 A poor friend of science and also the ability to communicate these results both with the scientific community and with the general public. Dr. Allen Boss of the Carnegie Institute of Washington exemplifies both of these qualities. During his academic career, Dr. Boss has focused his attention on the formation of stars, solar systems, and planets. One of his first research topics was the modeling of the collapse and formation of stars from inner stellar gas. The models developed were tremendously important because they went well beyond the earlier qualitative models of stellar formation that were common at the time. He has played a very influential role in the recent discovery of planets outside of our solar system. As a member or chair of many NASA, National Academy of Science, and National Research Council committees, he has helped to shape the research program for the search for extra solar planets. Before the 1995 discovery of the first planets orbiting other stars, his computer simulations predicted the size and relative location of where these Jupiter-like planets were expected to be formed. Since the discovery of several planets of this variety, he has played an influential role in analyzing and interpreting these results. In addition to over 100 publications in academic journals, Dr. Boss has been actively involved in communicating these results to the wider audience. He has written several review articles for Scientific American, Nature, Physics Today, and Encyclopedias. He has also just recently completed a book for the mass book market. Past few years, he has frequently been quoted in the New York Times, Time magazines, many magazines, and newspapers. He has also given numerous radio and television interviews and invited talks. Dr. Alan Boss is one of the most influential scientists in the field of formation of the solar system and planetary systems. It is my great pleasure to welcome him here today for his talk, Forming Solar Systems Here and Elsewhere. Thank you, Tom, which you very graciously didn't say was it my prediction about locations of planets were spectacularly disproved by the very first planet that was found. I missed the position by about a factor of 100, which is terrible even by astronomical standards, but you'll find out more about that as the talk recedes. Today, or yesterday actually, you may not know this is actually a very significant anniversary in the history of the advance of science, because on October 6, 1995, the first robust discovery of an extra solar planet was announced in Florence, Italy. And that event has propagated us into what is easily called a new era in science, because we have now entered, we are now two years into in fact, the era where we will discover more and more extra solar planets and begin to not only discover them, but to characterize them and find out whether or not they can possibly be abodes for life similar to Earth. So what I want to talk to you about today is, first of all, give you a little bit of the background on how the history of extra solar planet searches has come about and brought us to our present level of knowledge. I'll tell you a bit about the methods that are used to detect extra solar planets. And then I will put that all into the context of how we believe that extra solar planets form as well as the planets within our own solar system. To try and see whether or not the theory agrees with what observations have told us. And towards the end of the talk, I will give you a glimpse into the future and let you have a peek at some of the telescopes that NASA intends to build that will allow us to satiate our curiosity about planets around other stars. But first let me point out to you why it is so hard to find extra solar planets. This is shown here on this plot which shows the amount of light which comes from the Sun on the top curve as a function of wavelength versus the amount of light that comes from the Earth, Jupiter and Uranus. This is plotted on what is called a log plot such that each unit on this scale on the left corresponds to a factor of a 10 difference in luminosity. So you notice first of all that at visual wavelengths which occur here around say 5000 angstroms, the Sun is about a billion times brighter than the Earth. At infrared wavelengths it's not quite so bad because the bodies have their own thermal emission which is emitted and reduces this discrepancy and brightness down to about a factor of a million. But even a factor of a million is of course a very hard factor to overcome. And that's why it's exceedingly difficult to contemplate detecting extra solar planets directly. It's not to say that it's impossible but it is difficult. It can be done. And the instruments I will tell you about towards the end of the talk will actually culminate in an instrument which will attempt to detect extra solar Earths directly. In the meantime I will tell you historically about two other methods that have been used to look for extra solar planets. And they rely on not direct detection that is detection trying to find photons to come from the planet itself. But there are so-called indirect detection methods where you infer the presence of the planet by looking at the effect of a planet on the star on which it orbits. The very first of these methods is called the astrometric method. Historically it's the one which was first developed. And it relies on a very simple principle namely that as a planet and a star orbit each other they orbit actually around their common center of mass at the center of the system. So while the planet orbits around generally speaking a much larger radius than the star does the star can still move across an appreciable distance. It's not terribly large however. In fact in the case of our solar system Jupiter makes the sun move back and forth over a distance of only a radius of about 400,000 miles or a distance about equal to the radius of the sun itself. And in fact if you look at what all the plants of the solar system do to the sun's position over about a 65 year period of time shown here you'll notice that the sun does indeed wobble back and forth over a distance similar to its radius. And you'll notice that there's sort of a 12 year periodicity to this about every 12 years basically it returns to the same point. And what that's telling us is that there's a Jupiter mass planted in the system. Jupiter has a period of 12 years that makes the sun wobble back and forth every 12 years. But there's more information in this diagram than just that. There's also something else going on namely Saturn is also making the sun wobble back and forth. And if you could follow the motion of the sun with enough precision you could disentangle the motions and know that there was not only a Jupiter with a 12 year period but also a Saturn mass planet about three times lower in mass with an orbital period of about 29 years. So the goal of astrometry is to be able to measure the position of a star on the sky to such exquisite accuracy that you can actually disentangle these motions and infer the presence of planets around it. Now they give you some feeling for the degree of accuracy you need to achieve if you have a system that you're trying to measure that's about as far away as say the nearest stars which are on the order of perhaps five to six light years away where one light year of course is the distance that light travels in a year it's also equal to six trillion miles. At that sort of a distance the angular movement of a sun-like star orbited by a Jupiter mass planet is about one milli arc second. Now let me explain what that is. Of course in a right angle you have 90 degrees. You take each of those degrees, divide them up into 60 minutes, take each of those 60 minutes, divide it up into 60 seconds and then take a thousandth of that second. That's what a milli arc second is. It's an excruciatingly small angle. It's about the angle that a dime would subtend at a distance of a thousand miles. But believe it or not astronomers believe they can measure that and they believe they've been able to measure it for some time. And this is the person who really is the master of the astrometric technique. I should say he was the master of the astrometric technique. Peter Vandykamp. He was educated in Holland, came to the United States as a young man and was a professor at University of Virginia for many years then went to Swarthmore College in 1937 and was made director of the observatory there. The next year he began a very serious program to look long term for low mass companions to stars. In 1938 he began that program. One of the first stars that you put on that program was a star called Barnard Star. Barnard Star had been discovered in 1916 by Edward Emerson Barnard who was an amateur photographer who was so prolific at finding comets that he was offered a permanent job as an astronomer at the first Lick Observatory and then Yorkies Observatory. And Edward Barnard had found more or less by chance a star that was literally screaming across the sky. It was moving very quickly across the sky and when something is moving very quickly across the sky that means it has to be very close to you. It was moving about 10 seconds of arc per year and that's a movement about 10,000 times larger than that one milli arc second movement that a planet