Julien Lesgourgues – TRANSCRIPT
When I was a PhD student, which was actually not so long ago, we only had a vague idea about the history and the composition of our universe. There was still a lot of room for speculation on various ideas.
But since then, we, cosmologists, have experienced a burst of new observations and discoveries over the range of just 20 years. We now understand several details about the history of our universe over the past 13 billion years. And believe me, it has been terribly exciting to be part of a generation of physicists that, at a unique moment in the history of mankind, have been able for the first time, on a reliable basis, to understand what our universe looks like on very large scales. Humans have been making assumptions about that since ever, but the actual reality of our universe has been unveiled to us precisely over the past two decades. And for me, this is astonishing.
So let me illustrate more concretely what I mean when I say that we now understand our universe. For instance, these two pies show, at two very different moments, the cosmic recipe, that is, the composition of the universe in terms of different particles, like atoms, neutrinos, dark matter, et cetera. What is remarkable is that we are able to pinpoint the amount of each of these ingredients with a precision of a percent, despite the fact that most of these species cannot leave any track in our detectors, and despite the fact that the universe is, of course, far too big for sending detectors all around. This is a bit as if by observing a cake and not even tasting it, we could tell its recipe with percent precision.
Second example. This chronology shows that our universe over the past 13 billion years went through four different stages, each of them with extremely different properties. During the first stage, our universe was very dense, very homogeneous and in an ultrafast expansion. During the second stage, the whole universe was extremely hot and bright – a bit like the interior of a star, everywhere – and the expansion started to slow down. During the third stage, everything became cold and dark, and the first galaxies formed, separated by huge regions of empty space. Finally, during the fourth stage, the expansion of the universe started to accelerate again.
Amazingly, we have very strong proofs that each of these epochs really occurred, and we can even date them with high precision. So, the fascinating question that I want to address today is: but how could we understand all these details from our tiny planet? And how can we be so sure, while hundreds of generations before us have tried to reach this knowledge without being able to go beyond the level of assumptions or beliefs? The answer is connected to the word “spectrum.”
We understand all this because we have been able to measure the spectrum of the universe. A spectrum is a quantity used by scientists to describe anything in nature that vibrates, or fluctuates with time, or varies over space, like any sounds, or any lights, or any image. The spectrum is the decomposition of the sound, of the light, of the image into various frequencies or wavelengths.
So, for instance, any sound has its own spectrum. To show it, let me open on my laptop a sound analyzer – that’s it – and just blow this flute. (Flute sound) OK, that’s it. On the right, you see the spectrum of the sound that we just heard, and each music instrument has its own spectrum. Even two flutes or two violins can be distinguished through tiny differences in their spectrum, so the spectrum is really like the fingerprint of a given sound.
But what goes for a sound also goes for an image. Here is a famous painting, “The Garden of Earthly Delights.” This image can be decomposed into a sum of patterns of different frequencies, and the spectrum is just the amount of each frequency or wavelength. So, for instance, here is the spectrum of wavelengths in this particular image. Again, it’s like the fingerprint of the image.
What is fascinating is that we are able to do the same for the whole universe. We can define its spectrum, and, since a few years, we are even able to measure it. So now I want to explain and show you the holy grail of cosmology, the quantity that we were grimming to observe when I was a PhD student: His Majesty the Spectrum of the Universe. But before that, let me recall that when we point a telescope in a given direction, we collect light from various distances, and the farthest that light is, the farthest back in history we can see. And if with the telescope we select the light coming from the farthest possible distance, then we start to see a very special type of radiation, a leftover from the Big Bang, called the cosmic microwave background.
It has been first observed in the 60s, bringing a Nobel Prize. But after this first discovery, scientists asked themselves a question: is this microwave background exactly the same in all directions? And it turns out that it is nearly the same, but with small fluctuations. This radiation can be a bit more intense in a given direction and a bit less in another direction. Since the 70s, cosmologists have tried to establish a map of these fluctuations, and it has been a very difficult task. The first map has been obtained 22 years ago by the COBE satellite from NASA.
On this map, different colors indicate slightly different levels of intensity in the cosmic microwave background. This is a map of the full sky that has been unfolded and projected over an ellipse, like we would do with a more usual map of stars and constellations. The problem with the COBE map was its poor resolution, the fact that it is very fuzzy. For comparison, imagine that we look at the same painting as before, but from very far away with a telescope having a limited resolution. Then, we might see a very blurry image like this one, and in order to understand what the image is really about, we would need to increase the resolution of the telescope like this.
Well, it had been exactly the same with the COBE map. Several experiments have tried to measure it with better resolution. The most recent and precise experiment has been the Planck satellite from the European Space Agency. It has recently produced this exquisitely precise map. The COBE instruments have been already using amazing technology, but to produce such a high resolution and high sensitivity map, it was necessary to use detectors, about 1,000 times more sensitive, so this is really relying on the best cutting-edge human technology.
And from that map, it is possible to get a spectrum. I promised that I would show you His Majesty the Spectrum of the Universe. Well, here it is. These red dots are what we consider as the fingerprint of our universe. It could only be measured very recently thanks to exquisite experiments like Planck.
But in science it’s not enough to have fantastic observations. Alone, this spectrum is not telling us anything about the universe. To make it speak, we need to confront it to a theory. In fact, in science, every time that theories match observations, we learn something new. Actually, over the 20th century, lots of theories have proposed various possible models to describe our universe, relying on their understanding of the laws of physics, but not yet on precise observations like this one.
Later, in the 90s, we started to develop some numerical codes able to simulate the evolution of the universe in each of these models. Actually, one such code is installed on my laptop. To use the code, you need to start from a possible model of the universe, then you press a key, and the computer will simulate 13 billion years of cosmological evolution in only about one second. And after that, it will compute the spectrum of the universe in this particular model. So, let’s play this game together.
I turn on the code, and I start from some assumptions from a given model for the universe with some numbers describing the composition of the universe in that model and some other properties. Then I press a key, and that’s it. The laptop has simulated 13 billion years of evolution and has computed the spectrum of this model. Now we can compare it with the true observed spectrum. We find a complete mismatch, showing that we actually started from a very bad guess.
But we can try other models. And we, cosmologists, have played this game over the last years in a systematic way, trying billions of simulations, and we could establish that there is one model leading to perfect agreement with observations. Let’s simulate it now. I go back to this code, and now I will run with the new assumption in which the cosmic pie is actually exactly the same as we have seen in the second slide I start the simulation, and I get this new spectrum that I can now compare to the true observed one, and the agreement is beautiful.