Direct Imaging of Exoplanets - Bruce Macintosh (SETI Talks)

good good afternoon and welcome again to the SETI Institute this is again our weekly colloquium series which we bring to you every Tuesday at noon and we're delighted to see new friends and old friends and familiar faces at our weekly talk series and we're continuing today on the theme of exoplanets where we had a nice talk last week from one of our own researchers and as I said last week I confess to having a bit of favoritism toward the exoplanet work and the work of Kepler but also GPI because I find the the results in terms of the number of planetary systems that we're discovering in the number of exoplanets we're discovering to be rather profound in the extreme so we continue the theme today and I'm delighted to be introducing to you dr. Bruce McIntosh from Stanford University he will be speaking to us his presentation is called a direct imaging of extrasolar planets and the Gemini planet imager of which he and his colleagues are pioneers in new instrumentation that is allowing this this imaging capability I will tell you that with my good friend and one of our trustees Pierre we we saw Bruce speak on this topic back in May at SLAC and it was astounding because I think that at that talk one of the first public displays of direct imaged exoplanets was was shown to a wide audience and it was pretty pretty mind-boggling very very impressive to see so a little bit of background on Bruce Bruce is a professor of physics at the heavily Institute for particle astrophysics and cosmology at Stanford he obtained his PhD at UCLA working on infrared instrumentation and searches for low mass companions to stars as a postdoctoral researcher and staff scientist at the Lawrence Livermore Laboratory he worked on using adaptive optics to sharpen images made by ground-based telescopes at the lick and Keck observatories he co led the team that made the first ever images of an extrasolar planetary system discovering four giant planets orbiting the young star HR 8799 he's a principal investigator for the Gemini planet imager and for those of you saw who saw the announcement a few weeks ago the press release from Stanford and also we put this on our own website he was the principal investigator and lead author on the paper about 51 arrey B which is a jupiter-like gas giant also directly image it using the G PI system and it's orbiting a star about a hundred light-years from here so pretty pretty amazing and interesting stuff and without further ado I'd like to ask Bruce to come up and tell us about his work and the Gemini planetary imager thank you okay thank you for the introduction and also the opportunity to come here and speak at SETI instead I'll be talking about direct image extrasolar planets in the Gemini planet imager GPI which is an instrument deployed on a telescope in Chile not this telescope actually the next telescope over but the view looking in this direction is better with the Condor in the foreground and this represents the work of a very large collaboration of people and I'm obviously could fill up the whole hour reading the names and highlighting the roles of the individuals which I won't do but in particular James Graham from UC Berkley I think actually did a an earlier talk on this except at the time he wasn't allowed to talk about our newest results so so I get to cover more and also SETI participants especially Frank Marquis who I've been working with in one form or another for more than 15 years originally on planets in our solar system until we got kind of bored of those and moved on to the extrasolar planets who's a P I here at SETI Eric Nielsen who's a postdoctoral researcher who works for Frank and myself sort of splitting his time between SETI and Stanford and Sarah blonde who was a summer student here who did some really excellent modeling of the system that I'll be discovering so you've heard a lot of exoplanet talks you don't need to be told why it's interesting to do it's one of the most interesting questions in my admittedly biased opinion in not just astrophysics but science and it's a question that's been people have been thinking about for a thousand years or more are there many worlds or is there about one world it's a question we now know part of the answer to we know there are a lot of other planets out there we know depending on how you count one or two thousand extrasolar planets orbiting other systems so our level of knowledge has expanded enormously in the past twenty years but our level of ignorance as is often the chansons case in science has also expanded enormously and that of these thousands of planets we really have no clear understanding of the process that formed them of the nature and the composition of weather solar systems like our own are somewhat rare incredibly rare or uniquely rare and many many other questions and a large number of researchers with a large number of techniques are working together to answer that question missions like Kepler have been absolutely crucial in it but the direct imaging that we do is also important now and going to be even more important going into the future and that's what I'll talk about again with history we'll just go back a long way this is before I went to graduate school the the view of the universe which basically consisted of the earth as being the center of the universe and a bunch of stuff that wasn't the earth going around and around it and you can kind of see why that's a natural model for the universe if you're standing and you're patient enough to stand all night you watch Stars rise you watch star set you get the feeling that it's all moving relative to you plus it's a natural assumption if you're human beings to assume that you are the most important thing there is the problem with that of course is that that model did a terrible job of predicting the way planets would move this is an animation put together by someone at OSU in a Ptolemaic view where the earth is at the center of the universe of the ways the planets have to actually move to reproduce what we see so the Sun the earth is sitting there in the center all the other planets are going around it as I start the animation and you can see our Sun Goes Around that's nice and easy but to reproduce what we actually see Mars for example has to make these giant great swoops because if you watch the position of Mars night by night sometimes it moves forward sometimes it moves backwards very hard to understand if all it's doing is going around the earth in a relatively simple way so if the earth was the center of the universe you have to have this incredibly complicated pattern of things referred to as epicycles on the other hand if you put the Sun where it's supposed to be in the center of the universe like the animation now shows it's a fairly natural behavior that's sitting on the earth watching the other planets go sometimes you're overtaking Mars and so Mars will appear to be moving backwards relative to you other times you are moving on the opposite side of Mars and it'll appear to be going forward relative to you so if you put the earth the Sun in the center as we all know the universe gets pretty straight forward and the first to really do that definitively Copernicus and Kepler produced quantitative models that use this idea of a heliocentric universe to explain the motions of the planets so Kepler also together with Tycho took positions of the planets night by night by night Kepler took his model in which they moved in elliptical orbits around the Sun and could use those to predict a week a month a year more in advance where the planets like Mars would be and for a scientist that was really the conclusive proof of the fact that this heliocentric model was correct so that that's sort of the theoreticians approach to understanding the universe the direct imagers approach is more like this Galileo the first person to really take in astronomical telescopes are really bad sign when scientists compared themselves to Galileo by the way so please forgive me for this but he actually took a telescope he looked out it and he saw one of his first Knights Jupiter and four little dots around it and then he looked at it the next night and the next night and the next night and he saw that these four little dots were moving that night by night Jupiter was surrounded by things that were at different positions and he very rapidly realized that means that things were orbiting around Jupiter that not everything had to orbit the earth but there were independent individual objects orbiting around something else in this case the moons of Jupiter going around Jupiter and it's interesting in some ways that that visual proof actually seeing other objects moons in this case orbiting other things a planet just like seeing other planets orbiting other stars was sufficiently visually compelling that Kepler who actually did the math to prove the heliocentric model correct didn't actually get into trouble for it the guy who actually looked through the telescope and said look little dots really they're not going around the earth that was something that the church could understand and got upset about and locked him away