You can see me in real life again wearing glasses again, I have an entirely new shirt on, and you can see me here in my office. That must mean only one thing, and that one thing is that we have another guest lecturer. Our guest lecturer for this video is Heather Knutson. Heather is in the office right next door to me. We could knock on the wall right here and talk to her. She's a Professor of Planetary Science here at Caltech and she works on extra-solar planets. As she'll tell you about in this lecture, she mostly works on trying to understand what's in the atmospheres of these extra-solar planets, which helps you understand what these planets really are. But, I'll let her describe this to you. Today, I am going to tell you a little bit about some of the studies that we are doing to characterize the atmospheres of extra solar planets. I can now show you all of the planets for which we have measured masses and radii. So, there are a couple 100 of these planets currently known. So here on the x-axis, I'm showing you the Planet Mass in units of Earth Masses, and on the y-axis, I'm showing you the Planet Radius in units of Earth Radii. So, you can see that the majority of the planets which had been studied with both of these techniques, are up here in the top right. These planets have masses that are several 100 times the mass of the Earth. That is about the mass of Jupiter. So not surprisingly, when we calculate the average densities for those planets, we see that they are consistent with the planets being mostly hydrogen and helium. So, these appear to be gas giants very similar to Jupiter in our own solar system. If we move down and to the left, we start to go to slightly smaller planets. Neptune is about 20 times the mass of the Earth, here. So, planets in this mass range, have a slightly higher average density. Again, consistent with something similar to Neptune, where they have a rock ice core with the hydrogen and helium envelope. The really interesting part of this plot is down here on the lower left. Earth is all the way over here on the left axis, one Earth-mass. So, down here on the left, you have planets which are quite clearly smaller than Neptune, but there's still a fair bit bigger than the earth. So these are planets which have masses that are anywhere from several times up to 10 times the mass of the Earth. We call these planets Super-Earths, even though we don't actually know if they're Earth-like. So, we can imagine Super-Earths having a variety of compositions. Just to illustrate that, if we just take a single mass. So, let's focus in here on the eight Earth-mass line. You can see that there's a number of planets that fall on that line, which have very different radii. So, that's suggesting to us that those planets have very different average densities and therefore, probably, very different compositions. So the planets up at the top of the line with larger radii are going to have lower densities, and probably more hydrogen and helium, whereas the planets at the bottom of the line are going to have higher densities and perhaps relatively larger fractions of rock in their interior. It's a really interesting question to ask is, what might the Super-Earths be made of, and what fraction of Super-Earths really are rocky versus being something more like say, a mini Neptune? So, if we zoom in on the Super-Earth part of this plot, so now this is the same plot that I showed you before but just having a smaller range in Earth-mass on the x-axis and Earth radius on the y-axis. So, what I've done is to over-plot all of the known transit planet systems as these black circles. I've also added these lines to show you at a given mass, what radius we would expect to see depending on what the composition of the planet might be. So, you can see focusing in on our eight Earth-mass line, that if you had a planet which was a lot of rock and iron, it would be down here at about one and a half earth radii, and that as you increase that to be a planet which is all rock, no iron, the radius gets a little bigger. Then as you add some water to that and take away some rock and iron, the planet gets bigger again until you reach a point up here where you imagine a planet which has no rock and iron but is entirely made of water. That gives you a prediction there. You'll notice there're some planets at this mass point which are actually even larger than that 100 percent water prediction. These planets must certainly have an atmosphere made of hydrogen and helium because they're too puffy to be explained by any of these four models that we consider here. So we think we know what's going on up at the top, these planets that are up high in this region are certainly mini-Neptunes with lots of hydrogen and helium in the atmosphere. So if we zoom in on this region here, you can see that up top you can tell that these planets have a lot of hydrogen, but in down in here, things are a little bit more muddled. Let me explain exactly why that is. So, for the planets in here, for instance, for this 97658 planet, it lies right on this half water half rock line. So, you could imagine it being something like this cartoon picture on the left, where it has a water-rich envelope surrounding a small rocky core. But the Caveat is that we haven't considered all