- Cosmology is the study of the nature of the universe itself. Arguably the most fundamental
question astronomy can probe, how big is the universe? How did it begin? What is the universe law? It's a field that can humble
the most narcissistic of kings. One that inspires personal transformations from nihilism to humanism and one that in recent
years has been claimed to be facing a crisis so severe it threatens to undermine
our very physical models of the cosmos. I am of course, speaking
of the Hubble tension, the so-called crisis in cosmology. Like any active era of research, this is one where the
needle is rapidly moving as new analyses and data are pouring in on an almost daily basis. Yet, despite all that, we
do have enough information in hand now to say that something
appears to be very wrong. So today, let's break
down what is this crisis and what might it really mean? No sensationalism, just the facts. The story really begins in the late 1920s when observations of
distant galaxies reveal that the further away a galaxy is the faster it appears to
be moving away from us. Both Edwin Hubble and George Lemaître independently realize that
the universe must be expanding in order to explain this. Like stretching a piece of
rubber with marked points. Every point thinks that every other point is moving away from it
during the expansion. And indeed the points furthest
away seem to move faster. In the early 20th century, the rate of expansion was
difficult to reliably infer, but over the years astronomers
have improved their ability to measure this expansion rate, which has become known
as the Hubble constant. The Hubble tension
refers to a disagreement about this value. In particular, this
number has been measured with two independent methods, which seem to give answers
which are sufficiently different that they're difficult to reconcile. Either one of these methods is wrong or there was something
wrong with our understanding of the cosmos itself. The first way we can
measure the Hubble constant is from the cosmic microwave
background, or CMB for short. This is the earliest light
we can detect in the universe emitted at a time when the cosmos was less than 400,000 years old. So that's 13.8 billion years ago. At first glance, the CMB
appears remarkably uniform with a universe appearing to
have an average temperature of about three kelvin in all directions. But subtracting off that average, the patterns and inhomogeneity
that are revealed encode information about
the shape, composition, and structure of the universe. Remarkably, just about
everything within the CMB can be described by a physical model using just six free parameters describing the mixture
of matter, dark matter, and dark energy as well as
the shape of the universe and its expansion rate. The agreement of this model to the data is frankly remarkable with the recent Planck mission revealing almost perfect agreement to
the model across the board. By running this cosmological
model forward in time one can predict what the current rate of expansion should be. Over time, we've got better
at making that measurement with early results from the WMAP satellite giving a result of about
72 plus or minus five, but now more precisely pinned down to 67.4 plus or minus 0.5 using
the Planck mission. I'll ignore the units here because it's really not that important for the current discussion and largely just distracts
from the main points. The real thing to take away here is that all of these measures
are within era of each other and look consistent with just improving that precision over time. So if this cosmological model is correct, which remember it is a remarkably good job of explaining the CMB data, then we really should expect that the modern universe
around us is expanding at its predicted rate. But of course we don't
just wanna trust the math. Astronomists actually wanna
double check that number. So we've been going out taking
the local universe around us and measuring its expansion rate to make sure that everything agrees. These methods largely build
upon what Edwin Hubble was doing almost a century ago, measuring the distances and
velocities of distant objects and putting them on a graph
to get the expansion rate. The devil is in the details though for measuring the distance to objects is particularly nuanced and challenging when dealing with
objects that are millions of light years away. If I want to measure the
distance to a nearby object, the way our brains do it is with parallax. Hold your finger out of arm's length in front of some distant
background like say a city skyline. And if you close one
eye and then the other, your finger will seem to
move from side to side relative to that background. That's because your eyes are in physically different locations and so the angle subtended from each eye to your finger is different. The closer your finger
is held to your eyes, the bigger the shift. However, as we move our finger away, the shift becomes less. So in this way, the parallax
shift gives us the distance to an object and we do the
same thing in astronomy, except the spacing between our eyes is replaced with the orbit of the earth from summer to winter. The earth physically shifts
by 300 million kilometers over six months. The Gaia spacecraft is a
master of measuring parallax and it has now done so
for millions of stars, but the further away an object is, the smaller its parallax will be eventually becoming so small
that we can't detect it even with Gaia. To get roundness, astronomers
need a new distance measure, a distance proxy and so that's where a special
type of star enters the story. One called a cepheid variable. All we really need to
