also in
Newtonian gravity. If I toss the apple, the
mutual gravitational attraction between the Earth and
the apple slows it down, brings it to a halt. Eventually it comes back. So in the case of the universe,
if the density of the universe is sufficiently high,
then the slowing down of the rate of expansion
will be quite large. And eventually the universe will
stop and reverse its motion. So it'll collapse in on itself. It will implode. So you could start
with a Big Bang, with a Big Crunch,
a dense hot state. Or you could say,
Big Bang, Gnab Gib, which is Big Bang backwards. OK, so in that kind of a
universe, in an empty universe, you have no slowing down. But if you have a dense
universe, then the separation between two arbitrary
clusters of galaxies-- the so-called scale factor-- it
increases, but at a slower rate and then reaches some maximum
height, then comes back down. That's of course a graph
of this apple's motion. So in that kind of
a universe, you'd be looking at the galaxies,
you'd be lying on your back. They're getting fainter
and smaller with time. You would say it's
a good universe. And then you would notice
something a bit peculiar. And right around
now, you'd start getting a little bit nervous. And then it would be,
oh, goodbye cruel world. So that's what the
universe would do. That would be the
fate of the universe if we live in a dense
universe, a universe where the average density
of gravitating matter is greater than some
critical density. This is known as omega matter. So if omega matter
is greater than 1, you have this critical
universe-- overly critical universe that
collapses on itself. But you could have
a universe that doesn't have quite
so much matter. And it would slow
down the expansion, and it would asymptotically
approach 0 as time approaches infinity. So this would be a universe in
a sense like the apple thrown at a speed equal to the
escape speed from the earth. It keeps on going away forever. It doesn't quite ever stop. And then you could have
a low density universe, which is analogous
to an apple thrown at a speed greater
than the escape speed. So it keeps on going away
forever, asymptotically approaches some non-zero
positive velocity. And most measurements
of the matter density suggested that we live in
a universe that's only 30% of this critical density. So if that's the case, and
you're lying on your back and looking up at
the galaxies, they'll keep on getting fainter and
smaller forever, all right? And that's a very different fate
compared with the Big Crunch. This eternal expansion
is a very different fate. And cosmologists, those
who study the structure and evolution of the
universe as a whole, want to know what the fate
of the universe will be, just because. All right. Well, you could measure the
average density of matter, compare it with this
critical density, and figure out what
universe you're in. But that's a hard thing to do,
because matter comes in clumps. It comes in these clusters. And the clusters have
dark matter in them. And then there's
voids that don't have much material in them. So it's hard to get a
representative volume of the universe. And moreover, you're not sure
you're seeing everything, even gravitationally you're not
sure you're seeing everything. So another technique
is to examine the past history
of the expansion. And I like to go back
to the apple analogy. If I measure the apple's
speed at many different times, and its trajectory,
then I can see how much it's been
slowing down, figure out whether it will ever stop
and implode on itself. All right. In a similar way, if we
measure the rate of expansion in the past at many
times, and compare it with what it is
right now, we could predict the fate
of the universe. The way to do this is
to essentially look back into the past and see
which of these curves the universe has been following. Clearly it's not
an empty universe. So this is just an
extreme approximation. But the universe, if what other
astronomers are telling us, should be behaving
something like this. We should see the
universe following that kind of a trajectory. So let's focus in on the
time period before now. We're 13.8 billion years
into the universe right now. So let's look at this, OK? Now, the definition of the
redshift is the following. A given red shift is simply
the size of the universe now, divided by the size
of the universe, or the scale factor, as it was
when the light was emitted. So redshift 0-- well, 1 plus
0 by advanced mathematics is 1, that just means we're
looking at the universe now. A red shift 1 means that
we're looking at the universe back when essentially
all distances were half of their present value. So arbitrary
clusters of galaxies were separated by half of their
value, of their present value. So here we go. There is now, t equals now. There's redshift 0. Suppose we look at redshift 1. Suppose we look at
galaxies at redshift 1. All right. Then you'll notice that
the lookback time-- that is, how far back in
time we're looking-- depends on which
universe we're in. Take a look at this here. In a dense universe,
we're looking back in time some amount,
some billions of years. And the distance is the speed
of light times the time. So if we measure
the distance, we can figure out
the lookback time. In a less dense universe, the
lookback time-- the distance-- is bigger. In an even less dense
universe, the lookback time and the distance at a given
redshift are even bigger. In an empty universe,
for a given redshift, you're seeing as far back
as you could possibly see. So we expect, then, looking
at different redshifts, we expect to measure
different distances, or different lookback
times, and the universe should be-- the data should be
following one of these lines. And the observer's view of
this theoretical diagram is the following. At small redshifts, you
just have Hubble's law. That just means that the
slopes of all these curves are the same right now. That's just-- the universe
is expanding right now, with whatever rate
it's expanding. That's just a given. And two clusters of galaxies
have whatever separation they have right now. That's just a given, too. That's why all these models
have to converge at now. Their slopes are equal, and
their separations are equal. So the slopes differ very
little at small redshifts. That's what you're seeing
down here, Hubble's law. But as you go to
bigger redshifts, you start seeing deviations
from Hubble's law. At a given redshift--
I'm sorry, I hardly ever use a laser pointer anymore--
at a given redshift, you can see the dense
universe-- well, that galaxy, is at a smaller distance than
in a less dense universe. And that's a smaller distance
than in a less dense universe. The empty universe, well,
that's where the distance is the biggest for
a given redshift. So the procedure is clear. Measure the
distances of galaxies having a wide range of redshifts
and see which of these curves the universe has been following. Various observers told us
we would find this result. But we wanted to see. So how do we do that? Well redshifts are easy to get. You just take spectra of
these little blobs here, all these are little galaxies. There's only a couple of
stars in our own Milky Way galaxy in the Hubble
ultra-deep field. So you take their spectra,
and that gives the redshift. That's easy. The distance is a
little bit harder. We will define,
what I will call it, luminosity distance
in the stock. That's simply the
distance of an object, d sub L, such that the
light that you see, the given brightness b,
follows the inverse square law. So that's just a
luminosity distance. There are many
distances in cosmology, and you've got our
remain consistent. So you need something
of known luminosity. You need to know L, and you
measure the brightness b. And that allows you to
derive the distance. The measurement of the
brightness is easy, but how do you know the luminosity? Well, you need what astronomers
call a standard candle, or a standardizable
candle, something that is always the same,
like a 100-watt light bulb. And it's just like the
judging the distance of an oncoming car at night. You've calibrated how
