South Pole Science: Probing the Beginning of the Universe with Microwaves

In this post, I take a break from describing life in Antarctica to share some of the science behind our experiments. Many scientists, our group included, have spent years seeking the answers to some of the most fundamental questions about our Universe. As cosmologists, we want to know what the Universe is made of, how it began, and how it became what we see today. Our telescopes at the South Pole are designed to probe the earliest moments of the Universe’s existence: to learn more about what happened when the Universe was only a trillionth of a trillionth of a trillionth of a second old. Our best ideas of how the Universe has evolved throughout its history depend on understanding this beginning.

The Big Bang Theory

Our current knowledge of physics tells us that everything in the Universe – light, matter, space itself – exploded outward from a single point 13.8 billion years ago. This idea is called the Big Bang theory and is supported by a wide range of astronomical observations, made both by telescopes that can see the same kinds of light that our eyes can (but are much more sensitive) and by telescopes that can see different kinds of light, such as highly energetic X-rays and gamma rays and lower energy microwaves and radio waves. All of the telescopes that my group uses can only see microwaves, so the sky they see looks very different from the one that we see when we look up at the stars at night. A picture of the microwave sky is shown below. Rather than being mostly dark, the microwave sky shines brightly in all directions, with the disk of our galaxy the Milky Way appearing as a glowing band that stretches across the entire sky.

A map of the microwave sky. The Milky Way runs along the equator of the map. The data compiled to form this image was collected by the Planck satellite. (Copyright ESA, HFI and LFI consortia, 2010)

To understand why the microwave sky looks like this, we have to understand where microwave light comes from. Some of it comes from glowing clouds of dust in the Milky Way, or from other “local” sources. However, it turns out that most of the microwave light we see was emitted when the Universe was very young, and has been traveling through space to reach us ever since. This light forms the (nearly) uniform glow that we see in the background of the image above.

With everything jammed together in such a small space, the early Universe was a hot, chaotic place: fundamental particles of matter (protons and electrons) and photons (individual “packets” of light energy) all bounced around together, constantly colliding and ricocheting off in new directions. This early Universe was so hot that it glowed, like a fireball. It was so hot that protons and electrons bounced around freely because they were too energetic to settle down and form atoms. There were no planets, no stars, no galaxies – the whole Universe was a hot soup of particles and energy that looked the same in all directions.

As the Universe got older, it expanded and cooled. Around 380,000 years after the Big Bang it became cool enough that protons and electrons could combine to form atoms. When a photon encounters an atom, it will generally continue straight past without interacting with it. Therefore, after atoms formed, light and matter decoupled and each photon was able to keep streaming in a constant direction, rather than bouncing around as before. Light travels faster than anything else in the Universe, but it doesn’t move from one place to another instantly. This means that the further away we look, the further we look back in time. Light takes several seconds to reach the Moon, so if you bounce a laser off the Moon’s surface it will take several seconds before you see it bounce back. Light takes eight minutes to travel to Earth from the Sun, so when we look at the solar surface we are seeing light that is eight minutes old. When we look at the closest star outside the Solar System, we see light that is about four years old. And if we look far enough away, we can still see the glow from the early Universe, which has been traveling to us for over 13 billion years – the oldest light in the Universe.

A photo of the CMB as seen by the Planck satellite. (The microwave light emitted by our galaxy has been removed, leaving just the light from the early Universe.) The different colors represent tiny variations in temperature but the scale has been exaggerated for effect – one very important early discovery about the CMB is that its temperature is exactly the same to one part in 100,000 (0.001%) across the entire sky.

A photo of the CMB as seen by the Planck satellite. (The microwave light emitted by our galaxy has been removed, leaving just the light from the early Universe.) The different colors represent tiny variations in temperature but the scale has been exaggerated for effect – one very important early discovery about the CMB is that its temperature is exactly the same to one part in 100,000 (0.001%) across the entire sky. (Copyright ESA and the Planck Collaboration)

Just like a hot coal that fades from orange to red to black, the color of this background light changed as the Universe expanded and cooled, getting redder and redder. Today, this light has redshifted so much that most of it is now in the form of microwaves (the same kind of energy used in microwave ovens to heat food). We call this light the Cosmic Microwave Background, or CMB. Our eyes can’t see this light, but other instruments can – for example, some of the static in your TV is from the CMB. Microwave telescopes like Europe’s Planck satellite have precisely measured the CMB, producing maps like the one above. Looking at the map, we can see that the temperature of the CMB varies ever so slightly across the sky, forming a pattern of “hot spots” and “cold spots” that correspond to areas that were slightly denser (i.e. contained more matter) than average and areas that were slightly less dense than average at the time the CMB formed. These tiny variations in density grew into the structures we see in the Universe today: galaxies and galaxy clusters.

