Just this year, astronomers for the first time detected water vapor in the atmosphere of a Neptune-sized planet. Although the planet lacks a rocky surface and orbits so close to its sun that the temperatures reach more than 10,000 degrees Fahrenheit, its discovery proves our ability to detect water vapor on distant planets. The next step is to find it on a rocky, temperate world.
Three exoplanet hunters came together to discuss what we can learn about these planets from our vantage point tens of light years away, and answer your questions about how close we are to discovering other Earths. Below is a modified transcript of the discussion. Edits and changes have been made by the participants to clarify spoken comments recorded during the live webcast.
- ZACHORY BERTA-THOMPSON – is the Torres Fellow for Exoplanetary Research at the MIT Kavli Institute for Astrophysics and Space Research.
- BRUCE MACINTOSH– is a Professor of Physics at Stanford University and a member of the Kavli Institute for Particle Astrophysics and Cosmology.
- MARIE-EVE NAUD – is the University of Montreal PhD student who led analysis that recently uncovered a previously unknown giant planet, GU Pisces b, using infrared light.
THE KAVLI FOUNDATION: I’d like to begin by asking about one particular exoplanet that’s about 260 light years away called WASP-43 b. Far from being another Earth, it’s a hot ball of mostly hydrogen gas about the size of Jupiter. Just last week a team of scientists using data from the Hubble telescope released the most detailed map yet of this turbulent planet, revealing quite a bit about air temperatures and about water vapor that seems to exist in the atmosphere. Zach, I’m hoping you can start this off by telling bit more about this discovery and why it’s exciting.
When we made this type of measurement in the past, it was just a type of a black and white map that told us how bright the star is as a function of longitude. With these new measurements of WASP-43 b, that map kind of becomes full color, creating this very vibrant picture of what the surface of the planet looks like.
And with this measurement, they used the Hubble telescope to watch the star for three of the planet's years – that seems like a long time, but the planet orbits the star once every twenty hours so it's actually not so bad – and as they were watching it go around, they measured the brightness in wavelength and colors. From that they could reconstruct this colored two-dimensional map of the surface of the planet. It’s really cool to me that this is an entirely new way of looking at exoplanet systems.
TKF: One of the things they found using this 2D colored map is water vapor. How unusual is the discovery of water vapor on another planet, and what does its existence mean?
BERTA-THOMPSON: The cool thing here about this full-color map is that some of those colors just go pretty much straight through the atmosphere while others are blocked by the presence of water in the planet’s atmosphere. And so, looking at the map, you'll see these little dips or divots in the light in those particular colors. Those showed water is present in the atmosphere of this planet all the way around its orbit, throughout its entire orbit.
But it’s not particularly surprising that we would see water vapor, since water is a very abundant gas. This planet is very, very hot and so we would expect there to be water in gas form. But the cool thing about this observation is it allows you to use that water vapor as a tool to probe different depths within the planet’s atmosphere.
Because we understand how water vapor works, we know that it’s high in the atmosphere. So when you’re looking at a water vapor feature, you're not probing as deep into the planet’s atmosphere where as when you’re looking at other colors. So this kind of gives you the three-dimensional structure of the planet. This is such a beautiful dataset and there’s these really sophisticated models built up to match it, allowing us to both explain all the complexity that we see in these new observations and learn a little something about atmospheric circulation on hot Jupiters.
BRUCE MACINTOSH: To be clear, when we’re talking about water vapor, we’re actually talking about 2,000-degree ultra super heated steam. As Zach says, the fact that it exists is not surprising because water is hydrogen and oxygen and those are among the top 10 common elements of the universe. The signature they detect is very clear, and they can do three-dimensional mapping both around the planet—known as longitude—and deep into the atmosphere by studying light as a function of wavelength, letting you map out new information about the structure of planets.
For example, it tells us that this planet always has one side facing towards its star and one side facing away. The side facing towards the star is incredibly hot, and that drives enormous winds that blow around the planet and heat the backside of the planet. You can use information from these beautiful spectra and the beautiful resolved maps to reconstruct that circulation and see how these huge jets are conveying the heat from the hot side to the cold side.