would induce in the position of the star. So that's really a very large ratio. When he found this motion people immediately figured that this star had to be very close as I said before because if you think of closeness an analogy is an airplane. You know that airplanes have to travel at several hundred miles an hour in order to stay in the air yet if you see a distant airplane it seems to be hardly moving. It seems to be almost stationary. If an airplane of course is very close to you you realize it's moving very quickly. So when astronomers find a fast-moving star like Barnard's star they immediately rushed out and measured its distance and they found sure enough that Barnard's star was a very close star. It turned out to be the second closest star to the Sun after the Alpha Centauri, Proxima Centauri system. So the distance of about six light years away. And the important thing about a close system is that the closer you are of course the easier it is to see that astrometric wobble. If you're twice as far away it's twice as hard to see. So you want to find the closest possible stars. Barnard's star then was a natural to be included in Vandy Camp's program of looking for extra solar planets. It was also a natural because the star itself is rather low in mass. It has a mass about seven times less than the mass of the Sun. And it's a red dwarf star a bit too faint to be seen by eye. And because it has a lower mass itself that also means it's easier for the planet to make that star wobble back and forth. So it's a very ideal system to search for extra solar planets with. Well Vandy Camp and his 50 odd assistants he had working with him through the years accumulated a lot of data on Barnard's star. And by 1963 they were ready to begin announcing their first result. And Vandy Camp showed his data here in 1963 and made a claim that he had found evidence for a first extra solar planet. The planet with a mass about 60% greater than that of Jupiter. What's shown here are his actual data points to divide it up into two directions on the sky. Right ascension and declination you can just think of those as sort of X and Y and a Cartesian coordinate system on the sky. And I should explain the data points to you. The open points are simply the same data points over here just carried forward because he believed that there was a 24 year period to the orbit. And so everything should just repeat. So he was showing the same data points twice on the side of the curve. And from the scatter of this data he believed he had found a periodic trend. He fitted it with a solid line. And the solid line is the motion that would be induced in the star. If Barnard's star was being orbited by this 1.6 Jupiter mass planet. More or less what one might expect perhaps for the first extra solar planet. Something just a little bit bigger than our own most massive planet. It had an orbit about four and a half times the distance of the Earth from the Sun which is quite comparable to Jupiter's distance from the Sun which is about five times the Earth's Sun distance. The only odd thing about it though was that it was moving on a slightly eccentric orbit or a highly eccentric orbit. If it was moving on a circular orbit these plots would be nice sinusoidal smooth curves. Instead they have a little bit of a cusp. So that's the key that it's actually moving on an eccentric orbit. Meaning the orbit is not circular. It's actually like a squashed oval. So this was generally accepted after a few years. New York Times got excited about it. It began appearing in astronomy textbooks towards the end of the 1960s and was quoted as the example of an extra solar planetary system that people had expected should be around and this was apparently the first detection of it. Well all was well and fine until about ten years later when along came George Gatewood, shown here. George Gatewood did a thesis project on Barnard Starr. He did not use any of the plate material that Vandy Camp used however from the scroll observatory. He used plates from the Van Vleck Observatory in Connecticut and the Allegheny Observatory in Pittsburgh. It was a completely separate solution and he tried to recover the same solution that Vandy Camp had done. He did this as a thesis work. The thesis advisor was Heinrich Eichhorn. The two of them were actually doing this work in the upper floors of the physics building at the University of South Florida in Tampa. While I was underneath taking physics classes completely unaware of their existence. This is an amazing case where two careers have crossed in time for no obvious reason. What they found, it's true. I remember him walking around campus vaguely but I didn't know what they were doing. And what Gatewood found is shown here. This is his analysis of the data. He did not have as much plate material as Vandy Camp did. He has ten times less plates but they were measured with not by human being but by a plate measuring machine that had been developed at the U.S. Naval Observatory which hopefully removed some of the systematic errors. And he also has repeated his data. These two big data points here are the same as these two here. The size of the data point is giving you some feeling for how much you should believe them. The larger data points mean you should put some belief in them. This is where you should really follow. The little data points you should follow less. This little piece here you should ignore altogether because that's a speck of dust. That managed to fall on Vandy Camp's orbit. So who says God doesn't have a sense of humor? If you do follow the, let your eye follow the rather large points, you will see that it really doesn't come very close at all to following Vandy Camp's orbit. A couple of them seem to, tantalizingly come close. But overall, Gatewood was not able to confirm Vandy Camp's solution for Barnard Star. The situation for Barnard Star got even worse that year when one of Vandy Camp's colleagues at Swarthmore, John Hershey, published a paper. Hershey had also been working with the same data for Barnard Star. He'd also been following a second star, a Gleesa catalog star, another low mass star. And he noticed when he plotted the positions of Barnard Star and the Gleesa Star that in 1949 they both took a jump in one direction. In 1957 they both took another jump in the other direction. Now, if just Barnard Star doesn't, you say, alright, that means Barnard Star has a planet. If the Gleesa Star is doing it the same time in synchronous, that means they both have to have exactly the same planet orbiting in exactly the same phase, which is just too improbable to believe. What that tells you is that there was a systematic problem with the data. And Hershey was able to isolate what happened in 1949 and in 1957. Those were the years that they took the telescope apart. Sad fact, but this was a refracting telescope. It had to be, it was not like a reflector that you could sort of clean off the surface with a dust of air. You had to take the refracting lenses apart and clean them and then put them back together again. And they thought they did a good job of putting them back together again, but Barnard Star proved that they did not get it back together again. And so at that point the belief in the planetary system around Barnard Star began to fade rather dramatically, as you can imagine. Peter Venicamp himself believed in the existence of a planet around Barnard Star. He believed that he was the only one who had been able to find anything because he had so much data he'd accumulated, 50 odd years of data, and he finally died in 1995 just before the true extrasolar planets were found. Still believing that he had found something, but I think he was one of the few people who believed that along with him. The second method of detecting extrasolar planets relies not on the positional wobble of the star, but on the velocity wobble, if you were, of the star, as it again moves around the center of mass of the star planet system. If you consider, say, a certain spectral line in the atmosphere of the star that emits at a given frequency, say an emission from something like an iron atom in the atmosphere of the star, when the star is moving in a direction transverse to our line of sight over here to the Earth, then we measure that wavelength at exactly its rest wavelength as you would measure it in the laboratory on Earth. However, on part of the orbit around the center of mass, the star is, of course, moving away from us, and that means that that wavelength is shifted, red shifted, towards longer wavelength, and in portions of the orbit it's moving towards us so that wavelength is shifted towards the blue towards shorter wavelengths. So there should then be a periodic modulation of the apparent wavelength of any spectral line you look at. It will wobble back and forth in velocity space. Now, that is not a very large wobble again, as you can imagine. It's not easy to detect. The measure of the shift in wavelength is given by the velocity of the Sun. In this case, if the Sun is being orbited by Jupiter, it will move back and forth by plus or minus 13 meters per second. Sort of a fast run. Very fast run. And then you divide that, however, by the speed of light, which is quite a bit larger, and that gives you a fractional change in wavelength of about