With thousands of exoplanets in the bag, the game has moved on to characterizing those exoplanets. From the Doppler method we get the mass and the distance of the planet from the star, from the transit method we get the size of the exoplanet. Going beyond this requires exquisitely detailed observations that have only so far been possible in a handful of cases. Spectroscopy is a critical tool in this work, conceptually we can imagine taking a spectrum that includes both the planet and the star. Then another spectrum that just has the star and subtracting them to get a spectrum of the exoplanet. That's a principle of how we might characterize an exoplanet using spectroscopy. In practice, this method has only been used for a handful of targets. It requires the extraordinary stability of the space environment and so has been done by the Hubble Space telescope, but actually only for Jovian planets, not for terrestrial exoplanets. It works like this, as the planet passes in front of the star, light from the star is blocked by the planet but some of that light filters through the atmosphere of the giant planet and comes to us. The spectrum of a star has absorption lines imprinted from the cooler outer layers of the star. And those absorption lines give the spectral elements and the chemical composition of the star. Typically, hydrogen and helium but also potassium, sodium, and other heavy elements in trace quantities. When we're just observing the star, we get the stellar absorption spectrum. When the exoplanet is in front of the star, some of the star's light is filtered through the atmosphere of the exoplanet and extra absorption is imprinted by that atmosphere. In principle, differencing the two will give the spectrum of the exoplanet atmosphere. In a handful of cases where it's been used, several chemical elements have been detected in the atmospheres of giant exoplanets, sodium, carbon dioxide and water vapor or steam. These are high Jupiter's and so water is in the form of steam. These observations are proof of concept for the eventual detection of biomarkers. Where we will try to look for oxygen or ozone being imprinted in the spectrum from the atmosphere of the exoplanet, that would be a sign of life. As data accumulates, more and more exoplanets have both detections by the Doppler method and transits. If you have a mass from the Doppler method and a size from the transit method, You can use the two to get a density. So the fraction of exoplanets where both pieces of information are available are extremely valuable. All you get however is one number, from one mass and one size, and so that's a mean density. By comparing that mean density to the mean density of rock or metal or water or gas. It's possible to say what the average composition of the exoplanet is, and in principle distinguish terrestrial planets from gas giants. COROT-7b was an example of an early planet where this technique was used. It turns out to have a similar density to the Earth and is a super earth several earth masses-sized object. However, it orbits its star in only 20 hours and is so close to that star that its surface is probably molten, it's nothing like an Earth-like world. Another Earth-like planet found in the last few years has a size just under three times the size of the earth and a mass of about six times the mass of the earth. Its mean density of 1.8 grams per cubic centimeter implies the planet may be composed primarily of water which has a density of 1 gram per cubic centimeter. This is exciting, the possible detection of a water world, this planet is relatively nearby, only 40 light years away. And in principle, if there are living creatures on there, our light waves and radio communications have swept over them since the dawn of the electronic age. Unfortunately a single mean density does not uniquely define an exoplanet, there's what's called a redundancy or degeneracy in the models. In other words there are many different chemical compositions and layerings of a planet that can produce the same mean density. It could happen if you had a mostly rocky planet or a planet with a gaseous envelope and a small metallic core, those would give the same mean density. So we need more information before we can confidently talk about finding water worlds or earths. In fact, the zoo of exoplanets contains extraordinary diversity. Based on these mean density measurements and other chemical indicators, there appear to exoplanets primarily made of metal, primarily made of silicates, like rock. Primarily made of carbon, and perhaps primarily made of water. An extraordinary diversity of exoplanets, including some completely alien from those planets in our solar system. Kepler has found earth-like worlds and giant worlds in many different orbits. Most of these orbits, like the early exoplanets discoveries, are very close to their parent stars. So the equilibrium temperatures on the rocky planets found by Kepler so far are extremely hot, these are not typically habitable worlds. Characterizing an exoplanet goes beyond the simple discovery and the measurement of a mass or a size. It either involves combining information, such as from the Doppler and transit method to give a mean density and some sense of what the planet might be made of. Or with much more difficulty obtaining a transmission spectrum of the atmosphere of the exoplanet. This is a proof of concept for the detection of biomarkers, the way we might actually find life on another world.