We can really only do this for giant planets right now, and it took an enormous amount of time on the Hubble telescope to generate this map. In the future, there will be successor to the Hubble called the James Webb Space Telescope and that might be sensitive enough to do this same kind of spectroscopic mapping on planets that are closer in size to the Earth – maybe not exactly like the Earth, but “super Earths” that are a little bigger than the Earth. If they were close to their star undergoing this kind of roasting, you could do the same spectroscopic mapping and understand whether they have water vapor in their atmosphere. In the case of a rocky planet like Earth, water vapor in its atmosphere would be incredibly exciting.
MARIE-EVE NAUD: That’s exactly right. It’s important to remember that WASP-43 b is a very hot and very big planet, so water there is not the oceans and rain you could imagine on a rocky planet like Earth. So, as Zach and Bruce said, this work is really about getting a better understanding of the atmosphere of these peculiar objects. It's not directly related to water and life and everything we know on Earth.
TKF: I imagine this Jupiter-type planet can help us understand planetary physics or planet formation in general.
BERTA-THOMPSON: Absolutely. I think the really cool thing here is that this is one of the first times where we’ve looked at a transiting exoplanet and, using this really huge dataset, we’ve really been able to explain what we’re seeing from the start. There’s another paper that came out at the same time as these Hubble observations where they used 3D climate models and basically just put a big ball of gas near the star and let it go and see what it would do. From that very basic start they pretty much predicted the features that that we're seeing in the data. That’s a really cool thing to me – we’re at this point where we’ve learned so much from looking at giant planets that we now have this basic understanding of how systems work from scratch. That helps us know what we should expect for future data.
TKF: One of our viewers would like to know how common these types of large Jupiter planets are compared to planets like Earth. Do we know that yet?
BERTA-THOMPSON: I can answer that for close-in planets and I’ll let Bruce and Marie-Eve speak about farther out planets because they’re much more the experts there. About one percent of stars have a very hot giant planet in a very short orbit. There are still some debates over that number, but we know they’re pretty rare. Yet they’re easy to observe, so that's why we focus a lot of our attention on them so far. But as far as planets with wider separations go...
MACINTOSH: It’s interesting to hear it described that way because I’ve been in this field longer than my young colleagues and so I remember when the only planets we knew of were these hot Jupiters because they're so easy to see. So it was an astonishing surprise that they were only as common as one or two percent of stars having this class of planet. From missions like Kepler, now we know that this is not the typical planetary system by any means.
What we still don't know is the number of planets that are even remotely like Earth; those are still just below the sensitivity of the instruments we use. The Kepler Mission has shown that there's a huge number of planets out there that are maybe two or three times the size of the Earth. There's nothing like that in our solar system: there's the Earth and then there's Neptune, which is about four times the size of the Earth. So the question of what those in-between planets are, how common they are, and whether they're more like Earth or more like Neptune remains the biggest uncertainty we have in understanding exoplanet populations right now.
We’re trying to push our observations down to the Earth size, but it’s going to take a long time to get there. More importantly, we’d like to make measurements where you don't just tell that there's a planet there but you actually try to say something about what it's made of, to use tools like the spectroscopy to measure its composition. So we know these things are very common, but we really don't know if solar systems like our own are equally rare or represent a big chunk of the solar systems out there.
NAUD: However something we do know – correct me if I'm wrong, Bruce – is that the there are probably many more small planets than there are big planets, right?
MACINTOSH: Below the two-Earth-radius level, whether they start to get rare or stay common is a little fuzzy with the current statistics. But you’re right: These super Earths or super Neptunes are incredibly more common than these roasting Jupiters, which is something that’s encouraging for the even smaller planets.
TKF: Digging into this a little bit more, a viewer asks: How common are Earth-size planets in the habitable zone of red dwarf stars in our galaxy?