three possible ingredients together. So, water, rock, and gas, which in this case is hydrogen and helium gas. So, you could imagine making another toy model for this planet, which has exactly the same mass and radius, by just taking a slightly larger, rocky core, and surrounding it with a puffy atmosphere made of hydrogen and helium. So, the problem is that, for planets which are down here, which have very high densities, we can be reasonably confident that they're all rock. For planets that are up here with really low densities, we can be reasonably sure that they are have a fair amount of hydrogen and helium. But these intermediate planets, which is where a lot of these Super-Earths lie, are very mysterious because they can have a pretty broad range of different compositions that are all capable of matching their average mass and radius. So, the key to distinguishing between these two scenarios here, is to be able to measure the composition of the planet's atmosphere. If we looked at this planet, and I could tell you that it had an atmosphere which was almost entirely made of steam water vapor, you would think that this is probably the water-rich scenario here on the left. Whereas, if I told you that I measured its atmosphere and it had a lot of hydrogen and helium, that would make you think that it was probably more likely the scenario on the right. I said before that Super-Earths could be like a scaled-up version of the Earth, which are mostly rock. That they could be a mini Neptune which is the scenario with the hydrogen and helium atmosphere. But I glossed over the fact that this water-rich case, actually, is a case where we have no solar system example. So, what do I mean by a water world? So, when I say water world, you might be picturing something like this image here on the right, which is this ocean image, maybe even some dolphins jumping out. In this case, the planets probably don't have an actual liquid ocean with a distinct surface on it, because the planets are relatively massive and relatively hot. The water in the interior is a super-critical fluid, and then as you move farther out, that transitions into a steam atmosphere without ever going through a liquid-water phase. Although we call these water worlds, they're not actually worlds with just a a giant ocean surrounding the whole thing. So, they're a little bit different than what we picture, and that's interesting because we're not quite sure what to expect from those worlds. So how do we measure the composition of these planets atmospheres in order to answer the question, is this planet a water world or a mini Neptune? So, one way to do that is by taking advantage of the eclipse geometry. So, if we wait at and watch when the planet passes in front of the host star, the planet has an atmosphere. So, we mentioned before that the planet is going to block part of the star's light, and by measuring the amount of light that it blocks, we can learn about the radius of the planet. But what I was neglecting to mention before is that if our planet has a thick atmosphere, that atmosphere is going to be opaque at some wavelengths, and transparent at others. So, if you could imagine measuring this transit depth at a wavelength where the atmosphere's opaque, you can see that the planet's going to block a bit more of the star's light than it would if we looked at a wavelength where the atmosphere is transparent. So, just to put this on a plot, you can see up top, this is a plot showing the expected transit depth. So, this is the amount of light blocked by one of these Super-Earths on the y-axis as a function of wavelength in microns on the x-axis. So, this is mostly focusing on this near infrared wavelengths. So, in this particular plot, they consider a range of different atmosphere types for their planet. So, these three colored lines on the top correspond to atmospheres with a lot of hydrogen and helium. So, you can see that varying the compositions of the other elements doesn't really change the shape very much. You have a lot of water absorption, you have a little bit of CO2 absorption in some cases, and maybe even a little bit of carbon dioxide and ammonia, depending on the planet. Something interesting happens when you consider atmospheres which have very little hydrogen and helium. Those are atmospheres like the pure-steam atmosphere that we talked about before, the pure water case. Or you could imagine something which is half water and half CO2, or even almost entirely CO2, something like Venus. So, you can see that even though these are very different atmospheric compositions, that they all look pretty similar as a function of wavelength, they're all very flat. Why is that? Why are the hydrogen-rich atmospheres varying up and down with these big absorption features, whereas these other atmospheres give us transit depth which is almost constant as a function of wavelength. So, the answer to that lies in this idea of this little ring of gas which is doing the absorbing in the planet's atmosphere. So, the thickness of this ring is a function of the scale height of the atmosphere. So you can think of the scale height as a measure of the puffiness of the atmosphere. Large scale height means, very low density, diffuse atmosphere. Small scale height means, very dense, very