know is that cepheids are massive stars that regularly pulse, causing their brightness to periodically increase and decrease. These are very bright objects detectable from across intergalactic space and it was Henrietta Leavitt who discovered that
their pulsation periods were intimately related
to their brightness. Leavitt turned cepheids into
so-called standard candles, which are invaluable to cosmologists. The basic idea is that you
first measure the distance to a nearby cepheid using parallax, which practically speaking means it has to be warm within the Milky Way in order for that technique to work. Step two, very simply you just measure how bright the cepheid
appears here on earth. Step three is that you calculate then how bright the cepheid must truly be given its distance using
the inverse square law. And step four, you observe
its pulsation period. Doing this for all the
nearby cepheids you can find, you can make a graph of
the cepheid luminosity versus pulsation period giving
this tight relationship. And that relationship known as the period luminosity relation or simply the Leavitt Law is kind of like your decryption key for measuring distances
outside the galaxy. Because having calibrated this law, we can now go out beyond the Milky Way, take a cepheid pulsation period and then immediately know
what its luminosity must be. Combine that with its apparent brightness and you can figure out
how far away it must be using the inverse square law. So it's distance. And that distance is one
of the two basic quantities we need to know in order to calculate the local expansion rate of the universe. So you can see that this
technique is very powerful, but eventually it stops working for galaxies beyond about a
hundred million light years. After that, the selfies are just two faint for us to detect in the first place. Now that might sound like a huge number, but remember the universe is about 90 billion light years across. So we'd really like to go much further. A key thing to keep in mind is that this technique
is wholly predicated upon correctly calibrating the Leavitt Law within the Milky Way galaxy. In this way you can think of it as being like the second rung in a ladder, which depends upon that first rung, which remember is correctly
calibrating that law. In fact, we could even go to a third rung, which depends upon the second rung, and in this case that third
rung is type 1A supernovae. Like cepheids these are
thought to be standard candles, objects with standard luminosities. But unlike cepheids, we've never seen one up close and personal to really know what that
standard luminosity actually is, which is probably a good thing because those events are so violent that a local instantiation
would be undesirable to say the least. And so in the absence
of any local examples, we are forced to figure out the luminosity of these type 1A supernovae
using the second rung of that distance ladder. So the cepheids. In some galaxies we see type 1A's go off and in those same galaxies
we also see cepheids. That means that we can use the cepheids to measure the distance to that galaxy, which of course means that
that must also be the distance to our supernovae. And then we can figure out how
luminous the supernovae was. So type 1A supernovae are
the third rung on the ladder and they are so bright that
we can look far further out into space out to distances of now up to a billion light years away. But like a house of cards, if any of the layers below are wrong, then our calculated distances
of the type 1A supernovae will also be wrong. And that in turn means that if we use them to calculate the Hubble constant, that too would be wrong. It's a precarious game. If you want to measure the Hubble constant to a precision of say 1%, and that means that you need to calibrate each of those distance slider rungs to at least 1% precision
or really even better. The chain is only as
strong as its weakest link. So with that intro to the
distance ladder complete, what does it actually tell
us about the expansion rate of the universe? Hubble was constant. A team led by Nobel laureate Adam Riess who call themselves SHOES have been leading an effort to measure the Hubble constant
of the local universe. Similar to what we're
seeing with the CMB value, the precision has been
increasing over time thanks to more precise parallaxes, refined analysis techniques, and more observational data, which by the way includes
over 1500 hours of time on the Hubble space telescope. As before, the fact that
the value gets more precise really isn't surprising
and each improvement agrees with the previous
estimate to within error. So far so good. But the problem is when we
put the CMB measurements back on the same plot. We see a growing discrepancy between them. That right there is the Hubble tensions, the crisis in cosmology. So something is very weird here. One of these measurements must be wrong. If it is the local universe one, then that implies that our distance ladder is most likely in error. There must be some
systematic calibration issue with what we did there. But if it were the CMB one that was wrong, then that would imply that
our cosmological model of the universe is itself flawed. That's perhaps the most
exciting of these two options. 'Cause it would mean that
there was new physics beyond the current model. So what could this new physics be? There's been plenty of
exciting suggestions such as a new particle
ignored by the existing model like the staron neutrino for example. Another popular idea is that
the strength of dark energy is growing over time, which would have some
important consequences for the fate of our universe
in the distant future. Others have suggested decaying dark matter or modified theories of gravity, non-zero spatial curvature,