bright the headlights of a car of known distance are. And you look at cars whose
headlights are fainter, and you figure out
their distance. If you're not very good at
doing this almost intuitively, you shouldn't be
driving at night. So Hubble did this
with a type of star known as a cepheid variable. And you might think, oh
it's a variable star, so that's no good. But it turns out that there
are certain categories of cepheid variables. And the point is, you know
what their true luminosity is. And let's just look at
that one right there. Let's call it Mike,
just for kicks, OK, because he introduced me. What if you know that Mike
is just like Betelgeuse, the left shoulder of
the great hunter Orion? Magnificent, luminous
star Betelgeuse is. It's relatively nearby. We know its distance. We know its apparent brightness. We know the luminosity. We know its oomph. If we then look at Michael
in that other galaxy there, we can figure out its distance
from the inverse square law. And then if we get the same
answer with star Vikram or something like that, we
start gaining confidence that we're using
the right technique. So this is what Hubble
used to determine the distances of galaxies
in the mid-1920s. And he was the one who
showed what many suspected, and that is that these
spiral nebulae are far, far outside our
Milky Way galaxy. And our view of the universe
expanded by a gigantic amount with that realization. Well, these galaxies are so
faint and fuzzy and distant that you can't see individual
normal stars within them. But it turns out, there
is one type of star that is visible,
even from distances of billions of light years. And that's the supernova,
an exploding star. Very few stars explode at
the end of their lives, but those that do are
critical to our existence, because we are made of the
carbon, oxygen, calcium, iron that get spat out of
exploding stars like this. So we owe our existence
to this phenomenon. Anyway, they become very
bright, in some cases-- up to a few billion
solar luminosities. So if our sun were to do
this, then sunblock of 50 just wouldn't cut it, folks. You'd need sun block,
or supernova block, of a few billion to
protect yourself. But don't worry, be happy. Our sun isn't going to do this. So the name of the game is to
find some of these supernovae in galaxies whose
distances we already know, because we can see relatively
normal stars like Mike and Vikram in them. So we know the distance
of this galaxy. We measure the apparent
brightness of the supernova. That allows us to determine
its true luminosity. But it's hard to find
these supernovae, because they occur only once
every few decades in a given galaxy. So if I were a
really cruel adviser, I would have each of
my students looking through the eyepiece
of a telescope at one and only one
galaxy, preferably at night-- you see
more galaxies at night than during the day-- until
that student finds a supernova. Then we let them graduate and
move on to greener pastures. Meanwhile I will have had
decades worth of slave labor out of said student. Well, if there are
some crimes that are so egregious that even
a tenured professor can and should get fired, then
that would be one such crime. Of course, I could
have my students look at thousands of galaxies. These are random,
independent phenomena. So if they look
at more galaxies, you'll find more supernovae. But that would be considered
cruel and unusual punishment as well. Fortunately, with
modern technology, we have a better way. We attach CCD cameras to the
eyepiece end of a telescope, take photographs of
thousands of galaxies, and then simply look for arrows. And where you see an arrow,
you see an exploding star. You see once, twice, three
times, four times, five times. By the process of rigorous
mathematical induction, I conclude that this
must work every time. So obviously, it's
not that easy, otherwise we wouldn't give
degrees for this kind of work. What we've done
is we've developed a robotic telescope
at Lick Observatory, just a two-hour
drive from Berkeley, or a little bit over an hour
from the Googleplex here. In fact, Mike
showed me the view, which was a bit hazy today. Couldn't quite see it today. But I'm sure many
of you have seen it. This is not a big
telescope, but it's been programmed to look at
lots and lots of galaxies. And it's one of
several telescopes at Lick Observatory,
Mount Hamilton. Go and visit it
one of these days. Great place to go. And my close
associate, Weidong Li, programmed this telescope to
look at nearly 10,000 galaxies a week, and to automatically
compare the new pictures with the old pictures. And through the magic
of digital subtraction, here's the template. Here's the new image. This is, by the way,
a negative image. These aren't all black
holes, otherwise black holes would be easy to find. Anyway, so here's a
supernova candidate. And here's something
that the software thought was a cosmic ray, just a charged
particle that hit the detector. And here's a poorly
subtracted star. The image quality varies
with time a little bit, as mirrors point in different
locations, their shape changes. So you get maybe a few
dozen candidates per night. And when you're observing a few
dozen-- 1,000 galaxies a night, you only expect maybe
a supernova candidate every three or four nights. So most of these are going
to be not real things. They could be cosmic rays, or
maybe an asteroid with Earth's name written on it,
hurtling toward us. One person's garbage is
another person's gold. Someday, astronomers
and engineers will save humanity by
finding the asteroid that's 30 years from hitting us. And then they'll go
and they'll deflect it. So anyway, we have maybe a few
dozen candidates per night. And then I use slaves-- I mean
undergraduate students-- who, with their superior
eye/brain combination, examine these
candidates, determine which ones are worthy
of further investigation because they're likely
to be supernovae. And I'm very proud of my team,
because I get students, even in their freshman and sophomore
year, hands-on experience with real data analysis. And occasionally I've even
had high school students, and they write home to Mother
when they discover a supernova, and it's just really great,
because they get very jazzed. And most of them
don't go on and become professors of astrophysics. That's probably good. There's enough of
us in the world. They go on to become
engineers, computer scientists, applied physicists-- people
who are more immediately useful to society. But the hook was the
cool stuff in the cosmos, and the research experience that
they get as part of my team, and other teams at
Berkeley and Stanford and other great institutions,
is really invaluable. And something I don't
need to tell you guys, but I think we are losing our
technological edge in the US. And we have to shape up, or
we will be quickly passed up. Anyway, so-- what's that? A supernova? Brightens over maybe three
weeks, fades over a few months. I'll do more
questions afterwards. I realize some of you
will need to leave, so I'm kind of
rushing a little bit. But I'll be very, very happy
to answer more questions later. But yeah, I sped up the process
and that animation so as to not bore you. So we find them, and then
it's important to take their spectra. And so we collect the
light, for example, with the 3-meter telescope
at Lick Observatory. And I have undergraduate
and graduate students involved heavily in that
stage of analysis as well. And the spectrum tells you
what kind of a supernova you're looking at. There are different ways in
which different kinds of stars can explode. And only some of them are
luminous enough to be useful, and standard enough. If you've got light bulbs
of a huge range of lumens, that's not going to be so good. The ones that are useful for
cosmology are the so-called type 1a's. You have this roller
coaster of a spectrum. There's the calcium
in your bones, the oxygen that you breathe. This is real life stuff. We are made of
star stuff, right? So type 1a supernova
is useful, because it comes from a very
weird type of a star. Our sun will become a white