a part in 10 to the 7th or 10 to the 8th. If you're looking at visible wavelength light at 5,000 angstroms, that means you're looking for a shift in the wavelength of about a 10,000th of an angstrom. That's, again, a very difficult measurement to make. But if you look at enough stellar lines, and some stars have lots of stellar lines, and you try to average over all those lines, you have some hope of being able to remove any systematic errors you're making and actually get a good solid detection of any radial velocity wobble in the star. The people who finally succeeded in doing that, convincingly, are shown here. These are the pair that won the race to find the first extra-solar planet around a solar-type star. Shown here is Michel Mayor on the right, and Didier Calo on the left. There are two Swiss astronomers from the Geneva Observatory. Mayor is very much a senior scientist in this business. He had spent several decades looking for low-mass companions to solar-type stars, but working with a rather low-precision system that couldn't detect extra-solar planets. But in April of 1994, they installed a new spectrometer on his observatory, which is the Hall-Provence Observatory near Marseille, a rather humble telescope in many senses. It's a less-than-2-meter diameter telescope. The observatory is located only at a height of 2,000 feet or so. But they installed a spectrometer on it that allowed them to measure to a precision spectral line shifts of about 13 meters per second, which is the magic number to be able to detect a Jupiter-mass planet. So armed with that new spectrometer, they went out and started searching for extra-solar planets. And they had something of a head start, because Mayor had previously done a survey to look for binary stars. He had catalogued all the binary stars in the neighborhood of the Sun, and he knew which ones were binaries and which ones were single stars like the Sun. And the feeling, at least at that point, was that the single stars were more likely to have planets, and so he started focusing on all the single stars that he could find. He started that program in April of 1994 by the fall of 1994. They began to get a peculiar signal from one of their stars, a star called 51 Pegasi. They began to think that there was something going on in 51 Pegasi. Unfortunately, about one month later, 51 Pegasi was no longer visible, because the planets are only visible, or the stars are only visible for about six months out of the year. The other half of the year, they're towards the Sun, and you can't see them. They suddenly had to wait for six whole months until July of 1995 for 51 Pegasi to become visible once again. They ran to their telescope, insisted on having eight full nights of observing time, and they sat on that star for eight nights and followed it. And one month later, they were ready to publish, and this is what they found. This is the discovery plot for the first extra solar planet. Here are the variations in radial velocity. These are actually variations, of course, in wavelength. It's been converted to an equivalent velocity shift to make it more convenient to interpret. This has been plotted in such a way that you have one portion of the orbit where the star is coming towards you, so things are shifted towards long wavelength, and another portion where the star is going away from you, and it's shifted the other way. And where the velocity shift is zero, that means the star is moving transverse to the line of sight. So we see that in all the data, this is several different orbital periods, they're all plotted on top of each other to give you some feeling for what the errors are and how reproducible the data is. Some data points aren't too terribly good like this one here, but overall there's a very robust fit to this curve, and it's a velocity variation of about 55 meters per second. The solid curve is the variation you would expect if 51 Peg, which is very much a solar type star, it's very similar to our sun, that's the sort of variation you would get if it's being orbited by a planet with a mass about one half that of Jupiter. It could be considerably a bit more massive than that, you can't really tell exactly because we don't know the angle of inclination to the star. If the angle of inclination is such that the planet is moving more or less in a plane perpendicular to our line of sight towards the star, then the mass of the planet could be considerably larger because the radial velocity technique only picks up the component of the velocity towards the line of sight to the Earth. So if the planet is making the star move back and forth mostly in the transverse direction, we won't know about that, all we know is a little bit that it moves back and forth towards us. So this radial velocity method only gives you a minimum, a lower limit on what the mass is. But since things tend to be randomly oriented in space, the minimum mass is actually not too far off from what the most probable mass is, and most likely 51 Peg is showing evidence for a roughly half Jupiter mass planet orbiting it. However, there was one major, major surprise. That is that the period of this orbit was not 12 years, it was 4.23 days. And that implies that the planet has to be very, very close to a star. In fact, it's orbiting at a distance from its star about 100 times closer than Jupiter is to our Sun, 20 times closer than the Earth is to the Sun. And I believe that Michel Mayor was able to find this because he came into the problem with a fresh mind. He came in thinking not like other searchers that he was looking for a 12-year period solution, but he was coming in looking for binary stars if he could find them. And he knew that binary stars have a very wide range of periods. Some binary stars are essentially touching each other. They have periods on the order of hours. Some of them are so far apart that they have periods, orbital periods where they go around each other measured in tens of millions of years. So he was prepared to look for a wide range of periods. He found them at 51 Peg. Now, what really makes 51 Peg a robust detection is that it was confirmed. This is the first time, of course, that someone has actually gone out and said, yes, you're actually right. And the confirmation was done literally the next week by a pair of enterprising astronomers, Jeff Marcy and Paul Butler, who ran to their telescope. If they look observatory, they had been doing their own search ever since 1987. They were prepared, ready to go. They were not observing 51 Peg, but after Michel Mayor's announcement, they began observing it frantically. They took five nights of data, and over a period of time they built up even more data, and they were able to reproduce essentially the same curve that Michel Mayor did and confirm that 51 Pegasus does indeed have a half-Jupiter mass planet. Here are Paul Butler on the left and Jeff Marcy on the right showing inside the dome of the Lick Observatory Telescope. That's the Shane three-meter reflector in the background that they used to do their confirmation, as well as several of their other discoveries, because they soon realized that they could have beaten Mayor to the punch. If only they had analyzed their data. They had been taking data since 1987, remember, but they were looking for 12-year periods, and the astronomical lore is that you need 12 years of data before you can find a 12-year period, so they didn't bother to analyze their data. They began analyzing their data shortly thereafter. They spent the rest of October, November, and December. I don't think they had a very good Christmas that year, crunching through all of their data. And they struck pay dirt in December, late in December. In early January, they were ready, loaded for bear. They showed up at the American Astronomical Society meeting and presented data for two new objects. One of them, I believe, truly is a giant planet, like 51 Pegasus appears to have. It's orbiting around a star called 47 Ursa Majoris. This is again a solar-type star. You can see their data goes back all the way to 1987. They had taken several data points a year, and you can see what happened here when Michel Mayor announced 51 Pegasus. They got very, very serious about 47 Ursa Majoris, and they nailed it. This is now interpreted as having a perhaps two Jupiter-mass planet moving around it, which again sounds like it might very well be a Jovian analog. And it has a rather long period, though. You see, it doesn't have the 4.23-day period. It's a period of a little over three, around three years, and that's interpreted as having a distance from its star of about twice the Earth's undistance. It's not quite Jupiter's five, but it's sort of more than normal spacing. It's not quite as shocking as 51 Pegasus. It's an extremely tight separation. So this looks like this was the beginning of what has become since then a flood of detections of extrasolar planets. But that wasn't all the excitement that happened in October. In fact, literally on that same day, October 6, 1995, a number of us were sitting at JPL at Stone Among Us. We were listening to the report of a group that NASA had put together to make a plan for looking for extrasolar planets. Of course, little did we know that halfway across the world, someone had actually found one. But we were