We've heard about the amazing contributions of the modest one meter sized telescope that's in the Kepler satellite, but I want to summarize what we've now learned about habitable worlds from Kepler and from the radial velocity surveys. For most of the past five or six years the lion's share of exoplanets have been discovered by Kepler using the transit technique. Whereas in the first 15 years of exoplanet discovery, most exoplanets were found by the Doppler method. Kepler is entering its waning years with diminished capability. So it's a good time to summarize what we've learned about Earth-like worlds. From Kepler data, we can look at the distribution of the sizes of exoplanets. Remember that Kepler can only detect the size of a planet not its mass. We need the Doppler method for that. The distribution shows that the typical exoplanet is something larger than the Earth but smaller than Neptune, but an enormous number of Earth-like and potentially habitable worlds have been found. Essentially, 700 super-Earths or planets within a factor of three of the Earth's size and another 200 that are truly Earth-like, since almost all of these are potentially habitable, depending on where they are in their orbits, that's close to a 1,000 habitable Earth-like worlds have been found just by one mission, and mostly in the last five years. An extraordinary yield. Kepler runs out of steam trying to find things much smaller than the Earth. The absolute limit for Kepler is something like Mars, but for the census of Earth-like planets Kepler will produce a definitive number when all its data has been analyzed. If we put all the data on exoplanets by all available detection techniques together, we can see the important distribution in terms of the size of the planet and the period of the orbits. Earth of course is the benchmark here. Notice that exoplanets do not uniformly fill this plane of parameters. There are potentially gaps. In fact, there seems to be a deficit for planets slightly larger than a super Earth and smaller than Neptune. It's not fully understood. However, with time we have eventually found some giant planets that are in solar system type architectures. Remember, it takes Jupiter 12 years to orbit the sun. So we would have needed at least a decade of data to even start to find Jupiters in normal Jupiter tight orbits in other star systems. They are being found, but it's still not totally clear whether our solar system's architecture is normal. One way in which our solar system's architecture does not appear to be normal is in terms of the eccentricity of the planets. This was suspected early on based on the first 100 exoplanets found by the radial velocity technique. But if now we look at the eccentricity and periods of thousands of exoplanets, we see the persistence of large eccentric orbits for exoplanets. Remember that in our solar system, most of the planets have eccentricities or deviations from circular orbits by less than 10 percent. Yeah, we see a large fraction of the exoplanets with eccentricities much greater than 10 percent, ranging up to 50 percent. It's not entirely clear why this is. For example, exoplanets with high eccentricities may not be long-term stable, whereas planets in our solar system are. This high eccentricity may indicate there's more chaos in exoplanet orbital systems than in our solar system. There are mechanisms which over time will reduce the eccentricity of the orbits in a planetary system and make them more circular. We believe this happened in the solar system, but for some reason it has not happened in most of the exoplanet systems. We still do not know if our solar system is entirely typical. Another way our solar system may not be typical is as we've seen the most abundant kinds of planets are super-Earths or sub Neptunes. We have no super-Earths in our solar system. This histogram shows the rate of discovery of exoplanets since the first discoveries in the late 1980s and early 1990s. You can see the extraordinary surge due to Kepler, and in particular in two data releases from the last three years over 2,500 exoplanets have been discovered and publicized. Kepler talks about exoplanet candidates to be conservative and cautious, but when they're followed up essentially 95 percent of exoplanet candidates from Kepler turn out to be real. So the overall census of exoplanets now stands at over 5,000. Kepler has an upside and a downside. The upside is the extraordinary efficiency of this very simple mission with a relatively small telescope. The downside however is because of the size of the CCD. Kepler was constrained to look at a small field of view to get enough stars in it to find exoplanets. That means that most of the stars in the Kepler field are quite faint, 14th to 16th magnitude. That's a factor of 10,000 times fainter than you can see with the naked eye. That means in turn that most of the exoplanets found by Kepler are quite remote, usually distances of several 100 light years and often several thousand light years. So following up those exoplanets for example, to measure their mass by the Doppler method is really difficult and requires large ground-based telescopes. This has been a problem that is finally being addressed with the TESS satellite, which will launch in 2017. TESS is designed to complement Kepler by looking at nearby exoplanets over a very large part of the sky, essentially a majority of the night sky. So Kepler will have its extraordinary data complemented by TESS, which will find brighter more nearby exoplanets over the whole sky that are much more amenable to follow up including the search for bio-markers with large ground-based telescopes. It's a very exciting development. The field of exoplanets has enormous momentum because of Kepler and the follow-up missions, and the capabilities of large ground-based telescopes. TESS is also going to be able to look for exoplanets around M dwarfs. Remember, there are a 100 red dwarfs for every sun-like star, and Kepler was mostly focused at looking for exoplanets around sun-like stars, but those M dwarfs of course are exceedingly faint. So only an all-sky survey like TESS can find M dwarf planets. It turns out that most of the habitable real estate in the universe is going to be around M dwarfs. So the exoplanets found around M dwarfs are extremely important also for a second reason. Because the star is so faint, blotting the star out to see the feeble reflected light from the planet will be easier with ground-based telescopes. We can anticipate a lot of exciting results in the next 5-10 years as TESS does its work and as the James Webb Space Telescope starts to follow up the best exoplanet candidates, in particular, the Earth-like ones. Here we see a distribution of habitable exoplanets, and the dark green zone indicates a conservative