 Please welcome our first speaker of the night. She is a fifth year graduate student at the University of Washington in the astronomy department. We're in the same research group, actually, and I can attest that she is a fantastic scientist. Please give it up for Arena Bootsky. We're going to talk to you about Cosmic Rain with Cosmic Rays. It's a project that I just spent the last five months working on in, or at the Center for Computational Astrophysics in New York. Now I'm back and I'm really excited to share it with you. But before I get into it, I'm just going to give you a brief outline of what I'll talk about today. So I'm going to start out slow and talk about what galaxies are and how they're entrenched in a certain galactic medium. I'll talk about what cool gas is and why it's so weird and so important for galaxy formation. I'll talk about cosmic rain and how we make simulations of that. And then finally I'll talk about cosmic rays and why they're important in the cosmic rain process and in reproducing observations that we see. So this is a galaxy. Galaxies are gravitationally bound objects of billions of stars and lots of gas and dust. Here this blue light is coming from massive young short-lived stars and this is a good sign that the galaxy is alive and well. But in order to keep making these young massive stars and stay alive, galaxies need a constant supply of cool gas. And if you look at all of the gas inside the galaxy and you add it all up, you'll quickly find that there's only enough galaxy to make maybe one more round of stars. So without a consistent supply of cool gas from somewhere else, galaxies will die out and look something like this. So this galaxy is red and dead, which means that all of the young short-lived stars that are blue and really bright have long gone and you're left with low mass red stars. What I want to understand is this cool gas and where it comes from and how it makes its way up to the galaxy. And so far I've been kind of misleading and showing you what a galaxy is because a galaxy actually looks something like this. Where all that pretty stuff with all the stars is right in the center and it's about a tenth of the actual size of what a galaxy is. And this galactic disk lives inside a much larger dark matter halo. And all the stuff from the edge of the disk out to the edge of the dark matter halo is called the circumgalactic medium. It's a very straightforward name, but it's also very long so I'll call it the CGF from now on. Although this is a rough sketch, this is actually to some approximation what a galaxy looks like if you were to look at it. Because in order to see something you need photons to come from that object and hit your eye or your telescope. No matter how large your telescope is you still need some minimal amount of photons to see it. But even though we now know there's more stuff out here in the CGM than there is in all of this, it's spread out really really thin. It's spread out so thin that the density in the CGM is one million billion billion times less than that of the air we breathe. So even with our best telescopes we cannot directly observe it. But luckily we cannot observe it indirectly and some very smart people have figured out that the CGM actually has this rich complex structure. And it's shaped by chemically enriched outflows from the galactic disk, relatively pristine inflows from outside of the galaxy. There's some ambient random stuff going on here and it's mixed in lots of complicated ways. And so being able to understand the structure and reproduce it in simulations is really important for making predictions for how galaxies will evolve. And so the good news is with these observations of the CGM we now know there's a lot of cool gas in the CGM. And it's enough to sustain star formation and that's great news. And also I should say when I say cool gas I really mean gas that's about 10,000 degrees Kelvin which is hotter than the surface of the sun. But for our intensive purposes this is considered cool relative to the other gas in the CGM. But the bad news is this cool gas is much lower density than we expect it to be. So we know the gas pressure we expect in the CGM. And from that pressure we can use the old ideal gas law you might remember from chemistry Pb equals Nrt. And so if we know the pressure and the temperature we can make predictions for the gas density. And so this is a plot showing gas density as a function of the distance from the galactic center. And these are predictions for the hot gas density and the cool gas density. And so what we expect to see is this cool gas should have densities right around here. But what we actually see is that cool gas density is way down here. It's about 10 times lower than what we expected to be. And that's a big deal. It means that this gas is not in pressure balance with the hot gas. And so that requires an explanation that we don't have yet. And so moving forward we have two things that we need to explain. One, even though it's great news that we see this cool gas, we need to understand how it forms and why there's so