privileged to hear a report from Sri Kulkarni of Caltech, who had found the first undisputable Brown Dwarf star, shown here in this picture that Ed Stone showed earlier today. Now let me explain to you why you should care about Brown Dwarf stars. Brown Dwarf stars are objects that form the same way that stars do, but they end up severely mass challenged. They end up with too little mass to burn hydrogen for very long on their main sequence lifetime. They can burn deuterium for a short period of time and gives them some energy for a few million years, but they never become hot enough because they're not massive enough to have what's called the main sequence lifetime that our sun is on, where a star can burn and put out prodigious amounts of energy for billions of years. So they're sort of failed stars. They form just like a star does, but they don't quite have enough mass. And that critical mass is about 80 Jupiter masses or about 8% the mass of the sun. These objects had been theorized for quite some time to exist and they had been hunted and there had been false detections that I don't have time to tell you about today, heartbreaking false detections, but they were never actually conclusively proven to exist until this image was taken here. The discovery image is shown on the left. This is actually taken by a rather modest telescope, a 60 inch telescope on Mount Palomar, but a modest telescope outfitted with a clever technological device, an adaptive optics system that allows the system to take out many of the turbulent motions that normally would obscure a sharp photograph, turbulence that's induced by bending of light rays in the upper atmosphere of the Earth. The group had actually taken this photograph one year before in October of 1994, but they had to wait a year to make sure that this was truly physically associated with this very bright star here. This is Gleesa 229 itself. It seems very bright. It's really a low mass red dwarf star. This is the brown dwarf. It's about 44 times away from its star than the Earth is from the sun. They had to wait one year because they wanted to make sure this was really physically associated with the star and wasn't just some chance background object. So one year later they took another picture to see if they were still located in the same place and hadn't moved apart from each other. Sure enough, one year later they were there in the same location. The observing team was actually still on the mountain on Palomar when Sri Kulkarni came down to Telus about the discovery. He was so excited about it. Shortly thereafter, Hubble Space Telescope took this image on the right to show that it could do an even much better job of imaging that brown dwarf. This is a diffraction spike from the optics on the secondary. It's not anything real. But what really clinches this direct detection of a brown dwarf star is not just that this has to be a rather low mass object, perhaps 20 to 50 Jupiter masses. But what really clinched it was the fact that they could actually collect enough photons from this to do a spectrum. And that's what the spectrum looked like. Now they did use a good telescope for that. They went to use the 200-inch hail telescope on Palomar to collect enough photons to get a spectrum. And here's what it looks like. The spectrum on the top of the planet compared to that of Jupiter. Jupiter has a surface temperature on the order of 165 degrees Kelvin that's cool enough to allow methane molecules to exist. And these methane molecules produce what are called absorption bands in the spectrum, these little dips that you see here. Gliese 229B has apparently the same methane absorption bands. That means that it has to be cool. And if it's cool, it can't really be a star. The coolest stars have temperatures on the order of a few thousand degrees Kelvin. In order for methane to be stable, this temperature has to be less than a thousand degrees Kelvin. So this is a true proof that the first cool brown dwarf star had been found. An object intermediate in mass between a Jupiter and a star, but also an object that can masquerade as a planet. It has the same spectral signature as a planet. But we wouldn't really call it a planet because we think it forms as a star and has much more mass than a typical planet. So it's important for us, and you'll see this as a thread through this talk to try and differentiate between what is a brown dwarf star and what is a giant planet. Okay, to summarize what had been found by the end of 1996, this is the plot showing all the discoveries in terms of two parameters, the mass of the planet plotted in a peculiar log scale such that the mass of the sun is here. This value here corresponds to the minimum mass for a brown dwarf star. We're getting smaller and smaller masses as we go to the left. Here is Jupiter and Saturn, Uranus and Neptune. There's the Earth. These two objects are two several Earth mass objects that were found around a pulsar in 1992 by Alexander of Wolfson, which is another important discovery I don't really have time to go into today. But pulsars are hostile environments. If you're looking for habitable planets, they're not the place you want to look. So I'll focus today on planets around solar-type stars. These are two of the new planets that were detected. Here are some putative brown dwarfs. These are normal stars like the sun. And you'll notice that the difference is between what I'm calling planets and these objects up here is first of all mass, perhaps. These objects are lower in mass by perhaps a factor of two or three. And they also occur on orbits which have close to zero eccentricity, meaning nearly circular orbits. Orbital eccentricity has turned out to be so far and so far still is a very important discriminant for trying to understand what is a planet and what is a brown dwarf star. Brown dwarf stars, like other stars, tend to be found on highly eccentric elliptical orbits. Planets tend to be found on more nearly circular orbits. So that's what observations have told us. Now let's go on and see how theory does on the report card. How well did theory anticipate these discoveries? So I'm going to first talk a little bit about the formation of stars. What's shown here is the Barnard 86 dark cloud in this region here in the direction of Sagittarius towards the galactic center, a very, very rich star field. This is cataloged by the same Edward Barnard that found Barnard's star by chance. What's important about this photograph is not the brilliant field of stars you see, but this dark region here, which is a clue that tells you that there's a dark, literally it's called dark cloud, full of gas and dust. The dust is what obscures the background starlight. And so within these dark clouds that we know that young stars are forming even today throughout the galaxy. Particularly at infrared and millimeter wavelengths, we can image these young stars and learn a lot about their structure. But what I want to focus on today are theoretical models of the process through which dark clouds convert themselves into stars because it turns out the theoretical models have been able to make a fair amount of progress in following the entire process. The observations by and large can give us snapshots and often snapshots with a rather coarse resolution. With theory, we can at least perhaps delude ourselves into believing we can do much better than that. So I'm going to show you first a simulation of what might happen when one of these dense clouds becomes dense enough that it can collapse. This is a cloud that's actually mild. You can't see it very well, but it's actually roughly spherical at this point. It's being supported by magnetic fields. It's rotating. It has a sort of a lumpy shape. Very much like we observe dense cloud cores to be like. If we watch it evolve for another, say, 100,000 years, it slowly begins contracting in the center. It contracts because the magnetic fields inevitably leak out of the cloud. As the fields leak out, the cloud is allowed to contract inwards. It contracts faster and faster. It eventually gets to a phase where it's no longer contracting, but it's actually collapsing. Supersonic motions develop, and in this particular cloud it becomes more and more bar-like. And eventually, about another 100,000 years later, we get a binary star system. Now that's reassuring, because as I hinted before, binary stars are very common in our galaxy. The Sun is actually a bit of an outcast and being a single star. And this process of fragmentation is believed to not only account for the formation of binary stars, but also allows you to make stars with smaller and smaller masses. Because these little clumps here have only perhaps 10 or 20% of the total mass that was in the entire cloud. They may gather some more of that mass as they evolve, but it's the process of collapsing and fragmenting into smaller clumps gives you a way to build smaller and smaller mass stars, and it gives you a way to build brown dwarf stars, for example. But there's a limit as to how small a mass you can go. And to understand that limit, just briefly look at this diagram. It shows you the temperature dependence of a cloud as it gets denser and denser as it contracts towards the region where it's actually has become a fully formed star. Because of the detailed thermodynamics