for so there is something specifically exciting about seeing objects whether they're moons or planets that isn't necessarily quite as exciting as plotting the positions of objects on graphs night by night of course back when they were doing this there were four planets in the solar excuse me six planets in the solar system briefly there were nine now we're down to about eight depending on how you count and with a clear distinction in our solar system between rocky planets nice small things you can walk on close to the Sun and giant planets big puffy gas balls far away from the Sun with the dwarf guys kind of in the middle to mess it up and of course a clear relationship that Kepler demonstrated between how long they take to orbit and where they're located in the solar system and so when I went to graduate school mostly nobody talked about planets because planets weren't very exciting for astronomers when I was in graduate school to the extent that they did we were taught a model of forming solar systems that did a really good job of making our solar system because that's embarassingly a pre Copernican way what we thought the rest of the universe would look like that model starts with a star surrounded by a disk of gas and dust as in this animation and that disk of gas and dust gradually forms the nice flat set of planets that we see in a more cartoony fashion the large disk of mcleod of gas that's going to form the Sun collapses down as it collapses down it starts to spin faster just from conservation of angular momentum and hence flattens out ultimately you have a Sun I'll try to be fair about pointing in both directions you have a Sun surrounded by this flat disc because the little proto Sun is hot the inner parts of the disc are hot the outer parts of the disc are cold and in turn that makes a difference to the composition of the disc close to the protosun the only way for something to be solid for something to be a rock or dust or a planet for that matter is if it's made out of something that's solid at high temperatures like rock or iron far away from the Sun where it's cold ice water can be solid hydrogen and oxygen carbon and oxygen very very common materials and so as a result early in this formation process it was thought there'd be a lot more solid material in close excuse me a lot more solid material farther out a lot less solid material in close in turn that makes it easy to make big planets far out where you've got lots of stuff to build up a core that the center of a planet rapidly enough that it can then absorb the gaseous material from the disk around it close in to the Sun where solid materials are rare it takes a long time to form something aside to the size of the earth by the time the earth formed all the hydrogen gas was gone so the earth didn't get a big hydrogen envelope so this model said you form planets gradually you build their centers first out of solid material and then you suck in the gas from the disk and in turn you can only really do that efficiently closed away from the star and of course this model like all scientific models had to survive being tested by observations it just took a long time to do that the first observations to see planets as you all know but if you've been to a bunch of planet detection talks were done through Doppler shifts where it's noticing that as a planet orbits around a star the star wobbles and it produces a tiny Doppler shift a tiny change in its spectrum towards your way from the earth which can be measured by precision telescopes that resulted in the discovery of a large number of massive planets like Jupiter the other dominant way of finding planets again you've been to lots of these talks you're familiar with or through transits if through an astonishing coincidence a planet lines up so that's exactly between the star and the earth as it passes across the disk of the star its shadow will cause the star to get just a little tiny bit dimmer as observed from the earth shadow that the star the planet cast onto the earth that tiny bit dimmer is about 1% for a planet the size of Jupiter which is actually pretty easy to mod measure with modern digital cameras and in fact you can see this effect even with a tiny telescope there are people who do this with networks of telescopes that are literally this big someone actually detected a planet using this method from a telescope in his backyard and pleasant and I believe and published a paper on it collaboration with real astronomers and with modern digital cameras on your home telescope you can successfully do this of course the odds of it happening for any given star astonishingly small because it has to line up just right so if you want to find a lot of these planets you need to look at a lot of stars if you want to see a small planet you need to look very precisely because the signal gets smaller the smaller the planet is and by far the dominant thing as everybody in this room knows for doing this is NASA's Kepler mission it looks at a hundred thousand stars by staring at a huge area of the sky and can see fluctuations in the star planet brightness at the level of a few parts per million small enough to see planets like the earth and I've told I'm allowed to show some graphs in this talk so I'm going to show a graph given that we have people with a sorted technical backgrounds this is a common way for astronomers to think about the populations of extrasolar planets it's a plot of planet mass versus semi-major axis where every dot represents a planet that's been discovered by different techniques and in this diagram you can see the planets that are discovered by the Doppler shift technique which is good at fairly big planets down to a few fairly small planets and very good at planets close to their star you can see the planets that are detected through the transit techniques especially Kepler with some color coding for planets that have been seen through both techniques and a spattering of planets or other techniques including my direct image planets which I'll get back to in a little bit I've highlighted roughly where our solar system goes and one of the striking things about this of course is it looks nothing like our solar system our solar system had the segregation rocky guys here giant ones here instead huge numbers of giant planets have been found close to their star and large numbers of planets have been found whose size is just in between the biggest the smallest of the giant planets and the biggest of the rocky planets in our solar system and here's another version of that just a graph showing from Kepler planet candidates as of January 2015 versus size the number of planets a few Jupiter planets Jupiter planets or bigger than Jupiter planets are fairly common but one thing Kepler is shown is this huge population of planets whose size is somewhere between Neptune's and the Earth's from Kepler that turns out to be maybe the most common kind of planet in the universe there's nothing like it in our solar system in addition there's just very no easy way to make these giant planets form close to the Sun because we think there shouldn't be enough material for them to do it there's no obvious way to make quite so many planets as Kepler sees down here and so planet formation as a theory is really there aren't any planet formation theorists here today are there good is really trying Jack lacera beat me up if I said this but it let's say it's still trying to recover from the observations and there's no single model that can predict all the planets that we successfully see but broadly speaking you can divide the models that still exist into two categories the first category are still based on the core accretion picture I showed before where you form the center of the planet first I'll show some slides of that and the outer parts later may be followed by migrating it as the planet forms in the middle of this disk it'll experience a thing like friction with the disk it's a gravitational friction but it's like friction that slows the planet down and causes it to spiral into the star and so that's a good way to get giant planets close to the star it's actually surprisingly natural that this should happen and surprisingly hard to stop it happening computer simulations that make it so one of the questions is why didn't it happen in our solar system why don't all the planets get eaten by their stars etc that's a slow we're slow means millions of years kind of process when it forms planets it forms them from the out from the inside out so it forms a core of a planet and then sucks up the outer material the composition of the planet I'll show in a couple of slides will vary a lot it also uses up a lot of the energy of the planet and that'll be important when I get to direct imaging planets and so it's sometimes referred to as a cold process because the planet ends up relatively cold there's a lot of variations on this it's very trendy these days to have kind of swirling clouds of pebbles that form the planets to to concentrate the mass and make them accrete more rapidly and finally we are pretty sure this is a good process for forming rocky planets a variation on this is still the best explanation