much of it in the CGM. Which will be the focus of this first part of the talk. And two, we want to understand why this gas is so low density. And as I'll talk about later, I think it's cosmic rays and the pressure support they provide. So one way we can form cool gas is through a process called cosmic rain. Which is something you might be very familiar with if you live in Seattle. And so if you look, if you consider some patch of CGM gas, here we have gravity pointing downwards in the page. And this gas has a very smooth profile where cool dense gas is at the bottom, hot diffused gas is at the top, and it's in a very stable configuration. So if you were to leave it alone, it would stay like this forever. But as I showed a couple of slides ago, the CGM is a very complex turbulent structure. And so no matter how smooth your patch of gas is to begin with, it's going to be disturbed by a wind or an outflow or an explosion, something like that. And one really important thing to making cosmic rain is having processes that let the gas cool and heat. But first to build up intuition, I'm going to talk about what happens if this gas is perturbed, but there's no cooling or heating. It's just gravity and buoyancy. So if the gas is perturbed, and there's just a fancy way of saying that these gas blocks are moved out of their equilibrium position. So now you have cold gas where it's surrounded by a hotter medium, or maybe hot gas that's surrounded by a colder medium. What's going to happen is that over time these cold gas clumps will sink, and the hot gas clumps will rise until they're back to where they belong in pressure equilibrium. But if you have processes that let the gas cool and heat, and you start out with some perturbation, what's going to happen is if the cooling can act faster than this buoyancy force, the cold gas cools faster than the hot gas, and so it becomes even more cold and more dense, that makes it cool even faster. And so you get this runaway process that ends up creating cosmic rain. And the really important balance here is the cooling processes, and having them act faster than gravity can correct for these perturbations. And so one way to study this is to run idealized simulations of this process. And so I set up idealized simulations. So this is a picture of the domain, which looks a lot like the idealized setup from the previous slide, only now it's copied over this midplane, and gravity is pointing towards the center. And it roughly corresponds to this piece of the CGM. And I'll say the galaxy isn't actually modeled because we're focusing on the CGM gas. And when I say I run simulations, I mean something like this. So on large enough scales, galaxies and the CGM can be approximated as a fluid, and that fluid gets divided into lots of grid cells, many more than are shown here. I just got tired of copy pasting. And they start out with some initial quantity. So you describe the density, the mass, the velocity, and maybe a few other properties of this gas at every cell. And then you provide the program with relevant equations of physics to tell how the cells should act over time. And then you let a supercomputer do the heavy lifting and check back in on it every once in a while, and you get science. And so this is a movie of four different simulations that start out with the same initial conditions. This is the color showing the density of the gas, and it's densest and coolest near the midplane, and slowly becomes less dense out towards the edges. And on the left are simulations where the cooling processes are really efficient, and on the right are simulations where cooling is relatively inefficient compared to gravitational forces. So if I run this movie, you'll see that where cooling is fast, you see rain drops forming and starting to fall back on the disk. I'm going to run that a couple more times. And on the right, where the cooling time is inefficient, you see the gas is moving. It's trying to find a new equilibrium, but it's not actually forming rain drops. And so these simulations so far were really exciting to see because they helped kind of confirm a lot of theories and other simulations that have done similar things. But so far everything I've talked about has been work that's been done before. And so what I did, well, we can check off, you know, we have a theory for how cosmic rain forms. We can reproduce it in simulations. But the thing I added was cosmic rays in simulations of this cosmic rain, and I'm going to show you how it influences cool gas. And so what cosmic rays are, are charged particles like electrons or protons that are accelerated to relativistic velocities in extreme shocks like supernovae. And electrons and protons are tiny, and we're dealing with really, really large scales, the scale of a galaxy. And so on those large scales, you might have guessed it, cosmic rays behave like a fluid. And they're a slightly different fluid than the normal galaxy fluid. And so the basic approach is just to model both of them at the same time. So where before the simulations just had this quote unquote normal galaxy