dwarf in about 7 billion years. But fear not, it won't blow up. A white dwarf is a
weird sort of matter called degenerate
matter, not because it's morally reprehensible,
but this is simply the term quantum physicists
give to a very, very highly compressed state of matter,
where all the electrons are basically in their own little
cubbyhole, pigeonhole, energy level. And they can't occupy
the same energy levels and the same
quantum states, and so they exert this
weird degeneracy pressure on each other. It's truly magnificent. It's unlike any sort
of thermal pressure. Anyway, if you have a white
dwarf in a binary system that's sufficiently tight,
then that white dwarf can steal material
from the other star and approach a mass
which becomes unstable and it undergoes a
thermonuclear runaway. And it happens at about
the same mass each time, 1.4 solar masses. And the thermonuclear
runaway occurs in about the same way each time. And it's very luminous-- four
billion solar luminosities. So these are excellent and
pretty standard candles. And they can be
seen very far away. But we want to look at
them in great detail, because it turns out they're
not perfectly standard. So here are some
spectra that my team took a little over 20 years ago. And we started noticing-- we
took these spectra at Lick-- we started noticing that the
spectra of these things are not all identical. So here's a pretty
normal one, 1990N, and there's this line here. But you see it's weaker
in that supernova. And then in the
supernova it's stronger. And then this supernova also has
this weird absorption feature, which is due to
titanium, it turns out. So they're vaguely similar,
but there are differences. Here there's some clear
differences as well. So not all type
1a's are the same. That turned out to be a
very important realization in the early 1990s. We can't treat them all as
being identical objects. And moreover, these
guys, it turns out, are on the whole subluminous. These are overluminous, compared
to the more or less vanilla flavored type 1a supernova. So if you don't take
that into account, and you're looking at some
distant type 1a supernova, and you don't know
the true luminosity, you will get an incorrect
distance, and hence an incorrect lookback time. So it's totally important
to take this detail into consideration. Not all the headlights
are the same, simply put. And you've got to know which
kind you you're looking at. Well, the next realization
was that, oh wow. Turns out the luminous
type 1a supernovae have slower light curves. That is, they rise and
decline more slowly than the subluminous ones. And the normal ones are
sort of right in here. So if you find a whole
bunch of nearby ones whose distance is known, and
you calibrate this relationship, then you can look at the
distant high redshift ones, measure the light curve
and say, aha, this is a 94-watt light bulb instead
of a 100-watt light bulb. Or maybe it's a
115-watt light bulb. So instead of
saying, well, they're all 100 watts with a
dispersion of 50 watts, you can say no, this one's 94
plus or minus 13, or something like that. You really improve the precision
of the distance measurements, and hence of the lookback times. And this is what really allowed
the type 1a's to become useful in cosmology, and to ultimately
lead to this amazing discovery. Well, this relationship
here was first pointed out by Mark Phillips and
then Mario Hamuy in Chile gathered a whole bunch
of type 1a supernovae and showed, or
sort of calibrated the relationship even better. And then my former post-doc
Adam Riess, who at the time was a grad student
at Harvard, used their sample, the
Chilean sample, to calibrate type 1a supernovae. And then he took a new
set of type 1a supernovae. OK, this is a different
setup here, in the top graph, and he applied this machinery. Here we have the supernovae,
if you treat them as a standard candle-- 100
watts plus or minus 50-- you see all kinds of
dispersion here. Moreover, there are
some huge outliers. When you take into account this
luminosity light curve shape relationship, and
when you also take into account dust and
other interstellar debris, through which
the light is going-- that I won't go into
the details of-- you decrease the
dispersion to one third of what it used to be--
15% rather than 45% or 50%. And these outliers go right
down to where they should be. So the technique works well. So that's what
gave us confidence that we could use type
1a's to do cosmology. And although my robotic
telescope did not contribute to the
original sample that was used to define
this relationship, it has been used since that time
to refine this relationship, and decrease our resulting
scatter more and more. So we're calibrating these
things better and better with time. And in fact, one of my most
recent graduate students, Mohan Ganeshalingam
did a thesis on this. And here you can see
the various light curves that we have collected
with Kate and you can see very clearly that
the more luminous guys have slower light curves than
the less luminous ones. And so here is Mohan's Hubble
diagram without any correction. It just looks like
crap, sort of. I mean, it's pretty bad. And you can't do precision
cosmology with that. You can tell the
universe is expanding, but you can't do
precision cosmology. Whereas when you apply our
machinery, look at that. I mean, that's fantastic. This is a relatively
nearby sample, but this is the
kind of confidence we need to get before going on
and doing this great redshifts. Now, it turns out
that Wayne Rosing, who used to be Vice President of
Engineering here, has gone on and started this thing called
the Las Cumbres Observatory Global Telescope Network. He's got a network
of little telescopes all over the world,
with which we can follow variable
stars and exploding stars throughout the
24-hour day/night cycle. Their motto, I think,
is the sun never rises on the Las Cumbres
Observatory Global Telescope Network. So Wayne, partly
based on the success we had with our Kate
telescope-- and in fact, he very generously
helped fund that telescope-- he has now
globalized this whole thing and is pursuing one
of his passions. So it's fantastic. He's a good friend of mine. Anyway, now that we've gained
some confidence with the nearby ones-- and it's 2 o'clock, but
you'll permit me to go longer, right?-- let's go and do
the actual cosmology now. Let's find these things
in distant galaxies. Now, let me remind you,
there are two reasons you want to find the
distant supernovae. The first reason is, by
the observed brightness compared to the luminosity,
you get the distance and hence the lookback time. And from the spectrum,
you get the redshift. So the supernova is critical
for two reasons-- a distance through the inverse square
law, and a red shift. So, we find these things with
a wide-angle cameras attached to relatively big telescopes--
this is in Chile, Cerro Tololo Inter-American
Observatory-- and modern, wide-angle CCD cameras actually
are these gigantic things. It's amazing what the
silicon revolution has done, not just for
computers, of course, but for astronomical
imaging technology. We now routinely have these
16k by 32k arrays of pixels. And this is about the
size of the full moon. Of course, the full
moon isn't rectangular, but this is roughly the
size of the full moon. And in this picture,
there are literally thousands of galaxies. Nearly every blob
you see is a galaxy. So if you take
many such pictures over the course of
a couple of nights, and then repeat the
process three weeks later, you will have taken the mug shot
of perhaps 100,000 galaxies. And in those three weeks,
some of those 100,000 galaxies will have produced
a type 1a supernova. Indeed, several dozen
will if you do the math. So we did that, and through
digital subtraction, here's a small subset
of one of those images, taken on the 7th of April, 1997. Subtract that from
the 28th of April, you get a bunch of noise. That's OK. Any measurement
process, as you know, necessarily has some
noise associated with it. But here, cleverly placed
in the middle of the square, is something that looks