habitable zone simply indicated by the range of distances from a star where water can be liquid on the surface of a terrestrial planet. But of course, most Earth-like planets or super-Earths will have atmospheres. Those atmospheres will contain gases that may be greenhouse gases like carbon dioxide or methane. So planetary atmosphere models are used to expand the habitable zone. The pale green boundaries in this diagram show the expanded habitable zone based on the presence of greenhouse gases. Overall, it looks like we have several 100 potentially habitable Earth-like planets available after looking for 10 years and from a total inventory of 5,000 exoplanets. These are the prime targets for follow-up in the search for bio-markers. Here are pictures, not true images but imagine the images of several dozen of the nearest habitable worlds almost all found in the last five years. A team at the University of Puerto Rico calculates what's called the habitability index which combines information on the planet size, its mass, likely greenhouse gases, eccentricity of its orbit and so on into one number. In this scale, zero is completely uninhabitable and one is the Earth, perfect. On this scale, Mars has a habitability index of 0.64. Most of the exoplanets you see in this diagram have habitability index as larger than 0.64 in the range 0.7-0.8, not as habitable as the Earth but more habitable than Mars, which we think may indeed still sustain life under its crust. It's important to recognize that even with 5,000 exoplanets and as well into the third decade of exoplanet discovery, this is still a young subject. We know of many habitable worlds, but we have not yet detected a single exomoon. Remember in our consideration of the solar system, we realized that the cryogenic biosphere, places that are quite remote from the sun where water may be kept liquid, under the surface of a giant moon of a giant planet, portents an enormous potential for life in the universe. Exomoons are just beyond the range of detectability, although several groups are trying to use Kepler data to find them. There's a third category of potentially habitable exoplanets of which we know nothing. In simulations of solar systems as they form, in the chaotic phase, it is quite often that planets are ejected from the system into interstellar space. It's fairly likely that some of those ejected exoplanets will be Earth-like or be able to hold significant atmospheres that keep a biosphere intact. So it is possible of course for life to exist without a star on some of these ejected worlds. They're called Nomads. In some of the simulations there is many Nomads produced as planets that are retained in the system. We know of no way to detect Nomads except by the microlensing method, and not a single Nomad has been detected yet, although some could clearly be habitable. Many people enjoyed the movie Avatar, by the director, James Cameron, about life on an exomoon of a giant planet. The search for exomoons is probably the current frontier in exoplanet work along with the search for bio-markers on Earth-like planets. There are two of the detection methods which have the potential to detect an exomoon, perhaps a moon in the size of Ganymede or our Earth's moon. One is the microlensing method which has extraordinary sensitivity, and in principle can get down to a moon mass. The disadvantage or microlensing however, is there's no possibility to follow up, because the microlensing is caused by one star system that passes in front of another never to repeat. People are combing through the Kepler data to look for the faintest hint of a secondary shadow caused by an exomoon from an already discovered exoplanet. Three groups are doing this, and any of them may succeed within the next few years. Either way exomoons is the next frontier of exoplanet research. We can take all the work done on exoplanets so far by all the different techniques, and in this microlensing, although it has yielded relatively few exoplanets, is a very important cross-check on the completeness of the transit method and the radial velocity method. We can take all this information to reduce the incidence rate of planets around all stars in the Milky Way Galaxy. This is going to give us the census of exoplanets and habitable Earth-like worlds in the Milky Way. Here's the histogram. You can see that essentially one in six stars in the Milky Way galaxy has an Earth-like planet. Not all those planets will be habitable, but that's an extraordinary inventory of terrestrial worlds in our galaxy. Finally, we can see where we've come in terms of the Drake equation. Remember that the Drake equation frames a way of calculating the number of intelligent communicable civilizations in the Milky Way galaxy at any time. The number of potential pen pals. The first few factors are astronomical, and they are now determined. Kepler alone has determined one of the factors in the Drake equation. We start with the total census of stars in the Milky Way, 200 billion, and now including super-Earths as well as Earth's, essentially half of those stars are likely to have an exoplanet. Drop by another factor of 10 to reduce the number of habitable earth or nearly earth-like worlds, 20 billion. Take a little while to let that sink in. Also, recognize that the Milky Way is just one galaxy in a universe with a 100 billion galaxies. So the number of habitable Earth-like or nearly Earth-like worlds in the universe is 20 billion times a 100 billion, and that if you're keeping track is 10 to the power 21. That's a 1,000 billion, billion potential biological experiments in the universe. An extraordinary number. You can see how unlikely it would be for all of them to be sterile for us to be alone in the universe. The next factors in the Drake equation are also potentially within reach. The next factor is how many times a potentially habitable world actually host life, and the bio-marker experiment we expect to conduct in the next five to 10 years should start to pin down that number, by showing what fraction of Earth-like worlds have their atmospheres altered by a microbial activity bio-metabolism. The later factors than the Drake equation are going to be harder to nail down, and that of course is what the study experiment is four. Either way, it's an extraordinarily exciting time in astrobiology. In little more than two decades, we've gone from a situation of complete ignorance about exoplanets to a census of over 5,000, including several 100 habitable Earth-like worlds. This extraordinary success is largely due to the Kepler satellite, but other missions are planned and large ground-based telescopes are hoping to do the bio-marker experiment. We can put a number on the Earth-like habitable worlds in the Milky Way galaxy, and it's an extraordinary number, 20 billion.
