Frontiers and Controversies in Astrophysics

ASTR 160 - Update 1 - The Kepler Mission

Professor Charles Bailyn: It's been over five years since the lectures for Frontiers and Controversies in Astrophysics were recorded. In the wide world, many things have happened in that time: we’ve had an economic meltdown; Barack Obama was elected President; my beard is starting to get grey. I'm not actually all that excited about that, but what I am excited about is the thousands of papers that have been published in the astronomical literature since that time.

In one of the first lectures in this course, I said that--I kind of boasted--that we would be exploring research that has happened within the past week, and that was true at the time, but it’s been almost 300 weeks since then. And so, it’s now an opportune moment to try and update some of this information, and I'm very grateful to the people at the Open Courses, Open Yale Courses program, who have allowed me to give a few update lectures, bringing the material that was presented in this course up to date from 2007 to 2012.

This is the first of three of these update lectures, and this one pertains to the first part of the course, which had to do with exoplanets--planets around stars other than the sun. And, so what I want to do now is try and bring you up to date about what’s happened the past five years on that particular topic. The state of play as of the 2007 is kind of summarized in these two plots, here. There are two main methods of discovering planets around other stars, both have to do with looking at the star rather than looking at the planet itself.

This exemplifies the so-called radio velocity measurement where you look at the Doppler shift of the star and you watch the star coming back and forth, first coming towards you, then going away from you, then coming towards you again; and the star is making that motion because of the reflex motion from the motion of the planet around the star. And so the star doesn’t move very fast because it’s much more massive than the planet itself, moves only 50 m/s or so, which in astronomical terms, is actually quite slow. But, from how fast it moves, you can tell how massive the planet is, because the more massive the planet is, the faster the star will move back and forth.

And so this was the first example of such a radio velocity curve of a star called 51-Pegasus, which was discovered--this was discovered in 1995. The orbital period here is very short--just over four days; the amplitude is such that the mass of the planet is something like Jupiter. It must be very close to the star in order to have that short orbital period, and so, these kinds of planets are called “Hot Jupiters”, because they’re Jupiter-like planets in really close to their star. And, the discovery of Hot Jupiters was this big surprise because no such thing exists in our own solar system, and that started the current boom in exoplanet research.

The other method of discovering exoplanets is exemplified here; this is the transit method. The idea here is that the planet will go in front of the star once in each orbit, and when it does that, it blots out a little bit of the star. And so, the star gets just slightly dimmer. You may not be able to read this scale here. The dimming effect here is about 1.5%, between 1 and 2% of the light of the star is obscured--that tells you something about how big the planet is, because the bigger the planet, the more light it obscures. This particular set of data is extremely accurate, as you can see, the distance from here to here is 1.5%, and so these points are really very accurate. That’s because they were obtained with the Hubble space telescope above the atmosphere, and so you could get very high accuracy.

This particular planet is also a Hot Jupiter; it’s got the radius of Jupiter, much bigger than that of the Earth, and yet, it only obscures just over 1% of the light of the star. So, these are the two primary methods of discovering planets around other stars. In one case, you discover the mass of the planet. In the other case, you discover the size of the planet; and there are examples where both have been done, and in that case, you know both the mass and the size, and so you can deduce the density. And that tells you something about what the planet is made of, in that rocky planets like the Earth have a much greater density than gaseous planets like Jupiter and Saturn and Uranus and Neptune. And so, indeed it turns out the Hot Jupiters are mostly gas-type planets.

So, one of the primary goals of exoplanet research has always been the discovery of Earth-like planets. Fundamentally, what we’re looking for is planets that can support life. Life, as the science fiction writers would say, “life as we know it”, carbon life, complex forms of life. It's thought that this is likely to occur in situations with intermediate temperatures. Not too cold--not so cold that everything freezes solid; not so hot that everything boils over, but perhaps temperatures appropriate for having liquid water, as we have on the Earth. And so, the goal has been to discover planets that are similar to Earth in their mass and their size, and therefore, their density. The kinds of planets that you might imagine to have a rocky core with an ocean and an atmosphere the way Earth does, and you want them to be in what is referred to as the “Goldilocks Zone.”

This is the part of the solar system that is not too cold, not too hot, but just right for life. And in order for that to be the case, if the star is a star more or less like the sun, the orbit has to be about the same size as the orbit of Earth. And therefore, the orbital period has to be something like 1 year. And, the idea is, it can't be a Hot Jupiter with only a four-day orbit, it’s got to be a year long. And so, there’s been an effort to try and find planets that look like Earth in the Goldilocks Zone.