like it might be real. And here's a Hubble picture
of it a few weeks later, and it's marked with
an arrow, so it's got to be a type 1a supernova. I wish it were that easy. You don't know that
it's a type 1a, and you don't know
it's the redshift until you take a spectrum. So this was my main
job on both teams. I'm a trained spectroscopist. I've been studying supernovae
since the 1980s, the mid-1980s, and I have access to the world's
biggest optical telescopes, the twin Keck 10-meter
telescopes in Hawaii. And I urge you to visit
those, if you've not done so. They're an amazing
thing to look at. Here's Fred Chaffee, a
former director of Keck, just showing you the size
scale of these gigantic eyes that are gathering
light from afar. So with the Kecks, I could
get spectra of these type 1a supernovae that at redshifts
of, say, 0.455 in this case. This is a lookback time of
about five billion years. So we're seeing this thing as
it was five billion years ago. And lo and behold, the
spectrum to within the noise looks very similar to a low
redshift type 1a supernova. So that's really great. So here's the punchline. Here are several of
these type 1a supernovae that have been verified
through spectroscopy, and whose redshifts
have been measured. And the punchline is
that they are faint. They're very, very faint. And you might say, well,
they're in these scrawny, pathetic-looking galaxies that
are obviously very distant. Here you can't
even see the galaxy before the supernova went off. Obviously, these are
distant galaxies. So the supernovae
should look faint. And that argument is
correct, but the point is, they look fainter than
they had any right to be. For a given redshift, you
expect a certain range of brightnesses,
depending on what the universe has been doing. And the measured brightnesses
were fainter than expected. Let me show you that. Here are the possible
models that we expected. Indeed, people told us that we
would measure probably this. a Theorists actually
prefer omega matter of 1, for reasons I can go into in
the Q&A. But what we measured, when we measured distance
luminosity, distance versus redshift, was this curve. That wasn't-- in
a multiple choice, that was "none of the above." And if you follow the natural
progression of omega matter, this ratio of the average of the
true density of the universe, critical density, greater
than one, 1.30 negative, negative matter density,
which seems a bit odd, because we exist. And last time I checked,
I'm gravitationally attracted to the earth, rather
than being spirited away like on [? SpaceEx
?] or whatever. So here's a picture from
Adam Riess' lab notebook, when he was in my group
in the fall of '97. He has this little thing here. Omega matter is negative .36. Oh, gosh, I hadn't
taught him well yet. It should be negative 0.36,
because the decimal points are easy to lose, right? I hope you've all
been taught that. But, anyway. So there it is, negative matter. Well, the negative
sign is simply indicating that we're witnessing
an accelerating universe rather than a decelerating universe. The wrong sign, right? So this is kind of weird. Well, I was actually
privileged to announce this result for the
Riess/Schmidt group for the first time
at a meeting in LA. And here's the
headline that came out. "Astronomers See a Cosmic
Antigravity Force at Work." We use this term
"antigravity" hesitantly, because people ask us, can we
attach this stuff, whatever it is, to our cars and levitate
over Bay Area traffic jams? And the answer is no. It's either a property
of space itself, or there's so little of it
that it'll never be harnessed. Never say never. But anyway, by December of '98,
the editors of Science Magazine proclaimed this to be the
single most important discovery in all areas of
science that year. And we were obviously
very pleased by this, but of course we're not
yet sure that it's true. But they said, well, look. Other physicists
and astronomers have had the better part
of a year to find an obvious flaw in
what you've done, and no one has found
an obvious flaw. So either you're right or you're
wrong for some subtle reason that will end up
teaching us something interesting about the universe. And indeed, of course that's
generally how science operates. Now, the caricature
of Einstein looks surprised here, not
because he's blowing multiple universes
out of his pipe. You might not have known that
there are multiple universes, and they come from the pipes
of theoretical physicists. Well, the first
statement might be true. And I'll be happy to
come back and tell you about the possible multiverse. But he's surprised
because this one universe is expanding faster
and faster with time, rather than more
and more slowly, as one would expect in
normal general relativity. And he's doubly
surprised because he has a sheaf of papers here where
there's an equation-- lambda equals 8 pi g,
Newton's constant, times the density of the vacuum. And you might wonder, who's
this bozo from Berzerkely talking about the
density of the vacuum. You were taught on your mother's
knee that the vacuum is sheer emptiness-- zero, zilch, nada. How can it have a non-zero
energy or matter density? Well, this was Einstein's
idea, not mine. And he was a lot
smarter than I am. So the point is this. Back in 1917, when he
developed the general theory of relativity, its
basic attribute-- that is, gravity pulls--
remains the same as in Newtonian gravity. So the universe should be
collapsing in on itself, because all the galaxies
are pulling on each other. Or, maybe the universe
was born with a big bang. And in that case, the
universe should be expanding. But in any case, it should
be a dynamic universe, yet there was at
that time no evidence for a dynamic universe. People thought it was static. And Einstein himself felt
that the static model was very aesthetically pleasing. So he conjured up something that
was not aesthetically pleasing, by his own admission. The cosmological constant,
which in this case, it repels. It's the opposite of gravity. So if you have something pulling
down and something pulling up and the net force is zero,
you can get a static universe. Or, if the universe started
in an expanding state, then you could have a negative
cosmological constant. And that could
nullify the expansion. So you could get a static state. Turns out this is an
unstable solution. You can't have a static universe
for very long in this case, because if you perturb this
galaxy a little bit inward, it turns out that the
gravitational force increases, and it continues to go inward. And conversely, if you perturb
it outward a little bit, then it turns out it
continues to go outward. So it wasn't a very satisfactory
solution in many ways. It also implied that the
density of the vacuum is non-zero, which seemed weird. There was no experimental
evidence for that. And moreover, something
that seemed so finely tuned like this seemed ad hoc. It seemed arbitrary. It did not make the
equations wrong. The cosmological
constant can be thought of as a constant of integration. So that in and of itself
isn't so weird, other than that it's a vacuum
energy that's non-zero. But even if there is this vacuum
energy, why the world should its value-- why in
the world should the value of the constant
be tuned to exactly match the attractive force of gravity? It just didn't seem right. So Einstein never liked it. He reluctantly included it in
his solutions for the universe. A dozen years later,
Hubble discovered that the universe
isn't static after all. So the whole physical and
philosophical motivation for this cosmological constant,
this ugly fudge factor, disappeared. Einstein renounced
the idea, supposedly, as having been the biggest
blunder of his career. Because had he not introduced
it, he could have been famous. He could have predicted that the
universe is in a dynamic state. So here he is, sad that he
ever introduced the idea. I don't know that that's
what he's thinking, but it might be
what he's thinking. What have we done the better
part of a century later? We've reincarnated the idea. Not to give a static