for our solar system we know that formed rocky planets because we're standing on one and in simulations it can do a good job of forming rocky planets and other solar systems just to show a cartoon of that and allude to how it affects the composition here's a cartoon of a star and a disc of material around it and in numbers I've written down the ratio of carbon atoms to oxygen atoms in that disk and that will be an important number that we'll get back to later on in random clouds of gas in the galaxy and our Sun that ratio is about 0.55 five or so if this model where we're solid material freezes out is true that ratio can actually change a little bit there's a place maybe 10 astronomical units away from the star for this particular star that I've done my cartoon about where water forms into solids and so out in the disk from there on there little lump grains of solid water ice far enough out carbon dioxide can also freeze and so there's lumps of frozen carbon dioxide and in turn that will change the ratio of materials in the gas if water ice forms that sucks up a lot of the oxygen and so there's not as much oxygen left and so the ratio of carbon to oxygen in the gas of the disk goes up where all the car where all the oxygen got sucked into solids farther out the carbon gets sucked into solids too but even more oxygen gets sucked into solids and so the ratio goes up close in where none of these ices are the ratio stays kind of the same it was when the when the disc formed in the first place in turn if I build a planet out of those materials by say taking those little lumps of ice making the core of the planet and then sucking up a lot of gas around it and maybe dumping some more lumps of ice in it I can get a envelope for this planet whose carbon to oxygen ratio honestly can be almost anything but can be different than the disk was in the first place so because you form a planet a piece at a time the exact order in which you mix in the solids and the gases can cause the planet to have a chemistry a ratio of materials like carbon and oxygen and hydrogen that's very different than the gas that you started with deform it and we see that very clearly in Jupiter Jupiter is clearly enhanced in heavy materials because a lot of what went into Jupiter were these ices that caused it to form in the first place the other way you could potentially form a lot of exotic planetary systems is a fast process referred to as global instability that I just like to call it collapse where the big disk of material as it orbits starts to form little instabilities little spiral structures in this simulation those in turn under the influence of their own gravity can compress even further concentrate down into little lumps there's some lumps forming that would ultimately become planets this process if it happens happens really fast and in fact it's a problem with computer simulations that it happens way too fast to be plausible right now but it would probably happen still fast not millions of years but tens or hundreds of thousands of years and when it happens you just basically scoop up a whole big lump of the disk and squish it down to be a planet and so the composition of the planet is the same as the composition you started with for the disk so you end up with planets that look like the original cloud of gas that they formed out of in the galaxies no extra carbon no extra oxygen or anything like that because it happened so fast the planets have don't have a chance to lose any of their energy and so they form really hot and again that'll be important as well get back to it later on and because it really just scoops up a bunch of disk material it's a really terrible way to make small rocky planets it's most astronomers would say this is not the predominant method of planet formation which is good news because if it was then rocky planets would be relatively rare so as people went on to study these planets we got very excited by the first results we have beautiful press release figures this was sort of one of the earliest Doppler Planet detections you can even see the artist has added little moon getting evaporated by it you think it's a planet just like Jupiter but of course when you see a planet through a technique like Doppler shifts you don't know that all you know is that there's a planet that's going around the Sun you know what its mass is absolutely nothing else Kepler measures in a different way here's another artist conception example a really nice Kepler press releases recently about planets that are analogous to the earth orbiting in the so-called habitable zone or Goldilocks zone of their stars this is I think the one from a year ago when Kepler found a planet just like the earth as opposed to the one from the summer where Kepler found a planet most recently just like the earth or the one two years ago or the one three years ago there's a series of Kepler results along these lines it's very exciting to think about the possibility they could be planets like Earth but again well the artist conception shows little continents and maybe tiny seas or oceans on it all we really know is the radius of the planet and nothing else the planets Kepler finding could be rocky ones we could walk around on super-earths they could be micro Neptune's with giant hydrogen gas envelopes and except in a few cases we don't have a good constraint on that and when we put these artist conceptions together they're really artistic license more than anything else so what we want to do in direct imaging to get around that is to study the planet to actually take to actually see it of course taking a picture doesn't mean that we'll map out these continents and oceans that's not something you can do with any telescope we can build but if you can see a planet and see the light from it you can at least tell what it's made out of you can feed that light into a spectrograph and measure its composition and you can see different kinds of planets the problem of course is it's incredibly hard the analogy that I'm required by astronomer law to use is that it's like there was a lighthouse and you're looking right next to it to see a firefly something that's a million or a billion times fainter actually saw somebody do this calculation once and it's not completely crazy depending on that the population of fireflies and the brightness of the lighthouse you use but it's been done the first time it was done is actually not for a planet but for something called a brown dwarf brown dwarfs are objects that are really we think of them more as failed stars they're a cloud of gas that was going to form a star but never got big enough never got hot enough to have nuclear fusion happening to heat it up and so they sort of sit there cooling down over time but more than not quite more than a little bit less than 20 years ago the first brown dwarf was discovered sitting there glowing sitting right next to a bright nearby star and ever since then because you can actually see brown dwarfs directly we've studied them in enormous levels of detail we wanted to excuse me move beyond that to study planets the phase one of this we did using the Keck telescopes so the Keck telescopes are two ten meter telescopes in Hawaii you're probably familiar with they're built in adjacent buildings but 99% of the time you only use one Keck telescope at a time mostly this one for me beautiful telescopes their primary mirrors the shape of a hexagon super polished segments all attached together they're on this mountain in Hawaii because that puts them way up above most of the turbulence in the Earth's atmosphere but not all of it and so even from the Keck telescope one problem looking for a planet is that when you look at a star turbulence in the Earth's atmosphere blows the star up into a big blob and you can't see the planet light coming from a star gets distorted by blobs of light in the Earth's atmosphere in all different directions we fix that with a technology called adaptive optics we have mirrors that can change their shape that can be bent by little electric actuators to take the rays of light from the star and bring them all back into a sharp focus in a non cartoon version the adaptive optics instruments kind of look like this the big telescope here gathers all the light brings it down to a focus the deformable mirror the mirror that can change its shape does that about a thousand times a second to take all the rays of light distorted by the atmosphere and bring them back together it does that in response to commands from something called a wavefront sensor that measures the turbulence in the earth atmosphere a thousand times a second computer converts that into instructions for the deformable mirror and successfully makes the image better we take a fraction of the light from the star and split it with a thing called a dichroic to that wavefront sensor the rest of the light goes down to where we can actually do science with it this kind of technology has been in routine operation astronomically on astronomical telescopes for about 15 years and it can take something like this which is an image of Neptune taken with a keck telescope under good atmospheric turbulence condition you can