of dust grains in these clouds, there is an initial phase where they increase in density by several orders of magnitude while still remaining rather cool at 10 degrees Kelvin. Because they're cool, they don't increase in temperature, thermal pressure can't stop fragmentation from occurring. The clouds are very prone to collapsing and fragmenting. And this is the phase where we believe binary stars form. But eventually the cloud becomes dense enough, and this is dense for an astronomer, but this is probably better than most vacuums and physics labs on Earth. At this phase, the cloud is dense enough that the dust grains begin trapping the heat that's produced as the cloud is compressed, and so therefore the cloud begins heating up. Heating up means the thermal pressure becomes stronger, and thermal pressure is what allows the cloud to stop fragmenting, it stifles fragmentation. And just to graphically illustrate that, I'll show you now four results of four different simulations that are started with smaller and smaller mass clouds that show this effect of fragmentation being stifling, being stifled. Here's a cloud with a mass of about two solar masses. It's collapsed down. These are density contours and a false color representation. The cloud is broken up into a binary, sort of like the one you saw earlier on. If you lower the mass by about a factor of 10, it starts heating up earlier. The thermal pressure stops inhibiting, starts inhibiting fragmentation. You start wondering whether or not that's really going to make a binary or not. If you lower the mass a bit more, all you get is pretty much a single central object with some spiral arms. Lower the mass a bit more, and you get pretty much what looks like a single star, which is fine for making the sun, but it's not going to be an object that's going to fragment and give you any smaller mass stars. And if you do the calculations, it turns out that the critical mass below which you can't fragment anymore is about three to ten Jupiter masses. So that's a theoretical prediction of the minimal mass of a brown dwarf star. Another important theoretical prediction is that brown dwarf stars and binary stars should form on highly eccentric orbits. And that's because we believe that binary stars form during this dynamic collapse and fragmentation process. It's a process where nothing is in equilibrium. The cloud has lost support. It's falling in on itself. Perhaps a bar-like cloud begins to fragment into clumps. The clumps are not in centrifugal equilibrium. They're not in a nice, finely balanced circular orbit. They're falling towards each other. And as shown in this diagram here, they actually begin life moving in towards each other and then falling back, coming back out again on what you would recognize as a highly eccentric elliptical orbit. So the second theoretical prediction is that stars should form on elliptical orbits. All right, now let's move on to planet formation. Planet formation is even harder to address observationally because we've been able to detect the disks in which planets form, shown here in this artist's rendition of the disk that perhaps formed our solar system. But we, as yet, have had no detections of protoplanets forming within the disks. And it'll probably be some time before we have sufficient resolution to actually find such planets. So once again, we're in the theorists' playground. We'll have to listen to what the theorists tell us about how planets form. That's summarized here for the terrestrial planet region. You have a single star that's formed at the center, a rotating disk of matter is moving around it. It flattens out because of centrifugal force. And planet formation begins with the growth of dust grains, tiny, interstellar dust grains with sizes no more than about a tenth of a micron. Micron is one millionth of a meter, so it's a very tiny dust grain. These dust grains are believed to coagulate, to run into each other and grow just by random collisions and to grow to sizes perhaps on the order of a kilometer within about 10,000 years. And that's due to two processes. Not only do they run into each other, but they also begin to settle down towards the center of the disk. You can see an enhancement in the layer here. They settle down towards the disk mid-plane. They become more dense. There are more of the dust grains around, and because there are more of them around, they have a higher probability of hitting each other and growing. So the initial phases are thought to occur rather quickly. In about 10,000 years, you'll produce kilometer-sized objects, and those are key because kilometer-sized objects are finally large enough that their gravity starts aiding you in the next phase of growth. If you have gravity pulling two bodies towards each other, if you have two bodies that would normally miss each other in a close encounter, because of gravity, they'll be pulled in towards each other and collide. If they collide, they have a chance of sticking together and building an even larger body. So there's a natural effect where when gravity comes into play, that the more massive bodies win out because the more massive bodies have more gravity, they deflect more objects towards them. They grow faster than the smaller bodies. There's a process called runaway accretion. It's basically the rich-get-richer type effect. And pretty soon you end up with perhaps a few thousand lunar-mass bodies, those bodies with about a hundredth the mass of the Earth, orbiting around the Sun. And that takes about 100,000 years. Shortly thereafter, we know from astronomical observation that these dust disks disappear around the stars. We don't know exactly how that happens, but we know from looking at young stars that by the time they're a million or 10 million years old, they're pretty much cleared of gas and dust. They look like a normal, more or less a normal star, but they're still contracting towards the main sequence, as astronomers call it. So the gas and dust is gone, and then the final phase of evolution, though, is a rather long one. It takes about 100 million years, and that involves the collisions between lunar-mass objects, which now have to move considerable distances over the solar system, and they run into each other again when rather fantastic collisions and produce the terrestrial planets. The key thing to remember is that in this case, because a terrestrial planet is made from the collision of many moons that come in from different directions. One body hits this way, one body hits this way. On average, the nudges it gets cancel out to zero, and that means that the planet ends up orbiting more or less on a circular orbit. So the prediction is for terrestrial plants, they should be on circular orbits. Just to give you some feeling for how planetary accumulation occurs, this is a painting of the impact that is thought to have produced the moon. The process of the final phase of planetary accumulation is a rather random process. You don't know exactly when collisions are going to occur or how they will occur, and because of that, you cannot make direct deterministic predictions of what the planet formation process will produce. There's a lot of chaotic dynamics involved, and what you have to do to study the process is run a lot of simulations and try and hope to statistically speaking understand how the process of planet formation works out. I'll show you now a couple of results from a colleague of mine, George Rutherall, who's run more simulations of the formation of the terrestrial planets than anyone else. This particular model shows you the results of 23 different simulations. So you have to mentally take what you see here and divide it by 23 to understand what you would get out of a single simulation. What he's plotting here are the final masses of the objects in units such that six on the scale is the mass of the Earth. So these objects here are earth mass. These are two earth mass objects. And here is the distance from the Sun. The Earth lies at one of these units. Venus is in here, Mercury, Mars is out here, and Jupiter is out around five in these units. What Rutherall has found is that with a rather typical guess about what sort of initial distribution of matter you have to make the planets in the first place, that you can... More chances... Your chances are actually quite good of ending up with a solar system where you have basically two terrestrial-like planets, one earth mass planet, one Venus-type planet, one smaller guy in close that you might call a Mercury, and one object out around here that you might call a Mars. So the outcome that we have for our solar system is, while not required, it is a likely outcome. We think we can understand how terrestrial planets form. You'll notice there are no planets forming out here in the asteroid belt because Jupiter is included in these calculations. And Jupiter's pull disturbs the orbits of these objects that are trying to grow and prevents them from forming a major planet. Rutherall has also studied the case of what happens if Jupiter isn't there. This is what you get. You get more massive planets. Remember, again, six is the mass of the Earth, so here you're getting two or three Earth mass planets at Earth's location. And you're getting them right on out through the asteroid belt. So we don't know exactly what we're going to find when we're able to start looking for terrestrial mass planets. But if we look for them in systems where there's no Jupiter, it's quite likely that we'll see even more massive planets and see them distributed over a wider range of radii than what are currently found in our solar system. Now, I just mentioned Jupiter. Let's get back to Jupiter's. This is a thumbnail sketch of how we believe giant planets form. You can start off with a disk of gas which has to produce kilometer-sized bodies which then grow into, in this case, roughly 10 Earth mass bodies. 