universe, but one which, on the largest
scales, beyond about 100 million light years, is
accelerating in its expansion. So here in this room,
the down arrow dominates. Everywhere in our solar
system, it's down. Everywhere in our galaxy, down. But as you get to distances
of 100 million light years or more, the up arrow
dominates and you have an accelerating universe. And so Einstein's biggest
blunder, as he referred to it, may have been, in some ways, his
greatest intellectual triumph, conjuring up such a thing. And if he were around right
now to see the evidence that we and others have amassed,
maybe his reaction would be something like what
was in Science Magazine. So we now interpret this
omega matter less than 0 to instead be the cosmological
constant is positive, or something like it. It's not necessarily Einstein's
cosmological constant. It could be something
having a similar effect that on the biggest
scales, it leads to this acceleration,
which is kind of weird. Anyway, you might
worry that, well, OK, this is only out to four or five
billion light years in time-- I'm sorry, four or five
billion years in time. What would be the
predicted effect if we were to look
back farther in time, if something like the
cosmological constant were really operating. And there the idea is
actually pretty simple. The curve looks
something like this. The universe decelerates
originally, and then after a while
starts accelerating. And the reason for
that is that long ago, when galaxies were
closer together, their gravitational
attraction for each other was bigger than it is now. And the repulsion, if
it's a property of space, or something within space,
was relatively minor. As the universe expanded,
the gravitational attraction declined with time. It's a 1 over r squared force. But the repulsion
increased with time. If the repulsion is caused
by something in space, or a property of space itself,
than the more space there is, the greater is the
repulsive effect. So in a sense, gravity
is coming downward. Antigravity, if you
will, is going upward. And at some point, they cross. And that's where the
universe starts accelerating, right around here somewhere. But the clear prediction was
that if we look far enough back in time, we should see the
phase at which the universe was decelerating, if
we found something like the cosmological constant. So Adam Riess became
principal investigator of a Hubble project that we
initiated to find and monitor very distant supernovae-- seven,
eight billion, nine billion, or even ten billion
light years away. And we found that
indeed, the data followed the predictions
of early deceleration. So we witnessed the phase
of early deceleration, and then about five
billion years ago, that transition to acceleration. A transition from
deceleration to acceleration-- that's mathematically
known as a jerk. In fact, a jerk is any time you
have a nonzero third derivative of position. So you have position,
velocity, acceleration, jerk. I actually didn't know that
until we made this discovery. So in a sense, we
witnessed the cosmos as having gone through a jerk. And the headline that came
out in the New York Times was "A 'Cosmic Jerk' That
Reversed the Universe." And there's my former
post-doc Adam Riess. So I start getting
these phone calls-- hey, who's this jerk
you work with who reversed the expansion
of the universe? Anyway, this isn't
the greatest photo. And Adam's mother was not
pleased by this juxtaposition. Right? You never read the
articles in newspapers, unless you're really interested. You go flipping through them,
you look at the headlines, look at the pictures. And only a small subset
do you have time to read. Well, so then what is this stuff
that's causing this effect? It's clearly not the visible
matter of the universe, because all normal
visible matter pulls. It's also not antimatter. You might think
antigravity, antimatter. Good try, but antimatter
has a positive gravitational attraction. It's also not dark matter. Many of you have
heard of dark matter. You can see it
there, there, there. Lots of dark matter. It's what binds galaxies and
clusters of galaxies together. And I can't help but tell
you a small aside about one of my heroes, Fritz Zwicky,
an astrophysicist at Caltech who died just
before I got there. Actually, a few years
before I got there. He was the first to look
at clusters of galaxies and realize that the individual
galaxies are moving around so quickly that there's got be
sort of additional material there, holding the
whole thing together. Otherwise, they'd
go flying apart. And yet we see lots of
clusters of galaxies. They're not just chance,
superpositions of galaxies passing through the night. So he suggested that there might
be such a thing as dark matter. And he was routinely
ignored, but he was decades ahead of his
time on a number of issues. One reason he was
perhaps ignored was that he was somewhat
arrogant and abrasive. Somewhat is being a
little bit polite. And he did not have
a great opinion of the intellectual capacity
of his colleagues at Caltech. And, you know, Caltech's
a pretty brainy place. So they didn't take well
to this implied criticism. And here, he may
be showing you what he thinks of their
typical brain size. I mean, I don't know that
that's what he's thinking, but he's on record as having
referred to his colleagues as "spherical bastards." Because they're bastards
anyway you look at them. And of course, a sphere
is the only object that looks the same from
all directions, right? I would not recommend that
you refer to your friends as spherical bastards, or
your Google colleagues here. You will quickly end up
friendless, or colleague-less. But anyway, I like Fritz a lot. Anyway, so dark matter
helps bind clusters of galaxies, and even
galaxies together. Indeed, it's critical
to the formation of clusters of
galaxies and galaxies. We wouldn't be here if it
were not for dark matter, even though we're not
made of dark matter. But the acceleration has to be
something entirely different. It has to be something that
pushes in a sense, but not enough to disrupt galaxies
and clusters of galaxies. So the name that was given
to it, for better or worse, was dark energy. It's dark, we don't see it. It's also mysterious--
in that sense, it's dark. Yet it's clearly
some form of energy. But the reason a
lot of us are not entirely happy with that
term is that if there's one equation that even people
on the street often know, it's e equals mc squared, or
at least they've heard of it. So we're forever being asked,
are dark energy and dark matter just two different
sides of the same coin? And the probable answer is
no, even though we don't yet know what the dark energy is. Anyway, in honor of the
Nobel Prize and stuff, Noelle, my wife, who's
here, made this t-shirt, "Dark Energy Is The New Black." And she gave it
to the rest of us, who did not get a share of the
one and a half million bucks. But you gave it to the
team leaders, as well. And she even got one to the
King of Sweden, we think, if the courier gave it to him. So what is the dark energy? The honest truth
is, we don't know. They're really hundreds
of possibilities. The possibility I
like most, and the one that's still consistent
with the data, is that the dark
energy is essentially the zero point
energy of the vacuum. If you have quantum
fluctuations, which we know occur-- they affect
the energy levels of the electrons in a
hydrogen atom, for example. That's the Lamb shift. Then you have this energy and it
can cause an expansion of space faster and faster if there's
a positive net energy. But theorists had
always said that, well, for every positive
energy fluctuation, there's a negative energy one. In the parlance
of string theory, these are thermionic and
bozonic fluctuations, it turns out, if you
want to look them up. And people had always assumed
that they all balance out, and the energy of
the vacuum is 0. And the main reason
for thinking that is that the back of the
envelope calculation suggests that
vacuum energy should be 10 to the 120th power,
which is 20 orders of magnitude bigger than a real Google. Or I shouldn't