And as of 2007, this was not technically feasible to do, because the longer the orbit, the harder these things are to find; the lower the mass, the harder they are to find; and planets with orbits of a year and masses and sizes similar to that of Earth, were simply impossible to find. And, the thing that has happened since that time is the development of instruments which are capable of discovering Earth-like planets. And the key instrument is a space telescope known as Kepler, named after the great astronomer of several hundred years ago.

The Kepler mission, launched in 2009, was designed explicitly to discover Earth-like planets in the Goldilocks Zone, and so, what this telescope does is it’s a series, a whole group of space, of cameras behind this space telescope. Each one of them imaging an adjacent part of the sky, looks at this little part of the constellation of Cygnus, and it’s going to look straight at that piece of the sky for about 5 years. And, the idea is it’s going to take a little picture of this part of the sky every minute for the next 5 years. And, the idea is that in this patch of the sky there are maybe 100,000 stars. And every time a planet goes in front of one of those stars, you’ll see one of these little transit dips; and if you keep track of them every minute of every day for 5 years, then, if you have a 1-year orbit, you’ll see 5 such dips, and you can confirm the existence of such a planet. And, it’s specially designed to get enough precision that you could see not just the 1% dips that would come from a planet like Jupiter, but the 1/100th of 1% dips that would come from a planet as small as Earth.

So, in the first 10 days of the mission--as I said, it was launched in 2009--here’s what they discovered: this is a plot of observations from the ground of one of these Hot Jupiters, and you can see there's a lot of scatter in these points, but they’ve taken a great many of the points, and you can tell that between here and here, the general place where the points are is a little bit lower. I should say, each one of these tic marks is about 1% of the light. And so, you can sort of tell that as you go along here, there’s a lot of scatter, but once you get down here, it’s maybe 1% fainter for a little while and then it continues. And this line here is a model of how a planet should behave, and so, they inferred the presence of one of these Hot Jupiters.

Here’s what this exact same star looks like if you observe it with Kepler, and these dots here, which are indistinguishable from the line, are actual observations by the Kepler satellite. And you can see how much better it is to be observing from space than it is to be observing from the ground. This is really, kind of perfect data here, and then these bottom two plots here are the exact same data, just blown up a little bit. This is blown up by a factor of seven, so you go from here to here, that’s just under 1% of the light, and you can see how accurately it follows the model. And then, if we blow it up by a factor of 100, then this primary minimum here goes way down into the basement somewhere, and you can’t see it, but you can start to see other interesting effects.

First of all, there's a kind of general wiggle over the whole thing; there's a little more light here than here. That’s because, in this case, you really are seeing light from the planet, and what you’re seeing is the phases of the planet just like the phases of the moon. If this is the star and the planet’s in front of the star, then you see the back side of the planet and you don't see any additional light from the planet itself because the light you would see would be reflected from the star.

A quarter of an orbit later, here it is. Here, you’d see kind of a half planet; you’d see half of a planet, see this side lit up, that side in the dark, and you’d see a half planet. Here, you’d see a crescent, and as it gets towards this end, you see more and more and more of the planet because the phase of the planet is going from sort of new, to crescent, to half, to full. And that extra length reflected off the planet is why it’s getting brighter between here, where the planet is between you and the star, and here.

Then, an interesting thing happens at the exact opposite phase, namely. The planet goes behind the star, and so, all that extra light from the planet disappears for a little while, while the planet’s behind the star, and then it emerges out the other side. And you can see that you can see that little dip, and that dip was an important thing, not so much having to do with the light of this particular planet, but because that is the size of the dip that you expect from something the size of Earth when it does its primary transit. And so, this is a Hot Jupiter, it’s a very big planet, it’s got a nice big dip, but dips this size, which clearly can be seen are the kinds of things you’d expect from an Earth-sized planet, and so, within the first ten days, it was clear that we were going to be able to see dips from Earth-sized planets.