Our telescopes contain cameras that are designed to see specific colors (or frequencies) of microwave light. Last year we observed at two frequencies: 100 GHz and 150. The new cameras we installed this year will allow us to observe at a third frequency, 220 GHz. Other experiments look at even more colors – the Planck satellite observed the entire sky at nine different microwave frequencies, providing much more information than any single-color picture could give alone. Looking at the maps below, we can see that the Milky Way is much brighter at some frequencies than others – meaning that it is easiest to study the light from outside our galaxy at frequencies around 100 GHz, where the galactic emission is faintest. We can combine observations at different frequencies to separate the extragalactic CMB light from galactic light, allowing us to study cosmology and galactic astrophysics independently.

The microwave sky as seen by Planck at nine different frequencies. The bright line along the equator is light from our own galaxy, the Milky Way. Note how much brighter this galactic emission is at higher frequencies. At the highest frequencies, dust in our galaxy overwhelms the background CMB emission entirely. Copyright ESA and the Planck Collaboration.

The microwave sky as seen by Planck at nine different frequencies. The bright line along the equator is light from our own galaxy, the Milky Way. Note how much brighter this galactic emission is at higher frequencies. At the highest frequencies, dust in our galaxy overwhelms the background CMB emission entirely. (Copyright ESA and the Planck Collaboration)

Polarization and the CMB

The CMB was first discovered in 1964. This was a major result: the CMB was a key prediction of the Big Bang theory, and its detection represented a major validation of this theory. Since its discovery, scientists have used very precise observations of the properties of the CMB to refine a standard model of cosmology. Thanks in part to these measurements, cosmologists now have a very good understanding of the history and composition of the Universe.

When studying the light from the CMB, scientists can make different kinds of measurements. They can measure the intensity of the light (how bright the light is) at each point on the sky. They can also measure the polarization of the light. Light is a wave, so one can imagine it vibrating as it moves – like a person shaking a jump rope back and forth. Usually light vibrates equally in all directions, but sometimes various processes make it vibrate more strongly in one specific direction. (Imagine a person shaking a jump rope that passes through a picket fence. If the person shakes the rope up and down, the waves can pass straight through the vertical slats in the fence. But if the person shakes the rope side to side, the fence will block the motion.) Light that vibrates more strongly in one direction than another is said to be polarized.

Diagram of light polarization. The light source on the left emits unpolarized light, which shakes in all directions. The polarizer grating only allows vertically-shaking light to pass through, producing polarized light waves to its right. (Copyright Bristol University)

Diagram of light polarization. The light source on the left emits unpolarized light, which shakes in all directions. The polarizer grating only allows vertically-shaking light to pass through, producing polarized light waves to its right. (Copyright Bristol University)

The light in the CMB is slightly polarized. Both the amount and direction of this polarization change across the sky, forming a complicated pattern. Scientists break this pattern down into two different components: E-mode polarization and B-mode polarization. These two patterns are shown below. This particular way of dividing up the total polarization signal is useful because some physical processes produce only E-mode polarization and some produce both E-mode and B-mode polarization. By dividing up the signal in this way, scientists can determine how much polarization each process generates. Each process tells scientists something unique and important about the Universe.

E-mode polarization (top) and B-mode polarization (bottom). The polarization observed at every point on the sky can be uniquely described as the sum of an E-mode pattern and a B-mode pattern. Mathematically speaking, this is equivalent to saying that we can decompose the polarization vector field into a divergence-only (E-mode) component and a curl-only (B-mode) component.

E-mode polarization (top) and B-mode polarization (bottom). The polarization observed at every point on the sky can be uniquely described as the sum of an E-mode pattern and a B-mode pattern. Mathematically speaking, this is equivalent to saying that we can decompose the polarization vector field into a divergence-only (E-mode) component and a curl-only (B-mode) component.

In the CMB, the E-mode polarization component is much stronger than the B-mode polarization component. Almost all of the E-mode polarization signal arises from the small density variations in the early Universe. We therefore mainly use the E-mode polarization signal to learn about the Universe at the time the CMB was emitted, 13 billion years ago. E-mode polarization of the CMB was first detected in 2002 and provided additional confirmation of the Big Bang theory.

In contrast to the E-mode signal, which is overwhelmingly generated at the moment the CMB is emitted, the weaker B-mode polarization signal can tell us about everything from the early Universe to the very local environment within our galaxy. After the light from the CMB is emitted, some of the original E-mode patterns are changed into B-mode patterns before reaching Earth as the CMB light travels past galaxies and galaxy clusters. These objects are so massive that their gravity can bend the light’s path, changing its polarization. This process is called gravitational lensing. The gravitational lensing B-mode polarization signal tells scientists about how matter is distributed in the Universe on the largest scales, which constrains models of cosmic evolution. This was the first type of B-mode polarization to be detected in the CMB and its amplitude is exactly as predicted by the standard Big Bang theory.

Various processes in the Milky Way can also generate microwave light with both E-mode polarization and B-mode polarization. The most significant process at the frequencies we observe is emission from galactic dust. Some regions of the sky contain much less dust than others, but no part of the sky is completely dust-free. The exact amount of polarized dust was not well known until recently, when the Planck satellite published polarization measurements of the entire sky at seven microwave frequencies. Planck found that the B-mode signal from polarized dust was at least as large as B-mode signals from other sources, even in the cleanest regions of the sky.

