In my last article I discussed eclipsing binary stars; a configuration of two (generally close) stars that happen to orbit each other along our line of sight, thereby eclipsing each other every orbital period.
The most successful way of finding planets to datethe radial velocity or "wobble" methodrequires single stars since it detects the very slight offset motion of the star caused by the giant planet orbiting it. As discussed in my last article, the photometric transit method can be used to detect planets that cross two stars (the planet orbits both stars). These transits are more complex since they involve two stars orbiting each other while the planet crosses in front.
The general rule for planetary stability in two-star systems is that the axis ratios should be greater than 3-to-1. This means that if two stars that orbit each other are fairly far apart, a planet could orbit one or the other at a distance of less than one-third their separation. (This is the configuration of the giant planet that orbits 16 Cygni B, the lesser bright of the 16 Cygni double star system.) On the other hand, if the two stars are very close (usually the case for eclipsing binaries) then a planet can stably orbit both, provided its orbit is at least 3 times the two stars' mutual separation. It is this latter configuration that allows the detection of outer giant planets using the eclipsing binary timing method.
able --> I was at the observatory one evening, contemplating planets orbiting stars that orbit each other when I glanced at my watch. I thought of the ends of the hands on my watch as two stars orbiting each other. When I held my arm so that I was looking into the plane of the watch, I noticed the watch hands eclipse each other. I realized that the eclipsing binary is itself a clock!
I remembered from my undergraduate training that several people determined the presence of a third star in a system with an eclipsing binary by timing the binary eclipses. I thought that if one wanted to balance a giant planet and a binary star system on a teeter-totter, the giant planet would either have to be very heavy or far out on the board, or a bit of both. If you were to spin the teeter-totter, the size of the circle made by the eclipsing binary would be determined by the mass of the giant planet as well as the size of its orbit (i.e. distance from the center-of-mass, or fulcrum of the teeter-totter).
The eclipsing binary was also like a clock. If I quickly moved my watch away from you by about 1 light second's distance (186,300 miles) what do you think you'd see? With a big enough telescope youd notice that my watch showed a 61-second minute! Why? Though it took 60 seconds for the second hand to go around the dial, it took an extra second for the light to cross those 186,300 miles. Similarly, if I had moved a light second closer from out in space, you would have seen a 59-second minute.
The timing of the eclipses in the eclipsing binary would be displaced back and forth (60-second minute, 61-second minute, 60-second minute, 59-second minute, and so on) each orbital period of a giant planet orbiting it. By timing the eclipses (which are measured using the same imaging method used to look for planet transits) we could discover giant outer planets without them having to transit the stars!
We use a Global Positioning Satellite clock and software, and then we determine the exact time of the middle of the eclipses. We reduce these eclipse minima times to the Julian Day time (this is the number of days, or fractions thereof, since noon on January 1, 4713 B.C.; it keeps astronomers from things like leap years and pontifically banished days). Finally we have to convert the Julian Day at the observatory to the time in the center of the Sun so that we can compare all the times with each other irrespective of where the Earth was (otherwise this could cause an error as much as 17 minutes).
So, does this method work? Yes. We have a candidate giant planet (about 2 Jupiter masses) around the eclipsing binary CM Draconis with a period of about 900 days. But we'll have to wait until it orbits a bit more before we know for sure if its really there. What is really neat about this method is that the same data used in the photometric transit method can be used here as well. We just look at a different part of it.
In the next essay well look at another photometric method that came about, as I was contemplating the Moon.
advertisement