kind of tell that there's something there it's mostly round all it looks surprisingly not round for a planet and that there's some stuff going on on it you turn the adaptive optics system on and it looks like that so it can sharpen up the images by factors of ten or twenty to sort of Hubble's telescope quality images on our giant telescopes on the ground with that we can fix the light coming from the star but we still can't easily see planets Jupiter in our solar system is about a billion ten to the nine times fainter than this than the Sun and that's too faint to overcome even with adaptive optics even with the Keck telescope the second trick we use to find planets in addition to the adaptive optics is we look for the right kind of planets and the right kind means young planets so this process these processes I've just described for forming planets in either case initially big cloud of gas at the end small planet big cloud of gas has a lot of gravitational potential energy because it's spread out over a large area by the time you take all that mass and it ends up being a planet through whatever pathway it took that energy gets released in the form of heat in the planet itself so when planets form they're hot Jupiter now is sitting around at minus 200 degrees but when Jupiter was brand-new it might have had a temperature of 3,000 degrees or more glowing fairly brightly by the time is 10 million years old 1,500 degrees 100 million years old 500 degrees now minus 200 and that glow is detectable especially if you choose to look at infrared light which is the light that things give off when they have sort of warm temperatures planets become much much brighter so another graph this is a plot of the brightness of the planet expressed as a luminosity relative to the Sun is what this means so the fraction of the energy of the Sun that a given planet is giving off from something as bright as the Sun to sort a thousand times dimmer than the Sun all the way down to a billion times dimmer than the Sun versus age in billions of years so 1 billion years sitting here and each line corresponds to a different sized planet from 15 times the mass of Jupiter down to sort of Saturn size and each planet starts out bright and then just gradually gets dimmer and dimmer and dimmer as they cool down and if you catch them at 10 million year old ages the planets are bright enough so there are only 10,000 times or a million times fainter than their star which is still pretty faint but a lot better than a billion that Jupiter is sitting out way down here right now so by looking for planets that are young which we can do by looking at stars that are young the planets are potentially detectable a complication in this is that this is another one of those sort of pre Copernican graphs where people put it together assuming a simple process of planet formation a more sophisticated approach to this accounts for the two different models of planet formation that I talked about it in particular if a planet forms through the core accretion process form the core first suck all the gas in like we think our solar system formed that releases a lot of the planets energy early on and later on the planet is dim so one last version of this graph this here color codes show each different size of planets medium sized jupiter-sized planet really big ten jupiter-mass planet if they form through the fast collapse process then they're really really really bright when they're young and gradually cool down if they form on the other hand through the core accretion process like we think our solar system did they actually never very bright early on they kind of sit there percolating away and then gradually rejoin the cooling tracks of the other planets so a totally is a sort of mixed blessing on the one side it means we're we're sensitive with our early equipment to sort of ten to the minus five ten ten million times greater than the Sun we can only see planets if they start out hot but someday if we could see both kinds of planets maybe we could tell which way they form by looking at how bright they are versus time so a bunch of groups searched using first-generation adaptive optic system for planets like this this is one example of a success probably the most spectacular example using that the HR 8799 system that christian Marwa and a number of people including myself found in 2008 in 2010 so these are actual pictures of planets what you're seeing here is the star kind of behind a mask a lot of image processing has gone into it but a little bit of leftover starlight or these white blobs here and these black blobs here but the dots are actually infrared radiation from four planets the arrows are not giant arrows that were drawn on the sky they're there to guide your eye in the process the star is about 50% more massive than our Sun the planets are about 5 to 7 times the mass of Jupiter so they're big planets about a thousand degree temperature for the outermost ones maybe 900 degrees Kelvin for the innermost ones but they're real so we can actually now look at a whole other solar system from outside and see its individual planets and see them moving around in some ways this whole other solar system is kind of like our solar system we have four giant planets it has four giant planets our solar system has a belt of asteroids and a belt of comets the Kuiper belt there's evidence from infrared observations this solar system also has a belt of asteroids and a belt of comets at large angles and that there's a gap between those right where the planets are which makes sense because the planets would clear out asteroids and comets they're just like they do in our solar system other ways it's not like our solar system the total mass of these planets is probably 30 times the total mass of planets in our solar system their orbits are bigger but of course it's a bigger star so it kind of makes sense that that's going on and we can watch them orbit this is just an animation Cristiano oh I put together of data going back to their discovery in 2008 to some 2013 or so epochs the innermost planet is hard to see you can only see it when the weather is really good so it's not in all these blobs but there's the other three they're moving they're moving like Kepler's laws say they should move so this one is going really slowly because it's far from the star this one is going quite fast because it's close to their star it's not a surprise Kepler's laws of course or newton's laws they apply everywhere in the universe but the fact that we can see Kepler's laws from outside that we can see a whole solar system going around the way once upon a time we would have just had animations like the one I showed you earlier we can see in real data real life is actually pretty amazing so I said why we might want to directly image these planets is to think about their composition what they're made out of we had our press release figure 2 and I'll show one 4:51 arrow down a left but of course we don't see that we just see the dot of the planet but because we can image it we can actually say something more about its composition so I'll just remind people little about spectroscopy then every element in the universe has a distinctive footprint of light that it gives off the atoms that compose it have certain levels of energy they like to absorb or emit and so and so it will only give off certain particular color of light and you've all seen this in your highschool physics experiments or or college physics experiments if I have a cloud of hot gas and I feed the light from it through a prism I'll get a particular distinctive chemical fingerprint and that chemical fingerprint differs for different elements helium sodium neon etc if you're a scientist rather than looking at these color bars you might choose to make a plot a graph of the intensity of light versus wavelength if you have a cot light source behind a cold cloud of gas instead of seeing bright lines you see dark lines where this cloud of gas likes to absorb certain particular wavelengths and if I plot the intensity of light versus wavelength you can see the dips as the intensity goes down again due to the chemical fingerprints and so in turn if I can get light from my planets I can start to say something about the chemistry of their atmosphere we thought their chemistry would look a little bit like this so these are just stylized cartoons of the atmospheres of various objects from right to left the coldest stars have atmospheres that are mostly gas but have actually clouds of very materials that are solid at high temperatures like titanium oxide the brown dwarfs especially the hotter brown dwarfs those clouds live deeper inside the planet lighter-weight clouds are higher up and their atmosphere is full of carbon monoxide which you could see in a spectrum the cooler brown dwarfs and we thought planets would look like cooler brown dwarfs the clouds are even deeper in their interior up at the top where the atmosphere is cool you can get compounds like methane and so we thought we would see radiation coming up from the hot clouds through the methane and we'd see the spectral fingerprints the absorption lines of elements like methane when