The final source of B-mode polarization in the CMB is the Big Bang itself. Measuring this type of B-mode polarization is the main goal of my group’s experiments. Scientists studying the beginning of the Universe have theorized that in the first fraction of a second of its existence, the Universe went through a period of ultra-fast expansion, doubling in size many times over. This theory is called inflation. The super-rapid expansion of space sent shockwaves called gravitational waves reverberating throughout the Universe. As a gravitational wave passes through a region of space, space itself expands and then contracts slightly. This stretching and squeezing of space distorts any light in that region of space, generating polarization. The gravitational waves produced by inflation were still echoing through the Universe 380,000 years later when the CMB formed, so they should have polarized the CMB as well. Different inflationary theories make different predictions about the strength of the gravitational waves produced – and therefore about the strength of the polarization signal they impart to the CMB – but all of the theories agree that inflation should generate both E-mode and B-mode polarization. We focus on the B-modes because inflationary E-modes will be swamped by the larger E-mode signal from density variations. Once we detect B-mode polarization and verify that it comes from gravitational waves rather than gravitational lensing or galactic dust, we can determine which of these theories is correct. It is a very difficult measurement to make, since the amount of polarization is very small, but if our group is successful we will be the first people to ever see back in time all the way to when the Universe was less than a second old. It is impossible to reproduce such extreme conditions in any laboratory on Earth, so we will truly be exploring new physics.

Searching for Inflation

Last March, my research group published a very exciting result: after three years of staring at the same small patch of sky, our BICEP2 telescope had detected B-mode polarization. The signal was partially due to gravitational lensing, but we needed a second component from either galactic dust or gravitational waves to explain the full signal. Based on the best models of galactic dust available at the time, we concluded that the excess signal was too strong to be entirely due to dust emission. It therefore seemed highly likely that we had detected evidence of gravitational waves. If true, this would be groundbreaking: direct experimental confirmation of the theory of inflation. BICEP2’s maps of the polarized microwave sky are shown below.

The BICEP2 maps of the polarized microwave sky at 150 GHz. The total signal has been divided into E-mode polarization (top) and B-mode polarization (bottom). The B-mode map shows characteristic “swirly” patterns inconsistent with noise. Note that the color scale is different – the E-mode signal is about 5-6 times stronger than the B-mode signal.

The BICEP2 maps of the polarized microwave sky at 150 GHz. The total signal has been divided into E-mode polarization (top) and B-mode polarization (bottom). The B-mode map shows characteristic “swirly” patterns inconsistent with noise. Note that the color scale is different – the E-mode signal is about 5-6 times stronger than the B-mode signal.

Since BICEP2 could only see one frequency, we couldn’t distinguish gravitational waves from dust using our data alone – we had to rely on models from other sources, which were known to be highly uncertain. Within a few months after our discovery, we therefore started working with the Planck team to combine our data with theirs to better understand the origin of the B-mode signal that we had seen. Planck had their own polarization measurements of the entire sky at seven frequencies, but their data was noisier than ours so they couldn’t directly confirm our observations by themselves. However, by averaging over a larger area of the sky, the Planck team had already begun to realize that dust in our region might produce stronger B-modes than models had predicted, making a joint data analysis even more critical. By using Planck’s high frequency observations to constrain the polarization of the dust in combination with our BICEP2 data and new data from our latest telescope (the Keck Array), we were able to gain more insight into the polarized microwave emission in our patch of sky. At the end of January, we published the results of this joint analysis.

By combining our data with Planck’s data, we learned that the dust in our galaxy is more highly polarized than expected in our patch of sky. This means that the dust’s glow could explain most or all of the signal that our group announced last spring. Any portion of the signal created in the early Universe is too small to robustly distinguish from background noise, even after using all of the data currently available. While this may seem disappointing, even being able to limit the maximum brightness of the B-mode signal from the early Universe already tells us something new about fundamental physics and rules out some theories. We will need more data to get a clearer picture – and our group is in an excellent position to do just that! Our Keck Array telescope continues to build on the dataset collected by BICEP2 and is in better shape than ever, thanks to the upgrades that I helped make in Antarctica this year. We also deployed an entirely new telescope this year, BICEP3. Both BICEP3 and the Keck Array will continue to collect more data of increasingly high quality, and several other teams have built their own experiments with the same goal in mind. It is very possible that within the next several years we will have enough data to measure the emission from galactic dust well enough to remove it from our maps of the sky and peer through it to the gravitational waves signal that may lie beyond. Over the past few decades, the CMB has provided a wealth of new information about our Universe – and the most exciting discoveries may still be ahead of us.

The Keck Array, ready to explore whatever the Universe has to offer.

The Keck Array, ready to continue exploring the many mysteries of the Universe.

Additional Resources

My research group’s science website: Learn more about our work, read our scientific papers, or download our datasets.

Wayne Hu’s online cosmology tutorials: If this post inspires you to learn more about cosmology and the cosmic microwave background, these are a good place to start.

The Surface of Light: If you love Disney and/or would rather sing songs than read about physics, then you’ll probably enjoy this musical number about some of the research discussed in this blog post.

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s