fluid, I added a second relativistic fluid that is cospecial with the normal fluid and it exerts pressure on this galaxy fluid. And so I ran a bunch of these simulations, I'm here just a handful of them. So now all of these simulations start out with the same initial conditions except for that they vary in how much cosmic ray pressure is in the box. So on the left is the example of no cosmic ray pressure, on the right are simulations where cosmic ray pressure is 10 times more than the gas pressure. And so what you'll see is that this, the more cosmic ray pressure you have, the more different the simulation looks. And there are a few key takeaways. One, the cold gas pumps become bigger and they have less density contrast between the dense and the diffused phase. And in the simulation all the way on the right you see that the cold gas isn't actually moving down towards the midplane because it has so much support from this cosmic ray pressure. But if you look at a similar movie of the temperature, this looks a little, is similar but different. The shape of the perturbations is the same but the contrast is really there unlike in the density movie, play it one more time. Which means that the gas is actually cooling down to the same temperatures. And so we can quantify this a little bit by looking at the histograms of the temperature distributions. So here we have the same, now instead of looking at the simulations column by column we're looking at different rows. At the top are simulations with no cosmic rays and then they move down to the case with extreme cosmic ray pressure at the bottom. And this is looking at the distribution of temperatures. So this just means that about half of the gas mass is in the hot phase and about half of the gas mass is in the cold phase. And even as you include more and more cosmic rays this doesn't change very much. Yes there's some difference in this one right here but it's quite a small difference given how much cosmic ray pressure there is. But if you look at the density you have a very different picture. They start looking really different. And so in the case without cosmic rays up here you have two very distinct phases. You have a hot phase and a cold phase. One is low density and one is high density. And they're very separate densities. And the more and more cosmic rays you add the less difference there is between the density of the hot phase and the cold phase until you get to this really extreme case where the density is basically the same for both the hot phase and the cold phase. And so what this means is that the cosmic ray pressure is basically letting the gas cool without changing its density as would be dictated by the ideal gas law. And this is really exciting because this is starting to look like the weird observations we see where the cool gas is lower density than we expect and the cool gas density is pretty close to the hot gas density which looks like something like this. And so this isn't a done deal. The simulations I showed you have a very simplistic approximation to how cosmic rays move relative to the gas and so moving forward I've been working on different models of how cosmic rays move and seeing how that changes these predictions. But I think we're very close to actually constraining both the models of the cosmic ray transport and how they can match observations. And so in summary, galaxies need cool gas to live. The circumgalactic medium has tons of cool gas but it's at very low density and this low density might be explained by cosmic rays. And I didn't quite find a way to tie this movie in but I really wanted to show it. So this movie doesn't have cosmic rays but it does have cosmic rays and it is very high resolution. And I guess the tie-in is that or one thing I missed in the talk is that cool gas actually lives on very, very small scales that are basically impossible to recreate in large scale galaxy simulations. And so the really neat thing about idealized simulations is that you can really push the resolution on the smaller patch of gas to study this cosmic ray and so that's what it looks like if you map up the resolution. Awesome. Well, thank you. I'm going to bring back to you. So the question is what is plotted on the y-axis of the histograms? And the answer is these are the different simulations. So this is the simulation run without cosmic rays. This is a simulation with record pressures 0.01 times the gas pressure and so on. And so these correspond to the different panels and the movies I showed you. Oh, sorry. The actual, I see the y-value is the frequency or basically the number of cells that fall within each bin. And they're normalized so that they're the same or you're comparing apples to apples. How many cosmic rays there are? Great question. So Margaret asked how do people observationally constrain how many cosmic rays there are? And that's a really hard question for the CGM because we basically can't or at least haven't right now. And there are, we have some constraints based on gamma ray emissions from the galactic disk. And so the best we can do right now is we have