say a real Google, the original Google, right,
is ten to the hundredth. The vacuum energy should be 10
to the 120th, 10 to the 122, I think, if you want
three significant digits. And that wouldn't
even allow us to form. We wouldn't be here. Stars wouldn't be here. Galaxies wouldn't be here
if that were the case. So people had always
assumed it's 0. But if it's not 0 by
some weird circumstance that we don't yet
understand, then it turns out space itself has the
desired property of expanding faster and faster with time. It's a very weird thing. And the data are consistent
with this interpretation. A lot of theorists don't
like this interpretation. They think it might be some sort
of a new field, a little bit like the Higgs field
that's been in the news so much the past year or two. And in fact, a whole category
of models of this order known as quintessence, like the
Aristotelian fifth essence. You know, earth, air, fire,
and water, and the quintessence out there. And there are hundreds,
if not thousands of candidate theories. And the problem is,
all of them have additional bells
and whistles that are not super well
physically motivated. And the data do not agree
any better with those ideas than they do with the simple
cosmological constant. So I'm betting on the
cosmological constant, but I may well be wrong. What we're doing now
is trying to trace, in greater detail, the expansion
history of the universe to set observational constraints
that will be used to rule out some of the candidate theory. So that's what we're
busy doing now. The other possibility,
other than dark energy, is that general
relativity is wrong. For completeness, I do
need to mention this. Most theorists don't
like this possibility, but it's not yet
completely ruled out. So maybe there
isn't a dark energy. Maybe instead, general
relativity is wrong. But that would be a
pretty exciting thing too. Because we really do
think general relativity is very beautiful, and
has an amazing theoretical underpinning. Finally, you might worry that
all these conclusions are just based on supernovae. I mean, that would
be pretty weak. I mean, what if there was
something different about them long ago. They used to be less
luminous than they are now. We have to look for
that possibility. And there are many things
I'm leaving out of this talk. That's why they pay us
the big bucks, right? Yeah, right. So anyway-- in academia. So to figure out
the details, right? They pay us the big bucks. So in science, as you all
know, the more important the discovery is,
the more important it is to verify it through
completely independent techniques. And so because there
were two teams that made the same announcement
at virtually the same time, a lot of physicists and
astrophysicists took note. They started testing
this conclusion. And let me just briefly
tell you about some of the other tests which lead to
the same conclusion, basically. One is by looking at the early
afterglow of the universe, the cosmic microwave
background radiation. You can set lots of
interesting constraints on the properties and
constituents of the universe. And in particular, here's a map
I'm sure all of you have seen. It's a baby picture, an infant
picture of the universe, where the temperature
of the universe, to a good first approximation,
is 2.7 degrees Kelvin. But at the one part
per 10,000 down to one part per
100,000 level, there are small variations
in temperature. That's what you're seeing here. They correspond to small
variations in density. That's good, because it's from
those variations in density that galaxies and
clusters of galaxies were able to
gravitationally form. So if the universe
were completely smooth from the beginning, we wouldn't
be here talking about it. So from studies of the angular
sizes of these little freckles, you can tell that the
universe, over large distances, is Euclidean. It's flat. The sum of the interior
angles of a triangle is always 180
degrees, for example. And in general
relativity, that can only be the case if the
density of the universe is the critical density. And yet, we know the density
of normal matter is only 30% of the critical density. So in a sense, you need
70% of something else. That's consistent
with the supernovae. The something else
being dark energy. The other point is, you take
those little fluctuations, and then you let gravity do its
thing over billions of years. And gravity, that
great sculptor, sculpts the galaxies and
clusters of galaxies and voids. And when you do numerical
simulations, starting with this picture, which is
a map of the variations when the universe was
380,000 years old, and let gravity do
its thing, and you don't include dark energy,
the final distribution of galaxies and clusters
of galaxies in voids differs, in an observationally
significant way, from the observed
structure of the universe. Whereas when you include
dark energy in the mix, then the final structure
ends up looking like the observed structure. And this is the
observed structure, based on measurements of
literally millions of galaxies. So that's another
piece of evidence that the dark energy
really does exist. And there are other pieces
of evidence as well. So that's why we really
believe it's true. We think that the
composition of the universe looks something
like this right now. Dark energy is 70%
of the total pie. And we don't exactly
know what it is. In fact, we really
don't know what it is. Dark matter is most
of the remainder, and we're not sure
what that is either. We think it might be
little particles left over from the big bang. WIMPS-- weakly interacting
massive particles-- but they've never been
detected in a laboratory. I'm becoming mildly
concerned about that. The ordinary matter is only
5% of the total contents. And the easily visible ordinary
matter is only half a percent. So in a sense, we're the
debris of the universe. We're the afterthought
of creation. That's not to say you're
not important to yourselves, your loved ones, your
families, your friends. But you're not made of the
dominant stuff of the universe. The dominant stuff
is the dark energy, followed by dark matter. And we don't know what they are. So for the kids-- I go
and talk to kids sometimes about this-- I tell
them, if anyone tells you physics is dead and
there's nothing left to be learned
about the universe, you tell them, what about the
origin and detailed nature of 95% of the contents
of the universe? So this is one reason
cosmology remains so exciting. We don't know what
these things are. And moreover, dark energy may be
one of the few observable clues that will allow us to
discard some of the ideas for quantum gravity, of
which string theory is a generic umbrella. Quantum mechanics works
great on small scales. General relativity works
great on large scales, but the two together are
in violent disagreement. We are after a quantum
theory of gravity. Any quantum theory of
gravity that does not account for the acceleration
is not a viable candidate. So this is one of the
few observational tests we have of quantum
theories of gravity. So because of its importance,
it was recognized with the Nobel in 2011, and also because it had
been confirmed in so many ways. We don't know what
the dark energy is, but we are pretty
sure at this point that the universe really is
accelerating in its expansion. So here's a brief
history of the universe, just to close things off. We think it began with some
sort of a quantum fluctuation, maybe out of nothing. And then there were
quantum fluctuations within this rapidly
expanding universe that led to small variations
in the distribution of matter, which then grew over time in a
universe that was decelerating. And then in the last
five billion years or so, it had
started accelerating. And recent data have
led to an update of the age of the universe from
13.7 to 13.8 billion years. And it did not take
us 100 million years to gather those data. It's just that-- anyway. So the universe aged 100 million
years in the last year or two. So that's the history
of the universe. What will be its future? How will the universe end? Well, if the dark energy really