or exotic interactions between dark energy in dark matter. In fact, as you can see
there are quite a few options on the table. But before we start ripping
up the physics textbooks and getting too excited, there is of course a far more
mundane possible explanation and that is that our local
measurement is wrong, that perhaps the distance
ladder has been miscalibrated. This isn't just idle speculation. There are good reasons to be worried about the validity of the distance ladder. First, although the Leavitt Law wasn't even initially considered to be a simple one-to-one relationship between pulsation period and luminosity, we've since discovered
that it also subtly depends upon color and metallicity. That's frustrating because
there's actually only a few dozen cepheids in the Milky Way, so it's gonna be difficult to precisely measure
these new dependencies from such a small sample to begin with, as well as of course, challenging to measure those same properties for other galaxies down the road. Metallicity is a bit
particularly challenging topic because for cepheids, you
can't even often measure their metallicity directly. Instead, we rely on the
surrounding environment's oxygen abundance as a proxy. The dependency of
luminosity upon metallicity is characterized by this coefficient in the Leavitt Law equation. It's almost like the exchange rate going from dollars to Yen. If you want to know the value of your Yen, then you of course need
to know the exchange rate to make an accurate conversion. The problem is that astronomers don't really agree on this coefficient as shown by the range of recently published values shown here. And these differences can
introduce discrepancies in the Hubble constant
at the percent level. Now, there is some hope that JWST will provide a firm answer here thanks to its improved
spectroscopic capabilities, but for now we may have to wait. Another problem is frankly unavoidable. Cepheids are massive stars, which means they are young stars because massive stars
simply don't live very long. So why is this a problem? It's because stars are typically born in dense stellar nurseries that are nestled within
their galactic disk. This is the New York City of the Galaxy, a busy, bustling, dense,
overcrowded region filled with dust and
dirt all over the place. That dust though, is a major problem. The dust blocks out some
of the cepheid's light, and so we have to carefully correct exactly how much light was blocked out. If we assume the wrong
amount of dust though, then we'll get the wrong
luminosity for our cepheid. And so break our distance ladder. Who would've thought
that something as boring as calibration issues and dust could be potentially responsible for a crisis in cosmology? You've hoped for a revolution in physics and instead you got dust. And indeed, this might sound familiar because it's not the
first time astrophysicists have experienced that sensation either. Remember BICEP2 or Biogen
STAR, dust and dust. So to some degree this is an unavoidable
problem with cepeids. So perhaps one solution is just to ditch the cepheids altogether, to start over, go back
to the drawing board of our distance ladder, and get a new first and second rung. But to do that we would
need to replace the cepheids with something else. We would need a new standard candle. Enter giant stars. Stars like the sun, fuse hydrogen inter helium in their core, but eventually they'll
swell by hundreds of times as they exhaust a supply of
hydrogen fuel in their cores and instead start fusing the hydrogen around the shell of the core. This shell burning behavior
leads to a rapid ballooning in the size of the star, which in turn causes the
luminosity to greatly rise. It's exactly this process
that will eventually doom us all here on earth. Check out our previous video to see how. Eventually the pressure
in the stellar core becomes sufficiently high that
the thus far inert helium ash starts to fuse. This is actually good news because the star is now burning
fuel in its core once again out by helium fuel. And so now the star returns
to a more stable state and thus a smaller size. So if you track the luminosity
of the star over time, you'd see this gradual growth during the shell burning phase, followed by a reversal once
the helium fusion begins. That reversal point, the peak luminosity is called the tip of the red giant branch, and it's largely independent
of the star's metallicity unlike for cepheids. In essence, this reversal
point is a standard candle. It's a weird one though because it doesn't really
work for any single objects since it's unlikely you'd
happen to catch a star at exactly this moment in its life cycle. But if you look at the brightness values of a large number of giants, then you'd see a clear
maximum possible value that corresponds to the
tip of the red giants and that heat is our standard candle. This method is in many ways,
cleaning out of the cepheids. Since unlike the Leavitt Law, there's hardly any
dependency in mentality here, which remember it's very
difficult to measure at cosmic distances. Yet more, another major advantage here is that these stars are
by definition, old stars. For example, it would
take a star like the sun about 10 billion years to
get to this point in total when counting from its birth. Because of their advanced years, they have long since
left their star nurseries and have meandered in
and around their galaxy many times over. In fact, a large number of
them no longer even reside within their galactic disks, but instead have migrated
out to what we call the galactic halo, kinda like the quiet back
country of the galaxy where the spacing between
stars is far greater. This is incredibly useful because out here there's hardly anything to get in the way of our observations. There's no dust, there's