compact atmosphere. The scale height is a function of the temperature of the planet's atmosphere, its surface gravity, and MMU here is the mean molecular weight of the atmosphere. So, we can take a pretty good guess as to what the temperature of the atmosphere is. We know the mass and radius of the planet, so we can calculate the surface gravity. So, the only part of this equation that we don't know is the mean molecular weight of the atmosphere. If we consider two different models for our planet, one where it has a hydrogen-rich atmosphere, and one where it has something like a steam atmosphere, the hydrogen-rich atmosphere has a low mean molecular weight. So, this MMU is small, and H, the scale height is large, so the atmosphere is very puffy, and has a relatively thick ring of atmosphere that's doing the absorbing. So, because that ring is thick, the variations as a function of wavelength are large. So, this is this large amplitude signal up here. For the case of an atmosphere which is mostly water or CO2, those molecules have relatively large values of MMU, the mean molecular weight. So, as MMU gets bigger, H gets smaller, and the thickness of this ring shrinks. So, as this region gets smaller, it damps out the signal that we're trying to measure. So, by making this measurement of the depth of the eclipse or transit as a function of wavelength, we can learn something about the composition of the planet's atmosphere. So, this is an example from a recent work that I did with my group here at Caltech. So, this is a plot showing a measurement of the wavelength dependent transit depth for a Neptune-sized planet called GJ 436b. In this case, the planet has a temperature of about 800 Kelvin. So it is Neptune-like, but it's much warmer than our own Neptune. So we looked at this planet with the Hubble Space Telescope and we measured the depth of the transit when the planet passes in front of the star at wavelengths ranging from 1.2 to 1.6 microns. So, what we expected to see was this red model here. We expected to see this nice really big up and down variation in the amount of light that the planet's blocking as a function of wavelength. And so the main cause of this feature here is that there's a big water absorption band here. So if this planet really wasn't Neptune analog with a lot of hydrogen, helium and water vapor in its atmosphere, we would see something that was going to look like this red model. And so in black here, I'm showing you the actual data that we measured, and you can see that the transit depth that we measured is almost constant as a function of wavelength. That was really surprising because this is a Neptune. This planet is big, it's puffy, it has a low density, so surely we should see this nice signals. So, we came up with two explanations for what we saw in the data. So one explanation is that we were wrong, and this planet isn't really a Neptune analog, and that even though it has a mass and a radius which are close to that of Neptune, that in fact this planet has a hydrogen-poor atmosphere. So if you take all the hydrogen out of Neptune's atmosphere, you would get something which had a spectrum that looked a little bit more like this blue line here. So remember that makes your little annulus of atmosphere thinner, it suppresses the absorption features that you would expect to see. So that's one possibility, but we also realize that there is another possibility that we had to consider, which is this green model here. So, what if we had a Neptune which had a lot of hydrogen in its atmosphere but also had clouds? So those clouds are going to act to block out the signature of the absorption from the planet's atmosphere in the same way that making the atmosphere really heavy in shrinking down this annulus would do. So, in this case, there are two different scenarios. One, which is that it's not really a Neptune after all and it has a hydrogen-poor atmosphere, and the other being that it has clouds which hide the signal from our point of view. And so in this case, based on the temperature and mass of the planet, we think that the clouds could be something like zinc sulfide or potassium chloride which would be kind of salts, which is kind of fun to think about. So, there's one other technique that we have which we can use to study the atmospheres' extrasolar planets. From that technique is the secondary eclipse technique. So before, we're studying the transit when the planet passes in front of the star. But if we wait about half an orbit, we can also see the planet pass around behind the star. So, by measuring the depth of the eclipse when the planet passes behind the star, we can get a rough idea of the temperature of the planet, which is another thing that we might want to know. So just to give you an idea of what some of the temperatures in these planets might be, many of the planets that we study orbit very close to their host star, so they're quite hot. They typically have temperatures somewhere between a 1,000 and 2,000 Kelvin. And so for comparison, Venus has a temperature of about 700 Kelvin and the Earth is about 300 Kelvin. So, this is an example of a measurement that we made using the Spitzer Space Telescope at mid-infrared wavelengths, in this case it was about eight microns. So here, you're seeing