is the cosmological constant, it'll never go away. It will continue
to be repulsive. The universe will expand
faster and faster with time. A runaway universe, I call it. So if you want to see
a galaxy like this with your very own eyes
through a telescope, go up to Lick
Observatory, or Chabot, or Morrison Planetarium, or
your local astronomy club, and look through a telescope
soon-- within the next few tens of billions of years. Because beyond that
time, all those galaxies will be whisked away
to such great distances that we will no longer see them. There's a possibility,
of course, that the universe
will recollapse, because we don't yet quite
know what the dark energy is. But I think it's going
to expand forever. Robert Frost didn't know
of these two possibilities when he wrote his famous
poem Fire and Ice. I think it was actually based
on something in Dante's Inferno. But in retrospect, it's
very relevant to what I've told you today. The poem goes
something like this. Some say the world
will end in fire. Some say in ice. From what I've
tasted of desire, I hold with those who favor fire. But if it had to perish twice,
I think I know enough of hate to say that for destruction,
ice is also great and would suffice. So Frost would prefer
the recollapsing universe that ends up hot and compressed. But if the universe and
he had to perish twice, then eternal expansion
would be OK with him. And that's perhaps appropriate
given his name, Robert Frost, right, and ending in ice. Well, anyway, this
research couldn't have been done without the
help of many, many people and federal foundations. I'm very, very grateful to them. I'm also grateful to
all the opportunities I've had with like the Hubble
Space Telescope, the Keck Observatory. And in particular
Lick Observatory, because this is where
a lot of the data for the nearby
supernovae was gathered. And this is the
observatory where we really train our students
and post-docs. And they can be
principal investigators of their own projects, rather
than just helping me do things that I define from
start to finish. So they grow in their
intellectual independence by using these facilities. And my colleague Jeff Marcy has
discovered lots of exoplanets there. So I'm very much a supporter
of Lick Observatory. I'm actually currently president
of the board of an organization called Friends of
Lick Observatory. And we're actually
trying to raise money to support Lick
Observatory, in part because the University
of California is highly financially
squeezed right now and has told us that
we're doing great science, but they can simply no longer
support the Observatory anymore. So if you'd like
to join Friends, let me know or you know
which search engine to use to find a website
where you'll find this. And if you want to
support a place that's cutting edge in its technology
and stuff, let me know. I'm also happy to
arrange a visit for Google employees
and their friends. It's a great place to go. You can look through the
old 36-inch James Lick refractor at whose base
James Lick is buried. That's a wonderful experience. And at any time, if you guys
want to set up a Google Chat and ask questions of me, or just
ruminate about the universe, I'm happy to do that
kind of thing as well. So there I am. You can find me on Google+. And it's been great visiting. I hope there'll be a little
bit of time for questions. But thank you all for coming. AUDIENCE: So in your
graph, you have dark energy as 70% of the content
of the universe, Dark Matter is 25%--
but 70% of what? What is [INAUDIBLE]? DR. ALEX FILIPPENKO: So
let me go back to this pie. The dark energy is
70% of the universe. 70% of what? The pie is the whole thing. Whatever the density
of the universe is, OK, that's the whole pie. Now, that density
we now know was very close to the critical
density between what we call a positively
curved universe, a hypersphere-- like the
three dimensional balloon-- and a negatively
curved universe. Hyperbolic space. Closest two-dimensional
example I can think of for you is a horse's saddle. So we're right on
the dividing line. We're a flat universe
over very large distances, ignoring black holes
and stars and all that. Light travels along
Euclidean straight lines. So it's 70% of that density. But the point is, is this pie
could apply to the universe, regardless of what
its density is. It's just, here's the whole pie
and there are its constituents. AUDIENCE: [INAUDIBLE]? DR. ALEX FILIPPENKO:
It's energy en masse, through e equals mc squared. It doesn't really matter which
one you're talking about. So this has me and you in it. That's in the form
of ordinary matter. But it also has
something that we think doesn't have
a matter component. It is a pure energy,
sort of like light is a pure form of energy. Has zero rest mass. But dark energy is not light. It's just-- it's a different
form of pure energy which has what we call
negative pressure. And in general relativity,
it's the negative pressure that causes the acceleration. AUDIENCE: So I've been
somewhat confused for a while, I guess, about the
distinction between things in space-- galaxies and
whatnot moving through space and apart from each other--
as opposed to space itself expanding. And so you're describing
this expansion that causes redshift and
space itself expanding. But then you're talking about
the gravitational attraction that would cause deceleration,
pulling things back together. That sounds like it would be the
object in space pulling things together. DR. ALEX FILIPPENKO: So
that's a good, fine point. It turns out that in
general relativity, you can consider a uniform
distribution of matter. That's the simplest universe. And the self-gravity
of the universe itself causes the
fabric of the universe itself to slow down
in its stretching. This is actually
part of the framework of general relativity. That, in other words, if we had
found that the universe isn't behaving this way, we
would have been surprised, based on general relativity. Now, within the universe, you
can also have pulling motions. So the Andromeda galaxy,
our nearest big neighbor, is in our local cluster, our
local group, it's called. It is moving toward us. In fact, we're going to merge
in about four billion years. And after a couple
of billion years of looking like
a train wreck, we will turn into a different
type of galaxy altogether, an elliptical galaxy. So those are motions
through space. They're called peculiar motions. And they're caused by
individual galaxies, or clusters of galaxies, pulling
on each other. Nevertheless, for the
universe as a whole, you can treat that matter as
being uniformly spread out and ask yourself,
what effect does it have on the universe as a whole? And you get these different
amounts of deceleration. Or in the case of dark
energy, acceleration. Nevertheless, there are peculiar
motions within the universe, as well. It's like the raisins
tugging on each other, and moving through the dough. That's happening as well. Yes. AUDIENCE: So the space between
us and the Andromeda galaxy is probably still
expanding, although-- DR. ALEX FILIPPENKO: Yeah. The space between us and the
Andromeda galaxy is expanding, but the problem is that the
gravitational attraction is so great that it
greatly dominates. But yes, strictly speaking,
if there were no dark energy, we would be attracting that
Andromeda galaxy a little bit more than we are. You're absolutely right. But it's a negligible effect
within a cluster of galaxies, because the matter
density is so much greater than this dark energy density,
whatever the dark energy is. Yeah. AUDIENCE: If space
itself is expanding, how does that relate to
the microwave background? DR. ALEX FILIPPENKO: Yeah,
if space itself is expanding, how does that relate to
the microwave background? AUDIENCE: Did the
temperature [INAUDIBLE]? DR. ALEX FILIPPENKO:
The microwave background is the stretched electromagnetic
radiation from long ago. So when the universe
was 380,000 years old, its temperature had dropped
to about 3,000 or 4,000-- about 3,000 degrees Kelvin. At that point,