BERTA-THOMPSON: That is a very exciting question because, as I think we talked about last time, if you do find a transiting planet in the habitable zone of a red dwarf star then you do have hopes to do these kinds of observations to study its atmosphere and its properties. So that’s a very interesting number. The best work that I know of on this was done by Courtney Dressing, who’s a graduate student at Harvard, just up the street from here, where she used the M-Dwarfs that were observed by the Kepler telescope to estimate this kind of number. She’s still revising the estimates as she finishes up her thesis, but the numbers are significant – anywhere from 10 percent to 50 percent of M-Dwarfs have a planet that’s kind of like Earth in its habitable zone. So I won’t quote her final numbers yet, but at the very least they appear to be much more common than hot Jupiters – which is certainly very encouraging.
TKF: So it sounds like Earth-size planets are maybe just below the sensitivity level of what we can see right now. But if we were hunting for Earth’s twin, what are the main characteristics that we would be looking for? What would you hope to see on a planet that is a lot like Earth?
MACINTOSH: Maybe the first thing you would want to measure is just something as simple as its density to see what its bulk composition is. A lot of planets are turning out to have very low densities, which means that they're not mostly made out of rock and iron like Earth but instead are made of ice, water, hydrogen and so on. So the very first thing you want is just use the combination of the transit technique and Doppler technique to measure the planet’s mass and show that its density is consistent with a rocky planet like Earth with a real surface.
NAUD: Then what you would probably like to know is the basics of its atmosphere composition. You would like to know, for example, if there is water, if there are other gases that could indicate that there is something going on – like life, for example – in the atmosphere. But I guess anything that you could know about the atmosphere of this planet would be really interesting. The fact that there’s an atmosphere at all would be interesting!
BERTA-THOMPSON: Absolutely. At just the very simplest level, some of the first things we’d probably be able to measure would be how big the planet is. We’d want something just about the same size as Earth. We’d also want to measure how hot the planet is, and we could tell that by how far it is away from its star. Those are things that are relatively easy to measure, but they would allow us to start to say that this is really an Earth-like planet.
TKF: If we were to turn that question around and say that we’re looking at Earth from say, 100 light years away – in other words, the kind of this distance from which we’re looking at these exoplanets – what would we be able to tell about Earth? Would we be able to tell that it can support life?
MACINTOSH: It depends on what instruments you’re looking at it with. With the telescopes we have right now – with Kepler or the Hubble Space Telescope – we wouldn’t even be able to tell Earth is there. You can only see these Earth-size planets when they're close to relatively small stars and the geometry happens by an astonishing coincidence to line up so the planet blocks out the star. So to see something that's really like Earth around a Sun-like star even as far as 20 light years away, we’d need missions that don't exist yet but that hopefully will over the next 10 or 20 years. Twenty years from now, it’s not unimaginable that we would have this combination of knowing the radius and mass of the planet and having hints of spectrum of its atmosphere.
People have done really cool experiments where they try to see what the spectrum of Earth would look like by looking at it with another NASA spacecraft that happened to be flying far away from the Earth in our own solar system, or by looking at the light the Earth reflects off the moon. When you stare at those spectra, you can tell relatively easily the presence of oxygen or water. People have done experiments where a NASA spacecraft stared at the Earth from a long way away, and as it rotated they could actually see the Earth getting slightly brighter and slightly dimmer and slightly bluer and slightly greener as oceans and continents came in the field of view. So with a sufficiently large telescope and sufficient time, you might actually begin to see continents on the surface of the planet and that it had both water and dark areas.
BERTA-THOMPSON: Marie-Eve, weren’t you involved in some of those experiments?
NAUD: Yes, I did my masters thesis using a telescope we have here in Quebec called the Mont-Mégantic Observatory. It’s quite a small telescope, with a mirror that has a diameter of 1.6 meters, and we were using it to observe the moon because, as Bruce said, there is sunlight that is reflected from the Earth onto the moon so when you look at the portion of the moon that is not directly exposed to sunlight - that's called "Earthshine" - you can have a good idea of what the Earth would look like from far away.
