Thursday, March 26, 2009

Hunting planets

Image Credit: Fergal Mullally

My current observing run at McDonald Observatory is part of a project to look for planets circling white dwarfs. This project is led by Fergal Mullally, an astronomy postdoc at Princeton; I'm just helping out by taking some data.

How do we look for planets around white dwarfs? We have to be clever. Most planets have been found by measuring the Doppler shift in light from a star; the gravity of a large planet pulls its parent star around in a small circle, which causes the light from the star to be Doppler shifted as the star moves toward and away from us. But this technique doesn't work for white dwarfs, because it isn't possible to measure their speed accurately enough to detect the small motions caused by a planet. A planet can make a star move at speeds of about 20 or 30 miles per hour, but it's hard to measure the speed of a white dwarf more accurately than a half mile per second!

White dwarfs are stars that slowly cool over time. They used to be the cores of stars like the sun, but when the star ran out of fuel, its outer layers blew away in a planetary nebula, and the hot core of the star was exposed to the cold vacuum of space. Without any heat source, the white dwarf begins to cool off, just like a hot poker removed from a fire slowly cools off.

When most white dwarfs reach a temperature of about 22,000 degrees Fahrenheit, their outer layers begin to slosh back and forth. (This is due to complex physics, but if you are familiar with Cepheid variable stars, the mechanism is similar.) The sloshing causes the white dwarf to get brighter and fainter in a very regular fashion. For a handful of white dwarfs, this sloshing is so regular, we can predict the exact second the star will be brightest several years into the future! The variations in the brightnesses of these white dwarfs are so regular, they rival our best atomic clocks in accuracy.

Now comes the really clever bit. Suppose one of these regular white dwarfs has a planet. The gravity from the planet pulls the white dwarf in a little circle, just like the planets around normal stars do. This means that sometimes the white dwarf is a little close to us, and sometimes it is a little further away. Since light takes time to travel to us, and since the light will have to travel a little farther when the white dwarf is further from us, the light will take a little longer time to get to us. And remember that the brightness of the white dwarf is varying like clockwork. So, if we see the light variations arriving a little later than we predicted, that means the white dwarf is a little further away from us than it was before. And if the light variations arrive ahead of schedule, that means the white dwarf is a little closer to us than normal. In short, if the white dwarf has a planet, we should see the light variations arriving sooner than expected as the white dwarf moves toward us, and later than expected as the white dwarf moves away. (A variation of this method was used by Danish astronomer Ole Römer in 1676 to measure the speed of light; he used the predicted and observed times of eclipses of Jupiter's moons.)

So, Fergal's project involves measuring the arrival time of light variations from the most regularly-varying white dwarfs we know. He's been watching these stars for nearly six years now. Most of the white dwarfs don't show any early- or late-arriving pulses, so they don't have any planets (or at least none large enough to detect). But one white dwarf named GD 66 does show variations. A graph of those variations is at the top of this post. The squares with error bars show how early (negative) or late (positive) the light pulses have arrived at our telescope since 2003. The solid line shows how we would expect the time to vary if a planet about twice the mass of Jupiter were orbiting around GD 66 at a distance of about 250 million miles (2.75 times the distance between Earth and the sun). At this distance, it takes the planet about 6 years to go around once.

So, is there really a planet around GD 66? I think so, but we aren't sure yet. We must see the orbit start to repeat to be sure. In other words, we need to see the white dwarf start moving toward us again. Our best guess is that we should see the orbit repeat starting sometime this year. If we do see it repeat, we can celebrate Fergal's discovery of a planet. If it doesn't repeat, then we're not seeing a planet, and there's something about this white dwarf that we don't understand. And we can't rule that out yet, not until we see the orbit repeat.

And that's why I'm here at the mountain. Tomorrow I'll show you a little of the data I took. I can't tell you if the orbit is repeating yet or not -- there's a lot of analysis that Fergal has to do to turn the data I'm taking into planet orbits, and it may be a few months yet before the orbit starts to repeat. But I'll be sure to let you know!

No comments:

Post a Comment