10 Earth masses is believed to be a critical mass to reach because at that point, the body is massive enough that it can start pulling in gas from the nebula itself. We believe the giant planets are formed of roughly 10 Earth mass cores surrounded by, in the case of Jupiter, about 300 Earth masses of gas that had to come from the nebula. So the key thing is a timing issue. You've got to grow rather massive objects before that disk gas leaves. We know that disk gas is gone by the time the system is a million or 10 million years old, and the perennial problem in understanding giant planet formations to make them fast enough so that you can accrete the gas and end up with a Jupiter-mass planet by the time the gas is gone. This is not as well understood as the processes of terrestrial planet formation is, but it's also thought to produce giant planets on roughly circular orbits. Okay. One other constraint that can be applied to giant planet formation is where in the nebula might occur. This refers now to that spectacular misprediction I made. I referred to earlier. I have done some calculations where I try to calculate the temperatures within this. This is one such disk shown in the cross-section. This is a theoretical model that's calculated with a hydrodynamics code similar to the one I showed you earlier with the fragmentation calculations. This is a protostar sitting in the center. You can't see it, but there would be a one solar mass star sitting here, and there's a disk around here within which material can begin to accumulate to form planets. And with these numerical codes, I can calculate how hot the disk should be, and these are the typical results that have come out. One typically finds that the central portion of the disk, these are measured in degrees Kelvin, are rather hot. Perhaps 1500 degrees Kelvin is a typical value, and that's what you want for the inner portion of the disk, because that means you cannot have any ice condensing there. The dotted line shows you the ice condensation temperature of about 160 degrees Kelvin. The inner regions around where Earth and Jupiter would form are too hot for ice to condense. That means you cannot make an ice-rich planet in the inner region of such a disk. You can only make a terrestrial planet. That's fine. That's what we find in our solar system. On the other hand, the regions where ice begins to condense, where it's cold enough to have ice, is out perhaps around Jupiter's distance from the Sun, which is about five in these units. And ice is thought to be critical because you have to build those 10 Earth mass cores quickly, as I said in the previous slide. And you want to have as much mass around to help you build that, and ice is a critical component to do that, because ice is about two to three times more abundant than rocky material is. So if you don't have ice to help you reach that critical mass of 10 Earth mass cores, you're probably not going to be able to make a giant planet fast enough. So it's believed that you have to have giant planet formation only outside the ice condensation radius. Well, therein is the problem, because if you remember where 51 Peg was found, it wasn't out here at Jupiter's distance. It's right over here where my models would predict there's no way that a giant planet could form there. So what's wrong? What's missing? Well, I still think these models are correct, but what I think 51 Pegus's planet has told us the lesson it's taught us is that planets move after they're formed. This sounds perhaps like a theorist waving his hands, but it's not quite that bad. There actually had been theoretical predictions that planets should move well before 51 Peg was discovered. In particular, Peter Goldreich of Caltech and Scott Tremaine of Toronto Now published a paper in 1980 where they had done an analysis based on the effect of spiral density waves in Saturn's rings. These are waves where there's a moon inside the Saturn's rings and the gravity effect of the moon on the Saturn's rings produces wave-like spiral structures and these interact with the moons and it can make them move back and forth. They developed the analysis to handle Saturn's rings and then they just as sort of a side effect put in the numbers for what would happen if you made Jupiter inside this disk of gas in which we believe planets form and they found that Jupiter should fall into the sun within 10,000 years. It was viewed as something of an embarrassment at the time because of course 10,000 years is merely a blink and after having gone to all the trouble of trying to make giant planets, one didn't want to just toss them into the sun. They're after. So theorists took note of this but didn't really tell the observers that planets might move after they form. But now that 51 Pegas have been discovered, we have to let the secret out. There's another way of making planets move. They can not only move by spiral density waves as Goldreich and Tremaine had shown. There's also a model here by Douglas and John Papalaitzu which shows that when a planet becomes large enough it actually clears a hole in the disk. This is a plot showing how dense the disk is as the function of time coming towards you and the disk is actually growing in radius. It's getting larger and larger and at the same time becoming thinner and less massive. If you put a planet in just the right place it's massive enough. It sort of pushes gas away from it and opens up what's called a gap and that means that material can't pass past the planet's location and if the disk is expanding therefore the planet has to go along with whatever the disk does. Depending on where you put the planet initially the planet can either sort of stay put or it can start migrating inward towards its central star or it can start migrating outwards. So there are several ways of moving planets around after they form and the best guess now is that that's probably what's required to explain the existence of 51 Peg as well as several other so called hot Jupiters that have been found since then. There may be yet another way of making giant planets though and that is through not through this two step procedure where you have to build up a 10 earth mass core first then accumulate gas but it's conceivable that you can make them by watching the disk of gas coagulate or through an instability break up into clumps directly in a single step. It was thought that this process which people have studied in the past would only lead to strongly sheared thin spiral structures like this, sort of what you might think of seeing in a spiral galaxy I've done some calculations that suggest that it may not be implausible to believe that those clumps could actually go on and become more well defined and actually form a giant planet. This process is still very much under discussion and review if it does make planets it will again only make them rather far from the star so it's not a way of explaining the formation of hot Jupiters. Okay, now for some summary. This is where we were in September of 1995 this has showed you everything that had been discovered in terms of a search space that plots the masses of objects the planet versus their separation from the star so we certainly knew Jupiter and Saturn existed there was a possible brown dwarf star called Palomar Pleiades 15 it is still believed to be a brown dwarf star but it's so close to this hydrogen burning limit that no one can be exactly sure if it's truly a brown dwarf star or just a very low mass star. There was one other possible brown dwarf called HD114762 which is such a tongue twister that many people who worked on it began just calling it George for short. This seems to have a minimum mass of around the right limit to be a brown dwarf star but this is thought to be perhaps a case of a pathological case where the stars being looked at nearly pole on and so one really had a case perhaps where the object was much more massive than the lower limit it was really the true mass might actually be up here close to the hydrogen burning limit even though this was discovered in 1988 no one really took it too terribly seriously because there are arguments that imply that the system really was being viewed pole on. So that's where we were in September of 1995 just the month before Michel Mayor and Didier Kalo made their announcement. Here's where we are today that's impressive how much has been filled out. So let me walk you through the diagram of it. There are several other brown dwarf stars that have been found but they're not quite as convincing as at least the 229B object is. There are a lot of these brown dwarfs that have been found by Michel Mayor. There are a number of hot Jupiters and this is the very first one 51 