What is neat is that since the moon is not a perfect mirror, it reflects all the light mixed up and that's great because when we will have the chance to study the light from a distant exoplanet it won't be a perfect picture of the exoplanet either. So when we were looking at the Earthshine, we were able to quite easily detect the Earth's atmospheric oxygen and -less easily - detect vegetation on its surface. Vegetation is especially bright in the near infrared. I don't know if you’ve ever seen a picture taken with a near-infrared filter, but if you haven't, look on the web for a landscape image with a lot of vegetation that was taken with a near-infrared filter. You'll see that all the vegetation comes in very bright. That's something that we hope to see maybe in a distant exoplanet – but this is really far in the future. We were not even able to detect it for sure on Earth when we knew that we were looking at it!
TKF: So we’ve talked a fair bit about hunting planets that are like our own – and we’re getting there with those but we're not there yet – but there's also a huge amount of variation in the types of planets found outside our solar system. I’m hoping you can tell us about a few more exoplanets that don’t look anything like the ones in our solar system.
BERTA-THOMPSON: I think the weird planets that are most exciting right now are the “in-between” planets – the ones that are not quite rocky and not quite gas giants. There’s this really interesting regime of planets that we don’t even know what to call them. Some people call them super-Earths because they’re bigger than Earth and some people call sub-Neptunes because they’re smaller than Neptune. And so understanding the composition of those, I think, is really exciting. Only in the past few years have we learned about these planets.
NAUD: I agree completely because this category of planet seems really common from Kepler data, so it's even more interesting because they don't exist in our solar system. We're not even sure what they’re made of. Are they small gas planets or are they really big rocky planets? So this is a class of planet that interests me a lot. All of the community is interested in new kinds of planets.
MACINTOSH: There’s also an increasing amount of research in the area I work in: studying very young planets. A lot of techniques aren’t good at looking at young systems that are only, say, 10 or 20 million years old, systems in which planets are still forming. But techniques like direct imaging are good at seeing this. Groups have found that young stars are almost always surrounded by disks of dust and gas. Other groups have found evidence of things that might be planets in the process of forming in the middle of disks of dust and gas – in the same way we think of the planets in our solar system formed. These objects are actually sucking down gas so that the planet is swelling up past the size of Jupiter as it absorbs new incoming gas.
With the Gemini Planet Imager, we looked at a star that’s about 10 million years old – still very young – and although we don't see any planets around it, we can see evidence for an enormous asteroid or comet belt 10 or 100 times the mass of the asteroid belt in our solar system. It’s grinding up, making lots of polarized dust. We can also see that the edge of that dust ring is extremely sharp, as if there's something invisible to us keeping the space cleared up, maybe in the same way Jupiter keeps the edge of our asteroid belt defined so sharply.
NAUD: There’s another type of planet that fascinates me from my own research. Also in these young systems, there is another type of planet that’s bigger than Jupiter and that are farther out than Jupiter from their sun. This is another type of planet that we don't have in our own Solar System. You could call them maybe “mega Jupiters.” These are quite fascinating to me too because sometimes they are at a distance where they could not have formed in the normal way. We think of planets forming within the disk of dust and gas that forms around a young star. But these mega Jupiters are usually orbiting at a place where we think that disk was not present or not dense enough to allow the formation of a planet. How did they form? How did they end up there? These are questions that are intriguing about this type of planet that we don't have in our solar system.
TKF: A viewer has asked if there are many Sun-like stars that have been studied and, if so, whether those stars seem to have solar systems like ours.
MACINTOSH: That’s unfortunately still a question that we quite can't answer because our techniques can’t see a solar system like our own. There's been more emphasis on studying the M-Stars lately because it’s easier to see smaller planets around them. Kepler does provide statistics on planets around Sun-like stars but they still it doesn’t quite reach the point where any of the planets in our solar system are detectable.
It’s going to take future missions to really probe the Sun-like stars in depth. We're just barely at the level where we could see Jupiter around a Sun-like star using ground-based telescopes and Doppler measurements. We’re just getting hints of it.
One of the problems in this field is that to see a planet with most techniques, you have to wait until it goes all the way around its star once or ideally twice. Jupiter that takes 12 years each time and if you want to see three orbits of Jupiter, Zach and Marie-Eve are probably young enough that they’ll manage to measure that, but many of the astronomers in the field are not.