we looked at our planets so we decided to look for this here's a Keck image without a lot of image processing now showing what one of the planets looks like relative to the very bright star and this is the field of view of the spectrograph we used for it the spectrograph instrument called Osiris is a very special spectrograph it doesn't just measure the spectrum of one object at a time instead it takes a 3-dimensional picture where for every pixel in the image I actually get a little micro spectrum I get the intensity as a function of wavelengths so I can measure the spectrum of every object it looks at all at once over a small postage stamp field of view and I can show that by animating yes here's the the total image with a little dot there that turns out to be the planet all the streaks you're seeing in this image or scattered light diffraction and optical interference off the Keck telescope because they're a process made by diffraction each of these little streaks actually different wavelengths of light go to different positions in it these are like tiny little extended spectra where red light scatters far away from the star and blue light scatters close I can turn on an animation where I'm now stepping from long to short wavelengths and all of the artifacts all of the scattered light have positions that change as you go from short wavelength out to long wavelengths short wavelength out to long wavelength the planet always stays at the same position at all wavelengths and so you can use that to get rid of all the scattered light and say yeah there's the planet stock it up and so we extract that we do some math we turn it into a quantitative plot of intensity versus wavelength intensity wavelength and you can see in this this is near-infrared radiation because that's where the planets bright so long with visible light would be somewhere off here long wavelengths of light you can see some chemical fingerprints for example carbon monoxide absorbs very strongly here water this is absorption due to water vapor in the middle you don't we don't you make a big deal about water vapor because it's hydrogen oxygen is the most common thing in the universe of course there's water it'd be weird if there wasn't but but there's a huge water vapour absorption signature present there but there's not is methane is supposed to make a big absorption signature here and it didn't show up so we see the water we don't see any methane we see carbon monoxide and the overall radiation we see we think is from hot thick clouds of liquid metal so it's confusing that we were seeing these thick clouds and not seeing the methane what we think is happening is that the planet's atmosphere circulates pulling up the clouds to the top of the atmosphere where we can see them the clouds of liquid metal and pushing down the methane where it gets destroyed in the inside of the planet because we can study its light we can actually study it at very high levels of detail here's a zoom in on a different version of that spectrum here's the real data do to do to to to do and each individual little wiggle in this especially out here actually corresponds to a real feature from the spectrum especially of water and carbon monoxide but not of methane we can use those models that will through this in detail but to measure those things like the ratios of carbon to oxygen in it and we find that the planet is somewhat enhanced in heavy stuff carbon and oxygen compared to hydrogen in a way that says maybe it did form through the core accretion process like Jupiter did in our solar system that was interesting on the other hand the planet is way too bright to a form through that process back to this graph brightness of planet versus age the planets that form through the fast collapse process the dotted lines the planets that form the old-fashioned core way like our solar system the solid lines different colors meeting different masses of planets the HR 8799 planets kind of live here there maybe 2030 million years old there are only ten to the minus five times the brightness of the star and so it's really hard to make those through a core accretion process they're too bright to form the way we think Jupiter did but their chemistry says they form the way we think Jupiter did so there's actually a lot of confusion about how you could form these planets and in fact although there are a lot of directly image planets there's no clear consensus here's some other ones like beta Pictoris which is in the middle of a beautiful disk of dust a couple of others with serial numbers there's no clear consensus on how these directly image planets could have formed and no evidence yet that they form the way our solar system did showing this again number of planets versus semi-major axes all the Kepler discoveries all the Doppler discoveries if your direct imaging person you show these graph a lot actually makes you a little bit sad because of course there's a thousand kepler planets there's only a handful of directly image planets but some people's reaction is not to get sad but to figure out how to do better first one thing that's interesting to do is show only the planets that we actually do have spectra of it turns out for transiting planets like Kepler finds for a handful of cases you can also measure their spectra it's a separate talk that I'm sure you've had people do I won't go into it but every direct imaging planet you can also do it and so direct imaging is a large fraction of the planets that we've ever actually even gotten close to measuring their composition of and so that makes us feel better but really we'd like to find if not a thousand like Kepler which we're never going to do enough more planets to fill in this direct imaging region and it also to bridge this gap the farthest out we can see with Doppler techniques versus the closest we can see with direct imaging direct imaging Deezy planets when they're far from the star Doppler it's easy to see them when they're close to the star we haven't quite reached the point where those overlap but it would be enormous ly powerful if it did and finally we want to see more planets and maybe see the ones that formed in the cold way we've really got to see fainter planets down to say this kind of ratio of brightness and so that's the purpose of the Gemini planet imager it's an instrument whose construction we started in 2004 on the gemini south telescope in Chile it combines whole bunch of different pieces I'll go through in a second physically it's about the size of a person or a robot in this case this is the cube it mounts to on the back of the Gemini telescope it's in a big enclosure for example you don't want a speck of dust on any of your mirrors because that might look like a planet so the whole thing is in the equivalent of a class 100 clean room when it's under operational conditions computers that control the adaptive optics system leave in these enclosures off to the side another thing you don't want is the plume of waste heat coming from a fast computer going up through your telescope and making atmospheric turbulence so we put the computers in these boxes we circulate glycol liquid through the cabinets to cool them off and then the main body of GPI lives in this interior box it was built by a collaboration Lawrence Livermore where I used to work before I came to Stanford built the adaptive optics system and also ran the whole project a group at the American Museum of Natural History made some special masks to block the light from the star I'll show those in a couple of slides JPL built a very precise infrared sensor to measure small errors in the wave fronts of light coming into the instrument UCLA built our science instrument which is another spectrograph like Osiris that makes not just images not just spectra but 3d spectral cubes the group in Canada the National Research Council of Canada in Victoria built the mechanical structure that holds the whole thing together and equally importantly the software layers that connect all these different pieces that have to talk to each others about I would say about 15 person years of software inside the instrument we went to a lot of effort to super polish and keep all of our optics clean so they don't look like stars one key technology for GPI to make it make it better than the current systems is its deformable mirror so I showed the cartoon of a deformable mirror earlier what a deformable mirror for atmosphere turbulence really is is a thin sheet of glass about a millimeter so so thin they could have two very careful while you're holding it maybe this big or so with behind it a couple of hundred little actuators little piezo electric plungers that push and bend and warp the surface of it to respond to atmospheric turbulence 4gp we needed to do that not just a couple of hundred locations but at several thousand properly keep up and correct for all the little distortions in atmospheric turbulence if we built this with the same technology little electric plungers glued behind it the GPI deform layer would have to be kind of this big across and it wouldn't have fit in that enclosure on the back of the telescope and it would have cost on the order of let's say