So, that was the first ten days of observations from 2009; they’ve been at it for a little while now. Here are some of the results as of February of this year. They’ve discovered almost 2,000 candidate planets--when they say candidate planets, what they mean is they’ve seen 1 or 2 dips. They don't count it as an official planet until they see 3 or more dips all evenly spaced, but here are the candidates, and from each one of these dips, you can figure out how big the planet has to be. Some of them are Jupiter sized, there are a couple hundreds of those. There are 1,000 or so planets discovered that were Neptune sized; before this began, there were only 400 exoplanets known at all. They have 600 that are within a factor of two of the size of the Earth, and they have already about 200 planets the size of Earth itself. And so, this was- this has been a huge achievement, it’s greatly increased the number of planets known.

Here’s a little picture gallery--these are now confirmed planets, these are ones they’ve seen many dips of, and you can see they’ve rank-ordered them in terms of size here, and they’ve stuck a couple solar system objects in there just so you know approximately what’s going on. Here’s Jupiter, and so these guys are all bigger than Jupiter. This is Neptune in our own solar system. Down here is Earth, and here’s one example of a planet that’s actually smaller than Earth that they’ve discovered. And so, this is kind of the family gallery of planets, and they keep adding to this picture.

Here’s a different representation of the planetary candidates: these purple ones up here are the ones that were known before Kepler, and what’s plotted here is the diameter in terms of the diameter of Earth--so, these guys are about 10 times bigger than the Earth--versus the orbital period (how long it takes to go around).

And so, what we want is things with one Earth diameter with an orbital period of 365 days, which would be out here somewhere, so they haven’t quite gotten there yet. And, the purple--the dark purple dots here are the first set of Kepler candidates, the yellow dots are the current set of Kepler candidates, and you can see--there is a fair number of these things down below the size of the Earth.

Now, I want to give you an example of some of the particular objects they’ve found. They’ve found some very spectacular things, here’s one. This is the field of view of Kepler, and if you look at this particular star, what they discovered was six planets around this one star. This is a very confusing star because they found a lot of dips, but the dips didn’t look like each other. Some of them were deep, some of them shallow, some of them took longer than others.

So, they finally figured out there were six different kinds of dips happening in this system, each one associated with a different planet, and once they finally got this sorted out, they could plot the orbit of all six of these planets. And so, there's a piece of NASA propaganda for what this solar system looks like--three of the planets are having transits at once. That was part of the problem in analyzing this data, you had dips on top of dips on top of dips, and the other three are going to be transiting sometime soon. And, needless to say, no such picture exists; this is the kind of thing that the NASA artists concoct down in the basement to show--to give an visual impression of what this solar system must be like.

Here’s the size of these orbits, and the interesting thing is all six of these planets are well inside the orbit of Venus in our own solar system, they all have very short orbital periods. That’s not a surprise, because they’ve only been observing these things for about three years, and if you want to have multiple repeated dips, then obviously, the planets you’re going to discover are going to have periods that are small compared to the length of time you’ve been observing them. So, you couldn’t observe something like Jupiter, which has an orbital period of 10 years if all you’ve got is three years worth of data because you’d have, at most, one dip and you wouldn’t be able to confirm the orbital period at all.

So, here’s the summary of the Kepler-11 system, as it was announced in Nature a few years ago. It's got six planets; and it’s the closest-spaced solar system known at the time. These planets are all more or less mid-sized, a little bit bigger than the Earth, but a lot smaller than Jupiter, and they’ve now measured the densities of all of these, and it’s odd. Some of them are solid, rocky planets, some of them are fluffy gas planets, there’s a wide variety.

And so, now let me fulfill my promise here and tell you something that happened this week. As I'm speaking right now, it’s July the 2nd; on June 26th, they announced the latest in Kepler planets. This is a system #36, and they found two planets in this system, both with very similar orbits--14 and 16 days--so, they’re orbiting one right inside the other. And, the inner planet has a high density; it’s a rocky, Earth-like planet. Little bit bigger than the Earth, one and a half times the size of the Earth, but the outer planet is significantly larger, although not significantly more massive, and therefore has a much lower density, it’s made largely of gas sort of like Neptune. And so, these planets are quite different in kind, but in a very similar orbit--one right inside the other--and when they come close to each other, when one passes the other, they’re really quite close. Only five times further away than the Earth is to the moon, and much closer than the Earth is to any other planet at any other time.

And so, the NASA artists got very fired up about this and drew pictures. This is what they imagine the Neptune-like planet rising up over the horizon of the Earth-like planet. It actually isn't all that much like Earth because it only has a 14-day period, so it’s really quite close to the star, very hot, lots of lava and volcanoes--they imagine; we don't know this for sure. But, if you could sit on that planet and watch the rise of the other planet, this is the kind of thing it would look like, and it would be, you know, substantially larger than the full moon, as seen from Earth. So, quite a spectacular system.