BERTA-THOMPSON: Marie-Eve and I should have started a while ago then – we’re already behind!
NAUD: The very first measurement of an exoplanet using the Doppler technique was in the 1990s -1995, to be exact - so we couldn’t have seen much more than one full rotation anyway. It's a pretty young research field.
BERTA-THOMPSON: Right. Absolutely. In terms of Sun-like stars, I would say we’ve looked at them a lot but, as Bruce said, it’s really hard to determine if they have Earth-like traits. But we have been focusing a lot on smaller stars, because that’s where Earth-like planets are easier to find.
If we’re interested in really figuring out whether a planet has life on it – something that’s still very far down the road – it would be a little bit easier if we were talking about an exact twin of Earth, something that's just like Earth around a star that’s just like the Sun, where we feel like we know what we’re dealing with. Whereas if you put a similar planet around an M-Dwarf, a much smaller star, it's really a different kind of system that you’re dealing with.
But if we do someday have the capability to detect the presence of life on another planet, I think it would just be so interesting. It would be such a cool experiment to have planets around these much smaller stars. And if we could actually find that there is life in very different environments than Earth, than that’s just a really cool test. I feel like we will have learned a lot about what you need for life if it’s possible to populate all these diverse environments.
TKF: To that end, we have a number of telescopes that are coming online soon or that are in the works. There’s the Transiting Exoplanet Survey Satellite, an all-sky survey looking for rocky worlds around closest bright stars, and then there’s the James Webb Space Telescope that Bruce mentioned, which will look at those same planets in more detail. So I'm hoping you can tell us a little bit more about what these telescopes will and won't be able to tell us.
BERTA-THOMPSON: TESS is doing basically the same thing that Kepler did, but in a slightly different way. Kepler was designed to tell us about the statistics of planets, so it looked at one very small patch of the sky. TESS, on the other hand, is going to look at all the brightest stars across the sky. That means it will find not just lots of planets, but it will find lots of planets that are easy to observe in more detail later. The closer the star is to us, the brighter the star is, the more light we detect from that star, and the easier it is to study a planet that orbits that star. So TESS is going out to find all the really easy to observe transiting exoplanets, which will give us the right exoplanet systems to answer our questions about the masses of small planets, the atmospheres, all of the evolutions, all of these cool things that we want to know about. TESS is really a recon mission, and it’s going to tell us where we should point the other big telescopes that are coming online.
BRUCE MACINTOSH: And then, after TESS finds them, in principle the James Webb Space Telescope can study their spectra using techniques similar to the ones used in the Hubble paper we were talking about earlier.
I also want to sneak in a mention of the next mission astronomers are starting to plan for after JWST, and that’s something called WFIRST-AFTA. That incredibly boring acronym that stands for Wide-Field Infrared Survey Telescope – Astrophysics Focused Telescope Assets. But it has an interesting backstory. This telescope will do the direct imaging, which has the advantage that you don't need to wait for the 12 years to see a Jupiter. You can just tell it's there and study its orbit.
But direct imaging really requires a big telescope – there's almost no point in doing it with a telescope smaller than the Hubble Space Telescope. And five years ago, as astronomers and NASA studied what their future space missions could be beyond JWST, they didn't think that they could afford an exoplanet-hunting mission that was the size of Hubble or bigger. The highest priority project selected actually had nothing to do with exoplanets; it was to study dark energy and cosmology in the expansion of the universe, something for which you want a huge field of new telescopes studying billions of galaxies at once. As they were maturing the design of this, though, another government agency that will remain nameless came along and said that they had a bunch of Hubble-size telescopes that they weren’t using sitting in a warehouse somewhere. They donated two essentially Hubble-size space telescopes to NASA, and NASA decided that, because they were designed in part for wide-field science, they would be a beautiful match with the dark energy WFIRST mission. But they’re also finally big enough that it's worth considering flying an advanced coronagraph.
And so studies are going on right now to decide if it's feasible to attach instruments to this other agency’s telescopes, potentially including a coronagraph that probably couldn't quite reach the level of seeing Earth-like planets but would be able to see the Jupiters around solar-type stars, would be able to see the super Earths around stars and begin to determine something about the their atmospheric composition.