five million dollars or so instead our deformable mirror is actually a microchip it's a silicon device called a MEMS a micro electro mechanical systems device which is basically laid down lithographically the same way microchips are with 4000 of these actuators in something kind of the size of a quarter I mentioned masks to block the star we kind of smooth the telus transmission of the telescope so it doesn't have sharp edges to diffract off of but kind of looks like this doughnut mask and then we focus the star light down onto a mirror with a hole in it such this the star falls down the planet comes to a focus off to the side and we separate the two so we can measure the spectrum of the planet that way we built the whole thing at a bunch of different places then integrated it not far from here at UC Santa Cruz we installed the infrared spectrograph in 2012 the instrument has to operate in a fairly tough environment on the telescope in Chile or on the back of the telescope so as the telescope points in different directions the instrument might Bend and it has to stay aligned to within about 10 microns or so so that's not acceptable so we put it on a crane we cranked it all over the sky we saw how aligned it stayed when it moved which of course it does we measured those motions and we programmed the computers to compensate for them by steering mirrors in opposite directions similarly it has to work not just at a kind of room temperature but down to minus 3 or minus 5 Celsius which is the coldest is supposed to get at Chile so we built a giant meat locker around it and chilled the whole thing down to minus 5 Celsius to make sure all the computers worked and that it would stay aligned over a broad range of temperatures we did that successfully we packed it up and shipped it it's a sort of disconcerting experience to see 10 years of your life packed into crates my small child helped with the process packing and shipping put it on a track you can't drive to Chile turns out there's no roads all the way down there so the track took it to Los Angeles and then a land airlines 767 took it down to stopping briefly Lima and then continuing to Santiago and then another truck took it from Santiago to La Serena I did not watch this process because first of all I don't speak Spanish I would have been that helpful and secondly again ten years of your life on a track going up twisty mountain roads is so I was just sort of heavily drugged at home until it actually was delivered we could follow it and it made it there intact so it's unloaded by Gemini there's the telescope there's the Chilean Trek sat in their basement for a couple of weeks I did fly down to be reunited with it glad that it survived the trip bolted it on to the back of the telescope a nice thing about the Gemini telescope is here it is now on the back of the telescope like in the diagram we saw before looking up so light from the telescope comes down into it you can actually attach multiple cameras and spectrographs to the Gemini telescope these are some of their other cameras and spectrographs and so that's nice because what we were getting are is tuned up they could do science with their other instruments our first light almost two years ago now a large team of scientists and engineers from a bunch of institutions down there for it data coming out from GPI you will see a bunch of examples of this just like with the osiris instrument it's not an image it's a spectrum of every pixel so here's another animation from long to short wavelengths of infrared light this is just a Neptune in our solar system as you go through the spectrum at every pixel we get the intensity of light as a function of wavelength and in Neptune you can actually see some of this atmosphere structure for example these clouds here are way up near the top of Neptune's atmosphere and so they stay bright a cloud that deep inside Neptune's atmosphere gets much dimmer at the wavelengths of light where methane absorbs because Neptune's atmosphere is full of methane for early GPA science we concentrate on planets that were already known to exist this is the planet around a star called beta Pictoris beta pick B what you're seeing here in this GPM egde here's the mirror with a hole in it the star fell down that hole this is kind of a glint off the edge of the hole left over light and diffraction coming from the the star the four spots you're seeing here are not planets they're spots that we've induced in the image as a calibration and the reason we induce them is that when you find a planet one thing you want to know is how far is it from the star if the star disappeared because it fell down the hole that's hard to measure these four spots are actually lined up so that the intersection of them is always where the star is who can trace the star position very accurately the planet is that little dot here so that's a another big one about ten times the mass of Jupiter for comparison here's an image from a previous generation instrument that people who discovered that not the people discovered that planet another group after it was discovered planets visible planets more visible in the GPI image so we're somewhat happy but we're extremely happy because this took three thousand nine hundred fifty two seconds of telescope time and that took a 60 seconds of telescope time to make the same image so GPI is in fact substantially more sensitive than its predecessors and so we were happy and GPI has looked at a wide variety of objects I think James talked about some of them in his talk I don't have time to discuss it in detail including discs of asteroid or comet material orbiting stars Europa and Pallas images that Franck looked at young discs that might be in the process of forming planets a whole bunch of other planets people have known but what I'm here to finish up about is talking about our first new planet so here's another one of those spectral cubes I should get a version without the arrow to see if people can spot the planet because what happened literally was we started looking for new planets in November of last year so we're targeting a population of young stars about 600 of them to see which have nearby planets we had an observing trip in December right before Christmas all of us went home pretty much except for one grad student in one postdoc at Berkeley and the postdoc literally looking over the grad students shoulder at a computer in Berkeley saw a little dot that was kind of coming and going all these artifacts sort of hold still but this little dot is coming and going because the brightness of the planet is changing as a function of wavelength and here's its spectrum and it's changing exactly the way our model said cold planets should where light is being absorbed by methane and emitted by the hot clouds deep in the interior of the planet the star is about 20 million years old if it's really a planet it would be orbiting around where Saturn does and so the first question we asked ourselves is is it really a planet it turns out there's a large history of thinking you found a planet and being wrong in this game for a lot of reasons one of the most popular reasons is what we see here here's the images from G PI again and a Keck image we actually took an image with the Keck telescope so at least we know it's not an artifact from the telescope because it's there with Keck as well you see something very faint next to something very bright like a star one explanation is you're seeing something very faint because it really is very faint it's a planet right next to the star and planets are faint another explanation is it's very faint not because it's really faint but because it's very far away it's another star that through an astonishing coincidence happens to have lined up right next to your star and it's not that unseen coincidence if you do this for years and years and years and years and in fact I believe my ratio of actual discovered planets to background objects is about 5 to 400 or so over the course of the career so traditionally what you do is you come back a year later and you look for motion if it's really a planet it'll move with the star plus may be orbiting around the star if it's really a background object it'll either hold still or wander off in some separate direction we didn't want to do that partially because we've been doing this for a long time and we're not very patient but also partially because we didn't think we we needed to we did observe its position at multiple eras this is a graph Erik Nielsen here at SETI put together here's the year versus the position of the planet when I plot these I plot the separation between the planet and the star that's the top one and also the angle that the planet and the star make that's the bottom one we sought a planet in December of 2015 we confirmed it after Christmas in January of 2000 and excuse me December 2014 January 2015 but that's such a short period of time the great tract is the path a background object would have followed the cone is the range of orbits a real planet could have followed and you can see over the short period of time there's no way to tell the two of those apart we did dig through a lot of archives and