electrons combined with protons for
the first time ever, a process interestingly
known as recombination. Because plasma physicists
can ionize gases and then they watch them
recombine in their labs. But it's happening
for the first time. At that point, the
universe becomes transparent to radiation. Because you don't have
all these free electrons that can easily scatter
electromagnetic radiation, OK? At that point, 3,000 degrees,
well that's about the sun. It's half the sun's temperature. So the light waves back then
were optical and near infrared. They have stretched
a factor of 1,000. In fact, the stretching factor
since that time is 1,079. We now know it quite precisely. So the waves have stretched
by a factor of 1,079. Their typical
wavelength right now is in the microwave
region of the spectrum. They preserve their black body
spectral energy distribution. It's just that now that
black body corresponds to 2.7 degrees, which,
you will realize, is 3,000 divided by 1,079. So the microwave background
is the microwave background because of the stretching. It's the wavelength that it
has because of the stretching. AUDIENCE: [INAUDIBLE]? DR. ALEX FILIPPENKO:
So the temperature is going through a
jerk in the sense that the universe is expanding
faster now than it would have in the absence of dark energy. So indeed, the temperature
of the universe is going down
faster than it would have in the absence
of dark energy. Yeah. So the temperature, in a sense,
went through a jerk as well. Yeah AUDIENCE: Is that detectable? DR. ALEX FILIPPENKO:
Is it detectable? Very good question. I believe the answer is that
in principle it's detectable. But it has not
yet been detected. The way to detect
it is interesting. And indeed, there were
detections of the microwave background long before
Penzias and Wilson. It turns out that certain clouds
of gas in interstellar space had cyanogen molecules--
CN-- as well as others. And there are certain
transitions in the cyanogen molecule that are excited by
microwave wavelength photons. But they're excited more if
the temperature of the universe were slightly higher
than the current 2.7. So astronomers looked at
these clouds at redshift 1, when by the way the
temperature would have been 5.4-- 1 plus z times
the current temperature-- and lo and behold, the
transitions were clearly stronger back then, because
the absorption lines produced by those clouds of gas were
stronger than would have been the case had the temperature
back then been 2.7. So in fact, we've measured not
only the microwave background, but changes in its temperature. What has not been
done is studies of enough molecular clouds
at a variety of redshifts to see a transition in the rate
of change of the temperature. But in principle, I see no
reason that couldn't be done. That is very cool. You're the first-- I
can tell I'm at Google. You're the first
person ever, ever, to have asked me that question. So I feel like I'm on a
thesis exam or something here, where you have to
answer these kinds of things really well. Yeah. Other questions? Right there, yes. AUDIENCE: How do we
know the basic nature of the cosmic microwave
background that is coming from a
blast from the past, as opposed to, say,
the void glowing? DR. ALEX FILIPPENKO: Yeah, yeah. So various people
have tried to think of ways of reproducing
the microwave background radiation through local sources. And in fact, iron
whiskers, turns out, radiate in a way that
has some good properties. You can't get a perfect
single temperature background if you've got objects at
different redshifts glowing, due to whatever process. You'd need to somehow
finagle them all to have a slightly different
intrinsic spectrum, such that when you add
them all together at these different redshifts,
all their contributions, they end up giving
you a black body that is better than the thickness of
the pen with which you draw it. That's how good the data are. I mean, the measurements
from Coby and Planck and the W map satellites
are absolutely perfect black bodies. And so we've thought of
no way of doing that. And yet, if the
universe did begin in a hot compressed state, for
which we have other evidence-- primordial abundances
of hydrogen, deuterium, things like that
as well-- then you would expect this afterglow. Indeed, George Gamow and
his students and post-docs predicted it. So it's a natural prediction
of an expanding universe. And it's enormously
difficult and ad hoc to conjure up in any other way,
is what I would say, I guess. Yeah. AUDIENCE: So probably a naive
question about redshift. So the light waves
generate long ago because of the stretch of the
space got longer light wave. DR. ALEX FILIPPENKO:
That's right. AUDIENCE: Wavelengths. But when the space is
stretched, doesn't the rule for example, the way we measure
length, also get stretched? DR. ALEX FILIPPENKO: Yeah,
the lengths of the rulers are not stretched. Indeed we're not expanding. I mean, you might
after a big lunch, but that's your fault,
not the universe's fault. You're held by
electromagnetic forces. And again, it's a
spring that's so tight that you can't stretch it. We're held to the
earth by gravity in a way that's much greater. It easily overcomes the natural
tendency of space to expand. So even our local group of
galaxies is not expanding. Indeed, within it, Andromeda
is coming toward us. So if Hubble had only looked
at the Andromeda galaxy, he would have concluded that
the universe is collapsing in on itself. The expansion of space
is only visible starting, oh, 4 or 5 million
light years out. That's where the
density of material and the strength of
gravitational fields and things like that become low enough
that the stretching of space can start taking over. But we and the rulers
are not stretching. They're held together
by a spring that's so tight that you
can't stretch it. AUDIENCE: Or can I say that
actually the stretch is there. It's just too small--
dominated by the other forces. DR. ALEX FILIPPENKO: Yeah,
the stretch is there. That's another way of saying it. But it's so overpowered
by the other forces that it's not noticeable. I'm very careful
to say it that way, because even in
some textbooks, you read the incorrect
statement that, oh yes, of course our
galaxy is expanding. But because our galaxy is only
100,000 light years across, if you put 100,000 light
years into Hubble's law, you get a negligible
amount of expansion. That is a wrong explanation. You cannot put 100,000 light
years into Hubble's law. Because that would imply
that our galaxy is stretching with the rest of the
universe unimpeded. And that's not true. It is being impeded
by its local gravity. So just watch out. There's that subtle difference. And I didn't want
you to walk away thinking that it's only
that our galaxy is small, and therefore you can't
notice the stretching over a distance of only
100,000 light years. It's more than that. It's that our galaxy impedes
the stretching because of its self-gravity. AUDIENCE: Light is
redshifted by the expansion, then it seems like that light
is going to be losing energy. Does that energy go somewhere? DR. ALEX FILIPPENKO: Ooh! Boy oh boy. The light is indeed
losing energy. That's right. Light has energy,
and therefore in fact has a gravitational effect. In fact, you could have a
black hole whose interior consists entirely
of light, before it went into the singularity
and got eaten. But you can imagine creating
a black hole out of light, because light has energy and
therefore it bends space-time. So in fact, what happens is,
in an expanding universe, the light itself-- forget about
anything else in the universe-- the light itself is
impeding the expansion. The light is doing
work on the universe. And that is one way of thinking
of the redshifting of light. It is losing
energy, because it's doing work on the universe. Isn't that a cool thing? That's an interesting insight. Yeah. Very cool. And the universe is pretty
cool too, at 2.7 degrees now. MALE SPEAKER: So,
we're out of time now. We need to get Dr. Filippenko
to his next meeting. So thanks again. DR. ALEX FILIPPENKO:
Well, thank you so much. Thank you.
