A couple of weeks ago, on NPR during “Stardate” with Sandy Wood, there was an interview with some scientist associated with the mission launching to search for extra-solar planets.
I’ve had this question for a while, (even before the interview) but I remembered it in reading the thread on how we know the Milky Way is a spiral galaxy (barred).
My question is this, the guy on the radio said that they’d look at a star and see if there was a slight dip in it’s apparent brightness. Then, they’d look at it a year later and see if it dipped again. The mission is for 3 years. He actually used the term “a year later,” during the broadcast. I thought, “that’s great, if the planet they are looking at just happens to have the SAME year as Earth.” So, how does that work? There are some planets in our own solar system that take hundreds of years to orbit our sun. How would we be able to detect those planets around other stars?
I mean, most of time, up until now, they talked about how the star wobbles about a common axis with it’s planet. But if the planet takes 300 years to orbit the star, we wouldn’t even see one wobble in our lifetime! Are we only detecting planets that orbit REALLY fast?
The detection of extra-solar planets is a fairly hit-or-miss procedure. It’s not like they single out one specific star, then sit back and wait. There are probably quite a few false-positives and negatives. In short, the observers have to have luck on their side.
The planets they are specifically looking for are earth-type planets. Those are generally going to be about the same distance from their star as we are, give or take a bit. And since the stars we’re looking for are going to be similar to the sun, they’ll have roughly the same mass.
So the orbital period might not be a year, but it won’t be 100 years. It might be 14 months, or 10 months. Wait for the thing to transit the sun three times and you have three data points and can calculate the orbital period and the mass.
The coolest part is by measuring slight changes to the spectra of the light as som e passes through the planet’s atmosphere, we have a chance to actually detect the composition of the atmospheres of those planets. That’s very exciting. It also means we have ‘sensors’ that can detect the atmospheres of planets from distances of light years. That’s just too cool.
I’m utterly entranced by this subject. See www.nationalgeographic.com for a very recent discovery of an extra-solar planet out of re-analysed Hubble data.
Most of the planets discovered have been huge, big enough to be detected by the method of reduction in magnitiude, but I also recall reading recently that smaller planets have been identified using some other method.
The Kepler mission is going to stare at the same patch of sky for years. It’ll record all the blips for all the stars in that patch. Then they have to analyse all those blips to see if any are repeating at regular intervals. Three blips is the absolute minimum, but four or five will make it more certain. After that they’ll want to verify it by watching for the next expected blip with other telescopes.
I’d say the guy on the radio was simplifying to avoid a complex discription and this simplification was misleading.
According to Wiki, there are a couple of exoplanets that have been directly imaged, among them 2M1207b, the HR 8799 system, and Fomalhaut b. For such planets, we can figure out their orbits by seeing how far they are from the star — in fact, they have to be fairly far out for us to resolve them — and how quickly they move in the sky; the three examples above have orbital periods estimated to be in the hundreds of years. However, you’re right that planets have to be moving pretty quickly for us to infer their presence from their central star’s “wobble”; the longest orbital period of any planet discovered this way is about 19 years for 14 Herculis c. For comparison, Jupiter’s orbital period is about 12 years and Saturn’s is about 30.
Are you sure he said they test the apparent brightness every year? Because if they were testing the spectrum, that would make some sense. One way to test for a planet is to check if there is a tiny Doppler ‘wobble’ as the star is tugged by the gravity of the orbiting planet, but this Doppler shift is so incredibly tiny that we measure it when the earth is at nearly same point in space, so the interstellar matter between us and the system is close to the same. Just a way to cut down on error.
If he was talking about a dip in apparent brightness, then you are absolutely right, what he said makes no sense. The only way we’d see a dip in brightness is if the planet transits (passes in front of) the star in question, which is something you want to look for over a loooong stretch of time.
Kepler definitely looks for changes in brightness. That’s how it can detect earth-sized planets. Detecting the wobble of the star can generally only be done if the planet is much larger and/or very close to the star. That’s why so many of the planets that have been discovered so far are ‘hot Jupiters’ - Jupiter-sized planets orbiting very close to the parent star.
Here’s the description of Kepler’s technique from the Kepler home page:
Here are some very cool links if you’re interested in extra-solar planet:
On that page, you can list all the extra-solar planets that have been discovered, and filter the list by a bunch of criteria. Doing a search for ‘terrestrial planets’, you can see that we’ve already discovered 14 different terrestrial planets around other stars, ranging in size of 3.3 earth masses to 17 earth masses.
The Extrasolar Planets Encyclopedia contains tons of information about extrasolar planets and the techiques for finding them.
And here’s the Home Page for the Kepler mission. Lots of good information there, too.
Sim Lite is the successor mission to Kepler, scheduled to be flown some time in the next decade. It will use interferometry to detect wobble to a very high precision, and will be able to detect earth-sized planets and measure their masses.
The Terrestrial Planet Finder is the mission that will hopefully follow the SIM Lite mission. TPF will be able to measure the temperature and atmospheric composition of earthlike planets orbiting in the habitable zones of nearby stars.
Only a very small proportion of extrasolar planets will eclipse their star from our point of view (a little less than 1%, by my back-of-the-envelope calculation), so you need to watch a great many stars to see any, and we’ll miss the vast majority of them. But those few that do eclipse will do so repeatedly, every year (that planet’s year, that is, not ours), so you can get the extra data points you need to be certain of the detection.
Ignore what the guy on the radio said. He was obviously simplifyiing way too much.
There’s an article in the current Sky&Telescope about the discovery of a transiting planet that’s only 1.7 times Earth’s mass. It orbits its star in only 20.5 hours. It’s way too hot to have Earth-like life. Link: First Earth-Size Exoplanet