we discovered half a dozen other groups had looked at this one individual star to try and find planets none of them had seen it of course they would have published it we took for example some old Keck data from 2003 we analyzed it very heavily to see if we could see the plan which would have been about here we couldn't see the planet it makes sense CAC is less sensitive than us but that does actually help us tell if it was a background object the star is moving kind of this way and so if it had been a background object the star would have left it behind it would have been farther away from assuming the stars moving towards the planet so it if it had been a background object back in 2003 the star would have been further away and would have been just barely detectable by Keck and in fact we can sort of exclude a range of allowed motions for it this is a plot again Erik put together of now the separation between the planet and the star we know way back in 2003 the planet had to be within this range not right on top of the star we know where it is now and so we can partially constrain the fact that it's not at least a stationary background object so that helps but it could be a background object that's still moving and in fact the referee on our paper said you should reabsorb this freedoms it's very snarky about yeah this is the gold standard and if people publish a paper that doesn't go up to the gold standard in science they may be right they may be wrong but it'll be a bad example for the kids and everybody else will be publishing planets that aren't really planets we did not find that compelling as an argument what we did find compelling is that because it's G PI we have a spectrum and stars don't look like planets here's the spectrum at short and longer infrared wavelengths and it's got this enormous signature of absorption due to methane stars don't have methane in their atmosphere what does have methane in its atmosphere are either real planets or things called brown dwarfs objects that are in between planets and stars and the thing about the brown dwarfs is they're incredibly rare the odds that you would get a brown to find a brown dwarf you have to serve a pretty much the whole sky and you'll find 100 of them the odds that you'd find one right next to your individual star nearly zero and in fact the most compelling not the most compelling quantitative argument we did the calculations and we say that there'd be about one or two per million stars we look at background objects that we would actually find so the odds of being a background object are relatively small an example of that is that a hole but as I said a whole bunch of other people looked at this star to see if it had a planet or not they looked at it with cameras with huge fields of view that had just as likely as we are to find background objects in fact more likely the only place we can that they can't is right next to the star background objects should be anywhere planets right on top of the star so the fact that we founded with a planet imager it even says that on the label means it must be a planet it's I won't go through as I'm running along one time in detail but we know the age of the star very precisely because it's actually part of a triple system here's another figure Eric put together the star and it's little planet over here and then there's another pair of stars that are orbiting around them in this complicated ballet this pair of stars they're relatively low mass stars and you can measure their age very precisely we did put together very cool animation Jennifer patience of faculty more Arizona State University together with their planetarium and Danielle and Franck working through here put together an animation showing the system like 50 like HR 8799 this system has a comet belt comet equivalent belt and an asteroid equivalent belt and the planet is kind of in between them which is probably where planets should be as we zoom in close you can see the planet sort of sitting there Jupiter like but glowing because what we see is the infrared glow from it and in this figure as it zooms around you can also see the other two stars in the system just richer they're actually no Franck my know if they got the orientation of the Milky Way right or if this is just arbitrary that doesn't really matter miss the double system alright I should have pointed it out and I said you should be distrustful of all these gratuitous artists conceptions which is sort of true but what you know we don't have an image that looks like that we do have this spectrum and from this spectrum you can take models that tell you things like what the surface is what the atmosphere is like excuse me and in order to reproduce the bumps and Wiggles in this spectrum successfully we have to have an atmosphere that is kind of half covered by clouds that are cold at the top of the atmosphere but half without clouds where you see deep into the hot interior of the planet and this artist conception that Danielle and Frank put together actually reflects that so I don't know in detail that there's five bands or seven bands I don't know that the bands are elongated or there's a giant red spot but I do know that if you were to look at this planet with your eyes you would see a surface that's sort of half glowing deep on the inside and half too and opaque clouds on the top of the atmosphere and then finally in terms of its formation this is sort of the first planet that could have formed either way here's another version of those formation graphs brightness of Planet versus time all the previously discovered planets are way up bright where hot start planets could have formed this one is sitting a little bit under a million times fainter than the Sun where potentially a planet that formed like Jupiter would be sitting at this age on this particular graph we can't tell for sure it could be a very small planet that formed hot and just us isn't bright because it didn't have a lot of energy to start with it could be a big planet the forum cold and is not very bright because it lost a lot of its energy forming as its core as it accreted mass on its core but someday we may be able to tell the difference if we could ever measure the mass of the planet and there's some hope that in five years we might be able to do that accurately we could tell whether it sits on this hot start or cold start family of curves so to summarize to about twenty million years old we've seen it since it should say December 2014 excuse me temperature about 650 Kelvin two times ten to the minus six the brightness of the Sun the odds of being a background object are less than ten to the minus five the mass might be twice the mass of Jupiter at a separation of Saturn not quite our solar system but getting there and young enough that it remembers how it formed and then just as a hot off the presses I was in Chile what day is it I was in Chile a week ago and we saw that the planet is still there which is kind of reassuring before we went there Eric and his summer student Sarah did some modeling of the range of motions it should have it's actually pretty close to where the range of motions where the predictions say its motion should be and over the next six months we'll be studying in a lot more detail and so I think I'll finish there rather than going into the space mission stuff but I'm happy to talk about future space miss and other people hi Bruce nice talk so my question is what are the chances that Jeep I would be able to image a planet by reflected light rather than self linguist like a Alpha Centauri or something like that not to put too fine a point on it um it's just right Rosco knows the answer to this but it's a good question to ask and G PI wasn't designed to do this it was designed to look for the self luminous planets the signal from a reflected light planet is very small but not completely out of the question and especially if you took advantage of polarization so one capability G Phi has I didn't talk about at all is it can measure the polarization of light as well as its spectrum light that's scattered by turbulence in the Earth's atmosphere is unpolarized or light that's scattered by the dust on the telescope is mostly unpolarized light that reflects off a planet might be polarized and if it is that would help us pick that signal out so if planets have significant polarization fractions like ten or twenty percent then a one or two au planet around maybe even four or five stars would be within the detectable range if you spent persuaded the Observatory to let you spend ten or twenty or thirty hours on it so I'll do Epsilon Eridani you can do alpha centauri be a good combination for um they're models that predict the reflected light polarization the Europeans are maybe looking for it maybe not the self luminous planets we talked about might also be polarized but for reasons too complicated to go into right now but but GPI has shown to be very very good at making those precise polarization measurements at the part in a thousand level so we so are speckle suppression factor for polarized light might be ten to the third which does get us down if we have enough photons into the regime where we could