Peg B. It's been joined by several more with similar masses and slightly larger radii. Let me explain some of the aligns on this plot. This oblique line here which says e next to zero this demarcates the region where if the object is too close to the star then we can't really tell what its eccentricity was initially because its eccentricity would have been lost through orbital interactions over the length of the star. That is when a planet is very close to a star and is moving on an eccentric orbit it tightly flexes the star. The tidal forces are changing. The star is pulsing back and forth and that action dissipates energy and the energy comes out of the orbit of the planet and it produces a more circular orbit and so these systems here are so close together that regardless if they started out on highly eccentric orbits they would have been driven to zero eccentricity by tides over their lifetime so we have no idea what their initial eccentricity was. Systems to the right of that line though presumably their eccentricity now is what they started off with. You'll notice there's not much of a gap anymore between what we think brown dwarfs are and giant planets down here in fact depending on there may be no gap whatsoever maybe there's a continuum between these two classes of objects but there does appear to be one discriminator that I hinted at before and that is the orbital eccentricity. You'll notice that all the objects below about three Jupiter masses have the solid symbols implying that they have a nearly circular orbit with one exception 16-signity B whereas the brown dwarf stars above here are all on eccentric orbits so that is what theory had predicted should be the case and that still seems to be the case except for a 16-signity B when Jeff Marcy found that the object and found out that it was eccentric he was overjoyed because he thought he really had a way of puncturing theorists ego proving that they were truly wrong in this case however it turns out that there is a way for the theorists to wriggle out of this one once again 16-signity is actually not a single star it's a binary star actually several of these stars are binaries Taubu, Upsilon, Andromeda, 51, Rowan, Cancri and because 16-signity though has a fairly wide separation from its from its parent star it's far enough away from its parent star that the binary star over hundreds of millions of years of evolution can actually affect the orbit of the planet around its primary star and kick up its eccentricity so there's a very good theoretical understanding of why this particular planet should have an eccentric orbit now it's because it's in a binary star system so so far this picture is holding together but then again the next detection of extrasolar planet could kick the whole thing in the teeth and we may have to go back to the drawing board and redo our thinking of how planets and brown dwarfs form let me now give you that promised glimpse into the future obviously finding planets is now a growth business NASA bought into the Keck Observatory in Mauna Kea helped pay for the second Keck telescope NASA now owns one sixth of the observatory the Kecks are being used at this very moment to look for extrasolar planets by the radial velocity method there also are plans afoot to develop an interferometry capability for the Mauna Kea Observatory by adding on these outrigger telescopes these would be smaller telescopes would be linked together through these underground tunnels and their light would be combined constructively in such a way that you have effectively the resolving power of a telescope with this diameter remember the Keck telescopes are the largest telescopes on the earth at the moment they're 10 meters in diameter but if you can combine them with the outriggers into an interferometer you have a telescope 100 meters in diameter 10 times more potent and with that sort of accuracy you'll be able to measure the positions of stars in the sky to about 10 10 micro arc seconds about a factor of 100 better than what people needed to look for things like Barnard's star and that with that accuracy we'll be able to find not only Jupiters around all the nearby stars but also objects as low as mass as low in mass as Uranus and Neptune the next step in NASA's program is to go above the Earth's atmosphere because that does limit what you can do from the surface of the Earth and fly the space in an interferometry mission. This will again be an interferometer working an optical to infrared wavelengths this will be perhaps 10 meters in size this will work with an astrometric precision again of about 10 micro arc seconds possibly as low as 10 times lower perhaps 1 micro arc second that will give that telescope the capability to find not Uranus and Neptune's but possibly a few Earth's around nearby stars if they're there. But the real focus of it will probably be finding Uranus and Neptune mass objects and perhaps multiple Earth mass objects around nearby stars and then the ultimate goal is still pretty much a concept at this point it's called the terrestrial planet finder again an interferometer perhaps now 75 meters across about one and a half meter telescopes mounted on it this will be an interferometer but this one will actually have the ability to image Earth's it will detect extra solar Earth's directly it's believed with the current designs that if it stares at a star for about 24 hours it will collect enough photons to make a picture and return you an image showing you where Earth Venus Jupiter, Saturn and Uranus are you wouldn't really quite be able to perhaps pick up Mercury maybe not Mars but you could pick up an Earth mass planet orbiting around that star and then just 24 hours of when it's commissioned if you then find what appears to be an Earth mass planet orbiting in the habitable zone of its star that is in a region where water could be liquid the terrestrial planet finder can then stare at that particular object for about a month and it will be able to collect enough photons from that to give you a spectrum and the spectrum will look something like this this is actually a spectrum of the Earth taken by the Mars global surveyor which landed on Mars a few weeks ago it was looking back at Earth and it's sort of a core spectrum but it shows several very interesting features this is in the 10 micron region of the spectrum and the hope is that terrestrial planet finder will be able to look an extra solar Earth and tell us if it has carbon dioxide which is a hallmark of a terrestrial planet's atmosphere tell us if it has water which is a prerequisite for life and perhaps most intriguingly tell us if there's evidence for ozone on that planet ozone is a proxy for oxygen if we can find evidence for water and oxygen on a terrestrial like planet we will have a case made that we have at least a habitable planet if not an inhabited planet thank you once again you can place your questions on cards and send them to the table well we'll begin with questions from the panel any of the panelists have questions Alan there was an excitement recently about discovery of huge number of so called watery comets by Drew Frank would they fit into this general scenario or not oh goodness I'm not sure how they would fit in perhaps David Stevenson would like to react to that why are you passing it today I really don't have honestly don't have a good reaction to that of course the watery comets have have multiple problems that people have passed out that if their comets are coming in at that rate on the earth and we presumably are hitting the other planets as well then we should have evidence for water on the moon and other objects and so I think while there's apparently an observation there that needs to be explained it's unclear if the best explanation is watery comets coming in perhaps others have more to add to that story and the way that planets can move clarify that a little bit for me the spiral density is gravity not homogeneous that's right you have to have a what's called a non-axis symmetric a field which is not completely symmetric around the disk you have the planet pulls on the material which is closest to it which is a common characteristic of gravity it pulls the nearby disk into a bit of a lump the lump pulls back on the planet and because of that the planet can evolve with respect to the disk and what you the outcome depends on how big the lump is on either side of the planet and the typical outcome is that the lump on the outer side is larger and that means it's pulling on the planet in such a way that the planet loses angular momentum meaning that it falls inward so for most density distributions in the disk you would have predicted that the planet should end up spiraling inward and being lost on potentially a very short time period Dr. Stone have a question as I remember you talk about George Weatherill's calculation about the formation of terrestrial planets and as I recall he did this for a range of masses of the central star and came up with a fraction of times that one would expect to find an earth-sized planet do you remember what that fraction was well I think what Weatherill found was he started off with a calculation where all he changed was the mass of the central star and he varied the mass of the central star by factors of two down to half the