TKF: Bruce, these are instruments that otherwise might have been looking down on Earth instead out toward the rest of the universe, right?
BRUCE MACINTOSH: I think that’s a reasonable interpretation. The other agencies won’t say what they were going to use the telescopes for, but everybody’s inclination is to guess that is what they were built for.
TKF: So if we are going to see and understand exoplanets better, obviously we’re going to need better tools like the ones you just described. But we’re also going to need really skilled observations. Bruce, in the past you’ve said that the observing process itself is the best part of your job. I’m hoping you can set the scene for us and tell us a little bit about what it's like when you go about planet hunting.
MACINTOSH: In my case, these days I’m mostly going to the Gemini Observatory in Chile, so the process starts with agonizing long plane flights and connections and four-wheel drives up twisty mountain roads. Here's a picture of what you see when you get there. Modern telescopes are the most beautiful machines that I have ever seen – well, all telescopes are, really. These incredible pieces of scientific equipment comparable maybe to the big particle accelerators except in much, much cooler settings, on the top of lonely mountains.
On the left you can see the Gemini telescope itself with our planet finder bolted to the back of it and on the right is a photo of the group. A typical day involves trying to sleep all day, then waking up around dinnertime, having dinner, going up to the telescope, getting all the equipment ready, and watching the sunset. A typical night mostly involves sitting in a room staring at computer screens, unfortunately. But staring at those computer screens, you get to see little images that you can look at and say that’s an actual planet. With your eye staring at the computer screen a little bump represents a planet, orbiting around another star. We’re very happy and enthusiastic about it. And we analyze this data in daytime, work to improve the instrument, work to interpret the results we’re getting, working to understand the stars we’re looking at, but also occasionally going outside to actually look at the stars and remind ourselves that it's not just all about the computer screens.
TKF: And Marie-Eve, you’re actually in Hawaii right now to do some observing, is that right?
MARIE-EVE NAUD: That’s right. I’m at the Canada France Hawaii Telescope, or CFHT. As I said before, I had the chance to use the Mont-Mégantic Observatory in Quebec, but right now is actually the first time that I'm actually on-site at a major observatory. It’s incredible for me to see how things work here. Something that’s been quite surprising is that you can go up on the Hawaiian mountain here – which is called Mauna Kea and which also hosts another great instrument, Gemini North, which is the little twin of the telescope Bruce just talked about. However, all the observations for CFHT are actually made from a town called Waimea, which is at the bottom of the mountain.
It's all going very smoothly it's really impressive for me to see it works. It’s a bit like Bruce described – you try to sleep during the day because by night you have to make the most of that dark time, the time where there is no Sun. Here in Hawaii we are blessed by a wonderful weather – especially compared to the weather we have in Quebec! – so it’s incredible the number of hours we actually get to observe at the telescope. We’re also at such a high attitude that we don't have a lot of the weather patterns interfering with us; most of the time the clouds are below the mountain so you can see them when you are at the top and so it’s great. When you go high enough it's always a clear sky!
TKF: Marie-Eve, thank you for being up during the day for this call! Zach, what about you? Do you have a most memorable observatory moment that you can tell us about?
BERTA-THOMPSON: I’ve been to a few different observatories and I always love it. It’s such thrill, this excitement of getting yourself on a nighttime schedule and staying up late taking data and seeing things that nobody's ever seen before. You know that you’re probably taking the best pictures ever before taken of this star, or the best spectra. You know you’re doing something new and so that’s really exciting.
While I was a grad student I did a lot of work on a robotic telescope survey, and that’s kind of different. The telescopes observe without you, observing every single night whether or not you are there. So there were some less glamorous moments where my interaction with the telescope was mostly to go shoo the ringtail cats out of the telescope dome and clean up the little gifts that they’d left for me. So at those moments maybe observing wasn’t the best thing in the world, but most of the time it really is a thrill, especially at these very big telescopes where you’re seeing these incredible feats of engineering, making first observations happen. The skies are just incredible – it’s just so amazing to be able to go out, look up and see the whole universe above you.