PlanetQuest: Join the Hunt for Extrasolar Planets

Ever wantto discover a new world? That's what we are planning to have folks do withPlanetQuest, a distributed computing screen saver that will allow anyone tofind extrasolar planets on their own computer. Like the venerable SETI@home,but distinct from the new planetary-systems-generating program of the OKLOproject at UC Santa Cruz, PlanetQuest will enable users to discover realplanets around other stars using four different detection techniques. (By theway, there is also a NASA mission now named PlanetQuest. The two PlanetQuestsare distinct, but we share a common educational goal to promote enthusiasm forplanet detection.)

The firstplanet detection technique that PlanetQuest will use is the single star transitmethod. Sometimes a planet's orbit will align in such a way that it movesacross its parent star creating a shadow, or transit. For a sunlike star, thebrightness drops by about 1% for about three hours if the planet is about thesize of Jupiter and orbits very close to the star (with a "year" of about aweek). When a planet moves in front of a single star, the light drop isperiodic, and fairly easy to recognize. A difficulty with the single startransit detection method is, however, that stars (like lightbulbs) are brighterin the center than at the edges (and bluer--hotter--in the center also). This iscalled "limb darkening." It means that when a transiting planet moves across thecenter of a star, the light blocked can also look like two eclipsing stars justgrazing each other's outer edges. About two in three transit looking featureswill be grazing eclipsing binaries rather than planet transits, so one mustfind out if the star being observed is a double star or not.

Aspecialization of the transit method is the detection of transits acrosseclipsing binary star systems themselves. This can be very tricky since, inmost cases, the planet cannot be said to be moving across the two stars as muchas the two stars are moving behind the planet as they orbit each other. Thisproduces a predictable but non-periodic signal that many times does not lookmuch like a transit at all. The dimming and brightening will not be regular andtheir occurrence will, generally, not repeat in the same way because the starsand transiting planet will be at different places in their orbits when theplanet moves in front again. PlanetQuest's approach to this problem is to use amatching filter that will compare all possible models of planet sizes andorbits with the observations. This takes a rather huge amount of computationaltime, so is a perfect opportunity for the public to participate withdistributed computing.

The seconddetection technique PlanetQuest will use will be the eclipsing binary minimumtiming method. This method relies on the fact that eclipsing binary stars, asthey move in front of one another, are essentially a kind of "clock" that givesa time stamp to the observations. We watch the light of the stars (takepictures of them) and record the time exactly. (The plot of the brightness of astar as it varies with time is called the star's "light curve".) We have tocorrect for things like where the Earth is in its orbit at the time of theobservations, so we generally work with "heliocentric" time--the time in thecenter of the Sun when the eclipse occurred. This way we don't have to keepcorrecting for the light travel time across the Solar System to the Earth.

This methodwill be able to detect circum-binary planets of Jupiter-mass or larger becausethese planets will offset the two eclipsing stars (like a see-saw) as theyorbit them. The offset of the two stars due to a planet orbiting around themwill be detectable because the eclipses will be early or late, depending on thedirection of offset. PlanetQuesters will also then detect Jupiter-mass planetsthat don't even have to be orbiting across the disc of the two stars.

The thirddetection method is the gravitational lens planet detection. When there is avery close alignment between two stars, the star in front can bend the lightfrom the star behind due to bending spacetime, according to general relativity.Any planets in the foreground star will also bend some light (for a shortertime) and so be detectable also. Since the data for this detection method arethe same as that for detecting transits and binary eclipses (wide field,crowded star field images), then we shall also be able to detect planets usingthis method. The stars will, however, brighten instead of dim. One can tell,for example, a stellar flare from a gravitational lens event becausegravitational lenses will be achromatic--that is, not of any color. Bendingspacetime does not care what color the electromagnetic light wave is whilestellar flares, for example, are brighter in ultraviolet than in red light.

Finally,PlanetQuest will use a new kind of SETI (search for extraterrestrialintelligence) detector that will compliment existing SETI projects. All SETIsearches to date (as far as the author knows) are aimed at detecting the natureof the signal itself, rather than the content of the signal. In the case ofradio SETI, a narrow-band radio carrier wave is detected, and in the case ofoptical SETI a nanosecond pulse is detected. In our approach, we classify thesignals and produce a distribution of the frequency of occurrence of thesignals. Then we compare this with the frequency of occurrence of knownintelligent communications (this field of study is known as "informationtheory"). If there is a close match then we look at the structure of thesignals and their dependence on each other. For example, if one recorded babiesbabbling, and tried to find a connection (syntax) within the message, one wouldnot find such structure. But we know that if we record the vocalizations of anadult human, that there are grammatical and syntax rules that are causingcertain words to occur at certain times with relationship to each other. Thisis what allows one to fill in missing words from a copy where the copier waslow on toner and parts of the text are missing. We have applied this methodextensively to vocalizations of ground squirrels, squirrel monkeys, dolphins,humpback whales, and humans so far. If an extraterrestrial signal isreceived--if they are transmitting information-- then they too will have to obeythese rules of information theory.

We atPlanetQuest have already completed observations at Siding Spring Observatory in Australia and our second year of observations at the UC Lick Observatory in California, and are readying the alpha test software to begin our first tests of the wholesystem. We call the combination of our search engine and educational materialsthe "PlanetQuest Collaboratory" and hope you can join the search. EveryPlanetQuester will discover something--the nature of a star, or a new planet, ormany other possibilities--and get credit for that discovery in the PlanetQuestcatalog, which will be accessable online. So check out our web site, and signup for the PlanetQuest Newsletter. And soon you too can join the professionalastronomers in discovering totally new worlds on your own.

• Shadows and Silhouettes: Looking for Transits
• Images: Venus Transit Gallery
• Detecting Other Worlds: The Transit or 'Wink' Method
• Way-Out World: New Technique Finds Most Distant Planet Ever
• Mercury Transits Sun, Images on Web