mass of the sun and then up to about twice the mass of the sun he found that that change didn't really have that big of an effect on the outcome of the planet formation process you still ended up with roughly earth mass planets at more or less the same distance so the mass of the star isn't really as important as you might think for the planet formation process but where it comes into importance is in whether or not that planet is located within the habitable zone because as you go to a lower mass star and it's of course cooler it doesn't give off as much energy that means your planet would have to be much closer to the star in order to be warm enough for water to be liquid so what he found was that actually and vice versa if you have a more massive star then the star tends to be so hot that a planet at one earth mass would not be able to have liquid water so he invoked the concept of what's been called the Goldilocks zone or the fact that the mass of the sun is just the right mass that will allow water to be liquid and therefore for us to be here having this conference thank you Dr. Russell my question is whether in the calculation the theoretical models which calculate the formation of planets do they also suggest the formation of moons and satellites around some of these planets or is that a detail not taken into account it's pretty much a detail not taken into account although people have advanced ideas about it but in terms of the resolution I mean literally it's a question of resolution in a theoretical simulation these tend to be computer models we're not quite at the point where we can worry about the formation of moons we're still having a rather hard time forming the planets themselves in the first place but given the general processes as we understand them these objects didn't also form moons around them we have a question from the audience is our asteroid belt a stable system or is it likely to someday form into a planet well I believe the asteroid belt is unstable but it's unstable in the wrong sense and that the asteroid belt is actually grinding itself down right now it's going in the wrong direction the objects are getting smaller they're being reduced down to rubble so I don't think there's any chance that the asteroid belt would begin somehow reversing the arrow of time and beginning to form a larger planet a follow-up question is the existence of the asteroid belt due to Jupiter or the outer planets or is it a natural occurrence it's believed that Jupiter is actually the culprit that explains the asteroid belt for the reasons that I showed in the calculations from George Wetherl that if you have Jupiter there that's the force that stirs things up in the asteroid belt region and prevents a massive planet from forming there if you take Jupiter away you form a planet there if you move Jupiter in closer the asteroid belt becomes wider and extends to where the earth is so Jupiter seems to be the key thing and the key thing there is having Jupiter form early enough that it can disrupt the accumulation process in the asteroid belt region okay another question from the audience do you think that there is intelligent life somewhere in space a profound question yes I I think that somewhere within the universe there is intelligent life yes but that's not saying much because of course the universe is so immense that it would be almost it would be foolhardy to say anything other than that the question is whether or not there is perhaps intelligent life nearby and I think that is probably remote but perhaps somewhere within the galaxy there might be something that we need to communicate with but those things are very much in the distance and they are very much an object of scientific research at this point would any other panelists care to comment one of the questions that I had when looking at your spectra and your final or second to last slide would it be possible to tell if there was life on a planet if a terrestrial like planet would have found another star based on the spectrum that you would observe well that depends that's very much an area of active research it's believed that if you could get a spectral measurement not only of ozone and water but if you could also measure methane that would be very close to proving that you have to have life around in order to keep those molecules which would otherwise be very difficult times but on the other hand I've heard a few exobiology type people say that you know it's conceivable that they can think up a way of having a lifeless planet somehow still produce methane so there may be a way of wiggling out of it but if we could find methane plus oxygen that would be reasonably it would be arguable case that you actually found evidence for life I think it's having simultaneously carbon dioxide fully oxidized carbon methane fully reduced carbon you put that in context of oxygen and water it becomes very difficult to do that without some special way of producing methane or carbon dioxide in our case it's producing methane I have a question for an earlier question that came from the audience had to do with the asteroid belt one of the theories I remember hearing has to do with the meteorites and the asteroids possibly being the residue of a planet that broke up early in the formation of the solar system is there evidence for or against that not something that you would probably call a planet in modern terms there were certainly larger objects in the asteroid belt that are being broken down there's any coherent argument for saying that there was anything of course on the order of even the massive Mars I think the next largest asteroid series is still quite a bit farther down in size from Mars so there may have been series size objects that were broken apart if you want to call them a planet then that would do it a question from the audience if a collision caused the formation of the moon then how do you explain the lack of a flat side of the earth it's a good question in fact there are some people would hypothesize that that's where the pacific basin came from before plate tectonics explained that one of course when a Mars size object strikes the proto-earth it's a tremendously dissipative collision it gives off an awful lot of energy and the residual earth that would be left behind would be the outer layers at least would be completely molten and it would not be a place you would want to own real estate and so the post-impact earth would do some rather serious rebounding and bouncing around in its molten state and would eventually settle down into a roughly spherical object subject to whatever rotation it has Stevenson one of the experts Dr. Stevenson would you like to respond I will talk about this a little bit tomorrow actually I agree with what Alan boss just said the main point of course is that the earth has lost memory of such an event from the point of view of its shape or temperature this is an extremely traumatic event as a matter of fact the energy that is released when a Mars sized object hits the earth is so great that if you were an astronomer around a nearby star with an infrared telescope you would see this collision by the glow created because of the 2000 Kelvin material that is radiating energy to space following the impact event the earth would be a molten sphere it would have a silicate vapor atmosphere be fluidized and then cool off over a time scale of a hundred or a thousand years following the impact event the question from the audience was Pluto formed in the same way as the other planets well certainly it was formed as part of the same process it's somewhat different because it's out on the edge of the solar system and didn't accumulate anywhere as near as much mass as the outer planets and Pluto of course has been even suggested recently that it's really not even deserving of the name planet because it's actually one of a whole population of objects that's now part of the what's now called the Kuiper Belt which are massive massive for comets presumably commentary composition objects in the outer regions of the solar system that form there Pluto is presumably the most massive object of that population and along with its satellite Charon and so we believe that it formed through the same process of building up of icy planetesimals that smacked into each other and stuck and there's no need to do any other physics only thing you have to invoke is having enough mass out at that distance to have enough collisions occur over four and a half billion years to produce Pluto today okay thank you okay at the end of as we come to the end of this afternoon's lectures I would like to remind the audience that of the this evening's events from 6 30 to 9 p.m. we have the art show running in the Schaefer Art Gallery unveiling the imagination views of inner space at 7 30 p.m. the Nobel concert in Christchapel at 9 p.m. you have a chance to ask more questions of the panel in two places this evening an alumni hall at 9 p.m. story Musgrave Bob Russell rolled Zogdev and Edward Stone will be available to answer questions for approximately an hour and in Olin 103 Allen boss Sherwood Rowland and David Stevenson will be available to answer your questions also the Olin Observatory will be open from 9 until approximately 11 p.m. tonight if weather permitting thank you