the planet approach the star. Here is where the planet goes behind the star. So it disappears from view and then it reappears. So that decrease in light as the planet disappears and then reappears tells us how bright the planet is at these infrared wavelengths. And if we assume that the planet is emitting as a black body, we can translate that brightness into a temperature for the planet. So in this case, the planet has a brightness temperature of about 1200 Kelvin. So, just to sort of illustrate that visually for you, on the left, here is this image of a person taken with an infrared camera. You can see that where you see their skin, it's brighter because their skin is hotter. And then where they have clothes on, the clothes are colder and so it's fainter. So, in the infrared, if we assume that the planet emits as a black body, this brightness is a proxy for temperature. Another fun measurement that was published recently is the same kind of thing measuring the secondary eclipse, but this time in visible light. So this is measuring, in this case, the amount of light reflected by the planet at different wavelengths in the visible. So if we were to look at a hot Jupiter with the naked eye, this would tell us what color the planet would appear to be. So this was a measurement made with the Hubble Space Telescope as well. So in this case, surprisingly, even though we might picture hot Jupiters as being these kind of glowing red objects, it actually turns out that if we could fly a spaceship to the nearest hot Jupiter and take a picture, that it would appear to be blue, which is kind of interesting. So, this particular hot Jupiter was more blue than green in reflected light and it was also relatively dim. So, compared to most of the objects in the solar system, it would look pretty dark. So it would be dark and kind of bluish in color. We have evidence that this planet may also have clouds in its atmosphere. In this case, the clouds would be potentially made of liquid rock. So, perhaps the blue color is related to the presence of those rock clouds in the atmosphere. So the last question that I want to leave you guys with is the question of whether or not we could use these techniques to study a truly Earth-like planets? So things that are, this mass and radius theory, that are mostly rock with a thin atmosphere on top. So far we've been observing things like Jupiter, kind of down to Neptune and maybe even some Super-Earths. But to go from those down to something which is truly Earth size is still a big challenge, and that mainly has to do with the fact, that the planet is so much smaller than our gas giant planets. So one way to make our challenge a little bit easier is to cheat, and the way we cheat in this case is by changing the parameters of the problem. So, when we think about looking at Earth-like planets, we sort of picture them in something like our own solar system. So an Earth-sized planet orbiting a sun-like star. If that was the case the planet would create a very tiny eclipse when it passed in front of the star. However, if we cheat and look for these planets around stars which are smaller and cooler than the Sun, then that's going to make the planet bigger in the relative sense. So the planet star radius ratio gets larger if we look at these small star systems. So, if we want to study terrestrial planets using these same techniques that was describing and using the same telescopes that we have available, the best way to do that is to search for them around a relatively small stars because that allows us to find systems with a signal that we can actually hope to detect without building a new giant space telescope. So there actually are surveys that are looking specifically for small planets in these kinds of systems. So one example is the Mearth Project. Okay. Another survey that's going to really good at finding small planets is the TESS survey. So, this telescope will be launched in 2017 and it's going to survey the entire sky. It's going to find the brightest and closest transiting planet systems theory, including many systems around small stars. If we're lucky and we get some small star systems that are relatively nearby, we can follow them up with the James Webb Space Telescope, which is a new space telescope that will be launched in 2018. It's going to observe an infrared light and it will allow us to study their atmospheres using some of these same techniques. So just for fun, I always like to finish up my talks on exoplanets with an image of our own galaxy because I think it's also nice to sort of put in perspective what we've managed to accomplish so far with exoplanets. So this is just an artist's impression of the Milky Way where I have a little red arrow pointing to the position of the sun. And so, if I draw a circle which encompasses the region containing all of the exoplanet systems that we've studied to date, so all the transiting exoplanets that we know off, that region would look like this. So you can see that we have barely managed to go a little ways out into our own little local region of the galaxy. We're not even beginning to study extrasolar planets on the other side of our galaxy. So, we're really just beginning this process and I think it'll be a lot of fun to see what we find going forward.