counting exoplanets by their gravitational wells
Just like supermassive black holes, planets can bend light with their gravity. But do they bend it enough for us to detect them that way?
Depending on who you talk to, planets around alien suns are either somewhat rare due to the chaotic nature of planetary formation around infant stars, or even more plentiful than the stars themselves. Since exoplanets are rather small and dim, lost in the glare of their host suns, spotting them takes a lot of time and effort. Direct observation means catching a momentary dip in starlight from an object of indeterminate size at a time that's random to the observer. If a currently unknown planet orbits its sun every 237 days, how will you know to point your telescope at the right star every 237 days? There just has to be a better way of taking the galactic census so we can figure out what the average solar system looks like, and ultimately, what are the chances one may have the right conditions to host life. And there may be according to a group of astronomers who used a very familiar manifestation of general relativity to escape the normal fuss and bother of exoplanet detection, trying to find planets that orbit a little bit farther from their suns to get a rough measure of solar system sizes.
When we last talked about the physics of wormholes, we looked at microlensing, essentially the distortions in the appearance of an astronomical object caused by the gravity of something relatively small in front of this object in our line of sight bending the fabric of space. Usually we deal with gravitational lensing on the scales of galaxy and galaxy clusters and it's partially how we know that dark matter exists. At either end of the light distorting spectrum it's the same mechanism at work. Traveling photons are skewed by the uneven fabric of space and time. So, the researchers posited, if galaxies can distort the appearance of other galaxies and we see stars doing the same thing, what about planets orbiting stars? We know they also bend light in the wake of their gravitational wells, so a planet orbiting at some distance around a star should distort its halo. And so, after watching 100 million stars, they found evidence of exoplanets in orbit around their parent suns exactly as general relativity predicted would happen if they were there. Of course this is all easier said than done.
Since a star pumps out so much light and the planets have to orbit at just the right inclination to be spotted in the act of disturbing the halo, the 100 million stars had to be narrowed down to just 500 promising ones, and those 500 had to be watched for five long years until ten cases of direct microlensing were finally seen. But all that effort didn't seem to bring consistent data since the results seem all over the place. According to the tally, between 6% and 23% of stars seem to host a Jupiter-like world, between 23% and 74% have a Neptune-like body, and between 25% and as many as 97% might have a terrestrial planet around them. As the planet size gets smaller, the uncertainty increases wildly, so much so as to be almost meaningless for terrestrial worlds which are the ultimate goal of all planet hunters. The exoworlds are just too dim, too far away, and too small to register prominently on our existing instruments, and although the study does imply that pretty much all stars have a solar system of some sort, it can't actually tell us anything definitive about what sort of planet we could usually expect. Going by this survey, it could be anything from a turbulent gas giant to Earth 2.0.
Don't get me wrong, trying to use microlensing to find the statistical distribution of planets is a terrific idea. It's just that the universe keeps on placing interesting things too far away for us to spot with out current tools. We could even try this trick again with better equipment and hyper-sensitive telescopes to see if we can get more predictive and accurate tallies. However, it seems that until then, the candidate worlds seen by Kepler, and in the future, the Terrestrial Planet Finder, will provide us with the most accurate and predictive sampling of our galactic neighborhood since they can point to actual planets with an accuracy I doubt we could get from even the most precise measurements of planet-created microlensing manifesting around distant alien suns. This sort of survey would give us a more accurate picture of planetary distribution across the galaxy and allow us to build an accurate picture of a typical alien solar system. With such a model, we could look at any random star and have a decent idea of what we should expect to see orbiting it and at approximately what distances from it so we can better time our telescopes' observations in the hunt for another planet that hosts intelligent life.
See: Cassan, A., et al. (2012). One or more bound planets per Milky Way star from microlensing observations Nature, 481 (7380), 167–169 DOI: 10.1038/nature10684