no contaminating stars. If observing extra galactic cepheids is like trying to do
astronomy in New York City, swamped by the haze and city lights, observing extra-galactic giant stars is like being out in the Atacama Desert, clear inky black skies peppered
with crisp oasis of light. Famed astronomer Wendy Freedman has been advancing this method for years, which is not surprising
given that it was she who first used the Hubble Space telescope to measure the Hubble
constant two decades ago using cepheids. Suffice to say Freedman has been thinking about this problem very
deeply for as long as anyone. So her team recently used this completely independent
standard candle method, the giant stars as a way of recalibrating
the entire distance ladder in an effort to finally
resolve the Hubble tension. On the one hand we have the
67.4 plus minus 0.5 valley derived from the CMB using Planck. And on the other hand,
we have the supernovae plus cepheid based
measurement coming in at 73.0 plus or minus 1.0 coming from SHOES. So the real question is where does this new
independent measurement land? Does this third method reveal
who's right and who's wrong? Here's their answer, right
in the middle of the two. 69.8 plus or minus 1.7. Remarkably, it's actually
consistent with both. Whilst the CMB and the
cepheids measurements are five error bars apart or five sigma, the tip of the red giant branch method is effectively okay with both of them. It's 1.3 sigma away from the CMB method, which is completely fine. And 2.9 sigma away from
the cepheid measurement. That's a little high, but we could perhaps just explain that by adding in a little
bit more systematic error into the cepheid measurements. In our upcoming Cool Worlds podcast show, which you can look out
for in the new year, I asked Wendy what her reaction was when she saw that her measurement landed right in the middle. And I guess we would hope we were hoping that you would, your number would land on one side or the other and we'd be like okay, so now
we know where the problem is. When you realized your answer
was right in the middle, were you frustrated by that? How did you feel? - Well, someone, the immediate reaction was sort of like shock. Really? Did it really land in the middle? It was sort of odd. And we didn't attach this
absolute zero point until the end so we didn't know where
this was gonna land. So until we were you
know, ready to apply it and that was it. It was like one morning,
here's the answer. And it, so it was slightly a shock. - An interesting way of thinking about the Hubble crisis is to
imagine an alternate reality, an alternate reality where we invented and deployed the tip of
the red giant branch method before the cepheid method ever came along. So in this alternate reality
we'd have two measurements for the Hubble constant, one from the CMB and one
from the local universe. And both of them would seem to be an almost perfect
agreement with one another. There would be no tension. And instead we'd just be
celebrating the remarkable fact that these two radically different methods gave remarkable agreement about the expansion rate of the universe. So if then cepheids came along later we would likely just dismiss
it as a troubled method, especially given the fact that they are usually embedded within bustling galactic disks. So does this mean that there's
no crisis in cosmology then? Well, before we totally
give up on that idea, I think we still want to
improve our precisions just a little bit more. On the horizon there are
two big data releases from the Gaia Spacecraft that promise to improve our parallaxes even further, and thus will let us
go back and recalibrate that distance ladder
just a little bit better. Yet more, JWST is already
observing nearby galaxies in an effort to measure their distances using both the cepheid
technique and with giant stars as well as other methods too, to hopefully finally figure out whether they agree with each other or not. So rather than jumping the gun too much, let's just see what Freedman's
team find from that sample. You know, a crisis in cosmology
makes for a great headline. So much so that's proliferated across many science news outlets over the last couple of years. It's probably why you clicked this video. And indeed this video would
surely get a lot more views if I told you that physics is broken, or that the Big Bang model is wrong, or some other extreme conclusion. Or showed you thumbnails of JWST randomly firing death rate
out of it or something. Yes, that really exists, but the truth is often
more nuanced, complicated, and often less exciting than that tagline portrays. After all the answer
here may simply be dust, but a headline like dust may be a source of systematic error in
cepheid distance measurements just isn't as catchy as
a crisis in cosmology. I can't give you the answers
that you might want to hear. That's not how science works. To quote freedman, nature is there and she's gonna come
out in the way she is. Our predispositions don't count. All I can ever promise you on this channel is a sober analysis, but looking ahead there
will be answers here from Gaia, from JWST, we will resolve the Hubble tension. So until then, stay
thoughtful and stay curious. Thank you so much watching everybody. I hope you enjoyed this video. If you did, be sure to
like, share, subscribe. It really does make a difference. And if you really wanna help us out, then you can become a donor to my research team, the
Cool Worlds Lab right here at Columbia University. I want to give a moment just
to give a special thank you to all of our donors over
the last year in 2022. And here's looking forward to
an even more spectacular 2023.