hope to see these things can you tell us more about debris in 51 Eridani is it imaged or is it just from the spectral distribution it's just from the spectral energy distribution so we can tell these debris disks most often because the dust from the asteroids and comets colliding with each other absorbs visible light and radiates infrared light so the signature is you see extra infrared light from a lot of stars and you infer that that's from from dust material or aliens but dust is probably a more robust experience in general though people talk about Dyson spheres and things like that the 51 area case it's dust disc is actually pretty thin for such a young star it's prone ly about ten times as much as the amount of dust in our solar system which for 20 or 30 million year old star is quite young so it's Kurt almost certainly undetectable especially because the planet notionally clears a gap in it somebody I'm sure will be looking for it with the Hubble Space Telescope which is very good at seeing dust at relatively wide orbits but there's no detection yet so I know with brown dwarfs you can see patchy clouds or some kind of cloud modulation as the planet rotates do you has the spectrum change when you went back to look at it again or do you think it might just got off the airplane a week ago so um so ask me in a little bit people certainly do think it might for some classes of planets that you'd see the same kind of modulation as the clouds come in and out of field of view it's a hard measurement you could see how big the error bars are on that spectrum but it's certainly on our list is to try and monitor it and the the trip I was on I was in Chile for five nights and we only really had one good night so I can't tell yet but but I think it's unambiguously worth pursuing especially if it is at sort of 45 degree or more inclination great talk uh did you did you mention that you've got future work queued up in slides that you could show us in a minute I had a little bit about about future space missions especially the W first after I've space I wouldn't mind seeing a light or two if we've got time people can leave - yeah I assume they unlock the doors at this point w first yeah w first let me just do two or three slides and I also had a little bit about the thirty meter telescope so the broad question beyond is G PI is really limited still do these Jupiter size or bigger than Jupiter sized planets Kepler says that's not the most common kind of the planet in the universe how are we going to characterize them a lot of missions will do it including that the test mission the sort of son of Kepler but ground-based imaging has a role to play too the first way to do it is to put something like G pie on a really big telescope like the proposed 30 meter telescope in Hawaii or its equivalents in Chile that would probably get down to kind of Neptune II planets and scrape the super earth regime another approach would be future space telescopes Russ is involved in a proposal for one that would reach exquisite levels of sensitivity around very bright stars so to probe alpha centauri a broader version for slightly bigger planets is something called w first so the acronym it's going to get a real name when we fly it but for now it's w first wide field Infrared Survey telescope so the first part of its name tells you what it was originally intended to do which is to survey the whole sky at infrared wavelengths for dark energy cosmology stuff that a number of my colleagues find extremely I'm being videotaped aren't I stuff that a number of my colleagues find extremely exciting and so it was proposed and ratified by the sort of astronomical community as the most important thing we could do in the next decade it was originally going to be a so the second part of named after physically focused telescope assets that's even worse than wide field IR Survey telescope because it's not a word where does that come from so w first was originally going to be about a 1.2 meter i started kepler sized telescope extremely wide field of view infrared camera and it would basically do counting galaxies and measuring their shapes and over a three to five year mission it could only do that then something weird happened so we explained this is let's say you like hiking a lot you might buy hiking boots and they might be on sale so you might buy a bunch of hiking boots then you might forget you have them the open your closet and you discover you've got eight pairs of hiking boots clearly too many you have some friends so you say would you like some of my hiking boots you donate them to them the same thing happened with the Department of Defense in NASA except instead of hiking boots you should think Hubble Space Telescope sized telescopes so the department defense opened their closet and they discovered I do not know how many they have but they definitely had too many Hubble Space Telescope sized telescope assemblies hanging on a hook somewhere and they donated two of these to NASA so two point four meter telescope assemblies not whole spacecraft just the telescope in the optics but designed for a wide field of view so great for the dark energy science big optically good robust and available and if you're going to do that I'll skip this animation although it's very cool there are a couple of good things 2.4 meters first of all it's big enough that it can do the extremely exciting galaxies counting stuff much faster so it doesn't take all five years of the five-year mission which gives you some time to do other stuff like say look at planets and it's big enough that you could actually use to do direct imaging of exoplanets which is hard to do on a 1 meter telescope but very practical on a 2.4 meter telescope so the baseline is to add something basically that looks like gee pi a deformable mirror a coronagraph mass to block starlight and integral field spectrograph and fly it on this and do characterization of extrasolar planets in reflected light and in the usual math semi-major axis diagram space from some simulations dmitriy Saran ski at Cornell did it could reach down to maybe even earth mass planets but definitely to the sort of super-earth planets several times the mass of the earth and expect to discover get photometry maybe spectra maybe not but definitely characterization of a decent sample of those if it flies which it has a good chance of doing in 2024 or so so that's the w first sidelined well if there's no other questions I would like to run a one one last quick one excellent question so if you were to read no atmospheric turbulence but there's still imperfections in the mirror of the telescope it's better than the Hubble mirror it doesn't have the big miss fira collaboration but just the act of polishing it puts tiny little ripples in the mirror from the tools you use to polish it to actually see a planet without any correction the size of those tiny little ripples would have to be smaller than an atom they would have to be a tiny fraction of an angstrom or a nanometer you can't make a mirror that precise and you can't count on it staying that precise as you fly it in space so the deform mirror cancels the aberrations intrinsic to the optics as well as more complicated diffraction russ is wincing because i'm oversimplifying this but as well as more complicated optical effects intrinsic to the design of the telescope so it's to fix the little polishing the errors and imperfections that's why you can't do this with Hubble but you couldn't do this with any space telescope it's much easier to build it properly much easier to correct it rather than to try and polish it at that sub ultra mega nanometer level okay one last question I was wondering with GPI or maybe future surveys like this is it at all possible or likely that we could get a long enough time interval of observation of a single planet that we could look at variations in sorbets and actually infer whether there might be lower mass less bright planets in that solar system caused by those causing those variations that's a good question because Kepler does things like that of course in timing of eclipses it's unlikely so are the fastest orbiting planet that the groups I've been involved in have been discovering is sort of a 30 or 40 year period and so right now I'm trying to exercise better and eat well so that I can see it do a whole orbit but to see these perturbations you really need multiple orbits and very precise levels of pastrami tree but who knows we may get get that precise going on into the future all right well with that I would like to thank our speaker and thank crews for wonderful talk and this is the part where our speakers treasure most this is where we hand out the revered SETI Talks speakers mug these are highly valued and prized I think on eBay now they're about 50,000 a piece so anyway very rare but thank you so much great great talk all right

Channel: SETI Institute

Views: 42,129

Rating: 4.7176471 out of 5

Keywords: Gemini Planet Imager, GPI, exoplanets, SETI Institute (Organization), Astronomy (Field Of Study), Exoplanet (Celestial Object Category), Young Jupiters

Id: ABOr_Suyxvo

Length: 65min 25sec (3925 seconds)

Published: Fri Sep 11 2015