Simulating Milliseconds of Stellar Collapse: A Conversation with Christian Ott

Theoretical astrophysicists infer what sort of physical processes might cause the observed behavior of the universe; observational astrophysicists—astronomers—observe the universe to determine what is out there and how it is behaving.

Theoretical and observational astrophysics overlap more often than you might think. Astrophysicists with their varying specializations are in constant conversation with one another, weighing theory against observation and vice versa. Certainly this is true in the area of gravitational waves, first theorized by Albert Einstein nearly a hundred years ago as part of his general theory of relativity. While gravity is weak compared to other forces in the universe, gravitational waves actually squeeze and ripple space-time, creating physical effects in the universe that have not been successfully explained by any other mechanism.

There is excellent observational evidence for the existence of gravitational waves, including the behavior of the Hulse-Taylor pulsar, a binary system first discovered in 1974, and the recent finding by Caltech professor Jamie Bock and his coauthors that the cosmic microwave background has a polarization pattern specific to the gravitational waves that would have been released during the period of rapid inflation at the beginning of the universe. As of today, however, gravitational waves have not been directly detected, though not for want of trying. The Laser Interferometer Gravitational-wave Observatory (LIGO), a collaboration between Caltech and the Massachusetts Institute of Technology, is currently being refitted with a new technology called Advanced LIGO. When Advanced LIGO goes online in 2015, there is hope that it will be able to directly detect gravitational waves as they come to the earth.

Christian Ott, professor of theoretical astrophysics at Caltech, is eagerly awaiting data from Advanced LIGO. Ott formulates scenarios for what happens when stars collapse, and one result of stellar collapse is the rapid release of gravitational waves, just the kind that LIGO hopes to detect.

Much about the collapse of massive stars is well understood. But there are crucial hundreds of milliseconds in this process that determine whether a star will collapse into a black hole or into a neutron star, and these milliseconds are still a matter of highly educated and informed speculation. It is these fractions of a second that consume Ott's interest. His scenarios for stellar collapse are stories told with multiple terabytes of computer memory and petaflops of computing power—stories that are plausible, but whose truth is still unknown. One day detections of gravitational waves will help to confirm or contradict the models of stellar collapse that Ott is creating.

How did you get interested in astrophysics?

I've had an interest in this since I was a child growing up near Frankfurt, Germany. My father was an amateur astronomer. We had a small telescope at home, and we would look at the stars and the planets and the moon. After high school I chose to go to Heidelberg University to study physics and astronomy. As a freshman I read a book by Kip Thorne (Richard P. Feynman Professor of Theoretical Physics, Emeritus) in German translation: Black Holes and Time Warps. He has a way of explaining these things so that even a layperson can understand them, and I became fascinated with black holes, neutron stars, and regions of strongly curved space-time. Honestly, Caltech seemed to me to be some mythical place. I wasn't even daring to dream about a place like this, and now I'm a professor here. It seems crazy to me.

What spurred your interest in gravitational waves?

Heidelberg University has an exchange program with the University of Arizona, so I came to spend a year there during college. Shortly after I arrived, I was telling a graduate student about my interest in neutron stars and black holes, and he recommended that I talk to Professor Adam Burrows, now at Princeton University. I wasn't too excited about gravitational waves at that point. I remember that quite well. But Professor Burrows set me to work on calculating gravitational waves from supernovae. That was 2001, and I've been working on similar questions ever since.

What do you find exciting about supernovae?

What most people don't realize is that without supernova explosions, we wouldn't be here.

There are two kinds of supernova explosions. Type Ia, those that come from white dwarf stars, are responsible for about 80 percent of the iron in the universe, and core-collapse supernovae, or Type II, which come from massive stars, are responsible for the remaining 20 percent of the iron. Without supernovae, there wouldn't be iron for our blood; there wouldn't be iron in Earth's core; there wouldn't be iron to make steel. Type II supernovae are also responsible for most of the oxygen and carbon in the universe. Without this enrichment of heavy elements, there would be no life, there would be no planets . . . it would be a pretty boring place.

So blowing stuff up and chemically polluting the universe, as supernovae do, is crucially important. But for fundamental physics, it's actually more interesting to examine the collapse itself.

The physics of stars up to that time is pretty well understood: we know where the pressure comes from in the iron core of a star; we know about thermonuclear reactions. However, as a star collapses the core becomes unbelievably dense. Eventually the electrons, which are exerting pressure in the opposite direction of gravity, are themselves squeezed out in a process called electron capture. In electron capture, a proton and an electron combine to make a neutron and a neutrino, a tiny subatomic particle with no electrical charge. When all of the neutrons and protons are packed together that tightly, the nuclear force kicks in. Usually the nuclear force binds protons and neutrons together, but when you try to squeeze protons and neutrons too close to one another, the nuclear force acts in the opposite direction: it has the effect of an outward pressure against the gravitational pull of a collapsing star. We don't understand this mechanism very well, but if we didn't have the nuclear force, all stars would collapse to black holes. There would be no neutron stars or supernovae. As it is, there are three outcomes we know of when stars collapse: Stars can collapse directly into black holes with no supernova; they can experience a weak supernova and a collapse into a neutron star that then collapses into a black hole within hours or days; or there can be a strong supernova that leaves a neutron star behind, apparently forever.

What determines whether stellar collapses result in neutron stars rather than black holes?

You tell me.

You don't know?

It's what we call "an area of active research." Advanced LIGO should help us to answer this question. When you see supernovae with telescopes, you're looking at optical waves, and these come pretty late in the process of stellar collapse; it's not easily connected to what's actually happening deep inside the star. A star collapse is a highly energetic event and should create substantial gravitational waves. When we detect gravitational waves, we will get information about what is going on earlier in the process. Depending on the precise shape—the amplitudes and frequencies—of the gravitational waves we detect, we can get a finer sense of exactly what is happening in the core of a star when it collapses.

Gravitational waves would arrive on Earth up to a day before we would see the light from the supernova, depending on how far away from us the supernova occurs. The same is true of neutrinos. Although neutrinos are remarkably tiny, a supernova produces an enormous quantity of neutrinos that fly out into the universe. When a star collapses, 99 percent of the gravitational energy released goes into neutrinos; only a tiny portion of the remainder takes the form of gravitational waves. We can already detect neutrinos on Earth, and we have even detected them directly from a supernova in 1987 that occurred in the Large Magellanic Cloud, a neighbor galaxy of our Milky Way. If a stellar collapse occurs anywhere near us, we should detect tens of thousands of neutrinos.

So if you detected these specific gravitational waves or a lot of neutrinos, you could alert the entire scientific community to point their telescopes at the sky the next day to see the supernova?

No! I would tell everyone to turn their big telescopes away, so the instruments would not be destroyed! Imagine if Betelgeuse blows up in a supernova—it's a red supergiant star twenty times the mass of the sun. If that goes, it's going to be as bright as the full moon for an entire month. At the very least, astronomers would need to put filters on their telescopes to protect them from the intense light.