Now, you don't need to take the astronomer’s word for it, you can do this yourself at home. One of the things that they’ve done with the Kepler data is, they’ve put it all online so that people can go and find their own planets, and there is a website called planethunters.org, and if you go, you can log on and they will flash up a whole series of Kepler data for you. This is a plot of brightness of some star versus time. And you can see that this star is a variable star, it kind of gets brighter and fainter and wiggles around, but you can also maybe see that there are moments where it gets--suddenly--a little bit fainter, and these are evenly spaced in time.

So, here’s a situation where you have a variable planet--a variable star with a planet crossing in front of it. Now, it turns out the computers find it kind of hard to distinguish this kind of variability from this kind of variability. Human beings have no such trouble; we’re really good at recognizing patterns, and so, you can put all kinds of wiggles on a star, and if you say, well, look for the little equally spaced dips, no human being is going to have any trouble finding these. And so, this is one, in fact, that was discovered by a member of the public rather than a professional astronomer on this planet hunter site.

This is a newly popular thing to do with big, scientific data sets. The idea is to make it available to the public, have some kind of educational website which tells them what to look for, and all kinds of people will sign up to look for all kinds of things. And there's enormous sort of blog discussions and things about the kinds of things that they found. The galaxy people are doing the same thing with hundreds of thousands of images of galaxies. And so, it turns out that by crowd sourcing the data analysis, you can actually get a lot done.

So, if you want to find planets on your own, you can go to planethunters.org, sign up, and discover the next Kepler planet. So, there’s a lot of excitement around Kepler. They’ve found thousands of new planets, they found all kinds of exotic planets of kinds we didn’t expect to exist, all kinds of solar systems of multiple planets interacting with each other. It turns out, if you crowd source it, then the population at large will find even more planets for you, and we can, I think, confidently expect that every few weeks or months, there’ll be a yet more spectacular system announced by the Kepler people. But, there’s still one thing missing, and that is the true Earth-like planets. We found things that are as small, or smaller, than the Earth. We have also found things that are in the Goldilocks Zone. That is to say, having one-year orbital periods. By now, the mission’s three years old, they’ve seen a few things out there. We have not yet found something that is both as small as the Earth, and in the Goldilocks Zone. So, what’s going on? Is Earth perhaps unique? Well, probably not; we have discovered something. What we’ve discovered is that most stars aren't actually like.

[off camera noise]

We have discovered something. What we’ve discovered is that most stars are not, in fact, like the sun. Most stars are a little more active than the sun. They have more sun spots, bigger sun spots, flares, various kinds of activity. And when you have that, you can have a coincidence where a sun spot appears on the surface of the sun--that makes the star look a little dimmer--and if it just so happens that another spot appears at exactly a year later, and a year later than that, or whatever, they can accidentally line up and give you the impression that there’s an Earth-like star--an Earth-like planet around that star.

And so, if you see three regularly spaced dips, remember we’ve had a three-year mission so far. Then, it could be a planet, but it could also be three sun spots that just happen to show up at the right time. Every time you see an additional dip at the right time, it becomes less likely that it’s a sun spot because each additional dip has to be another coincidence lining up the sun spot to just the right time. So, the feeling is, that by the time you get seven, you’re in the clear. No combination of sun spots or other kinds of activity are likely to give rise to that. But that means that the mission had better keep going, because you won't be able to detect Earth-like planets with one-year orbits until the mission is seven years long, rather than the expected three to five.

And so, they petitioned to NASA to keep the mission going for longer than expected, and that actually was approved in April. And so, we now have the Kepler extended mission, and it will indeed go on for seven or eight years, looking at the exact same stars once a minute for many years in a row. And so, perhaps four years from now, in the sixth and seventh year of this mission, we will finally get the announcement we’ve all been waiting for: that there is a rocky, Earth-like planet in the Goldilocks Zone around one of these stars. And that will be the next step on the route to identifying planets that have the conditions appropriate for having some kind of life on them. And so, Kepler’s been this huge step forward. Thousands of new planets, all kinds of new things. We haven’t quite gotten to the point where it fulfilled its original mission of finding Earth-like planets in the Goldilocks Zone, but we certainly expect that to happen sometime within the next few years.

[ end of transcript ]