some idea of how many cosmic rays should be produced and how they might lose energy or escape the disk. And so you use, you inject simulations of galaxies with cosmic rays, let them escape and then see which models were produced at galactic disk and then say, okay, these probably apply to the CGM, but it's pretty poorly constrained right now. Yeah. The cosmic rays in question, are they coming from just specific gravity or any coming from other branches of the galaxy? Great question. The question is, are the cosmic rays coming from, within the galaxy or from outside sources? And so most of these would be coming from supernova explosions within the galaxy. Although in the simulations I ran, they just started out as a uniform ratio of the gas pressure. Oh, in the back. Ooh, great question. How are we able to detect the presence of galactic gas or of gas outside the galactic disk? I have a slide for this. One of these will be the right thing. I hope. No. All right. How do you, hey Nicole, could I ask you to hit that skip slide right before the CGM picture? A little one? Between four and five. So we can observe CGM gas indirectly. So if you have a telescope like the Hubble Space Telescope looking at a bright distant object, that's billions of light years away, chances are you can design your observation so that it intersects a galaxy or intersects the CGM of a galaxy. And then as this piece of, or like this light ray moves through the galaxy, it forms a, I can't play, there's a movie, it forms absorption features which are like shadows cast by, oh, thanks Nicole, by the atoms in the CGM. And so from the details of the absorption features, these are the guys, we can figure out the temperature, density, composition and velocity of the gas in the CGM. Yeah, so it's mostly hydrogen than helium. So hydrogen is like 75% of the mass and helium is 24, I don't know. And about, I think, up to a few percent other chemicals. Yeah, all the way there. The question, so the question is are the pressures high enough to induce star formation if so, how do the time, or are the time scales long enough to create supernova feedback? So the cool gas I talked about here still needs more time to cool before it can form stars. It's significantly hotter than what you would need. There are stars that form in the halo, and I'm not sure how that relates to this. Okay, I have time for one more question. Oh, sorry, I do see a hand. Great question. So the question was, is the dark matter boundary, the boundary at which gas can no longer be detected? The short answer is no. It's kind of hard to define. It's usually defined by where the density of the dark matter is some multiple of the average density of dark matter in the universe. And you can detect gas out beyond that radius. It's just difficult to see because it's so diffused. So there are studies going on right now trying to see just how far gas makes out beyond that radius. Awesome. Thank you very much. Congratulations again to our trivia winners. I'd like to introduce our second speaker of the night, a National Science Foundation Fellow at the University of Washington of Postdoc and a rare person who lives in Washington and also originates from Washington. Give it up for Dr. Keaton Bell. I say good morning because I'm observing this week, so my schedule, I stay up all night commanding telescopes over the internet and then I sleep almost through the entire day. So I just woke up a few hours ago. So the telescopes I'm using... My clicker is not working. It's going to be really interesting. Okay, so these are the telescopes I'm using this week. There's two of them pictured here. I took these pictures in my trip down to New Mexico to the Apache Point Observatory just last month. There's a 0.5 meter telescope here on the left. This is in this raised dome. I've been using that every night for the last couple of weeks. And then this 3.5 meter telescope I'm using. Occasionally this is in this big boxy building here. And that building is really cool. Super, thanks. This building is awesome. It's got a giant telescope up on the upper floor and the entire upper floor of the building rotates. And so when you need to point to different parts of the sky, that entire top of the building swivels. It's kind of disorienting if you go in there and then it turns and you come out and you're facing a whole different direction, right? But I'm taking a break from my observations tonight. I'm going to share with you a very important message and that is that the Earth is doomed, okay? I'm not talking about climate change. I'm not talking about right now, right now we're doomed but the Earth is going to be fine without us. I'm talking about in a few billion years, right? The Earth is going to meet its end and the existential threat is our own sun. So to explain why that is, I'm just going to really quickly run you through sort of a review of stellar evolution. You heard a little bit about how this works in the last talk. Space is full of these big clouds of gas and dust. And every once in a while, parts of this gas and dust clouds, they contract. Maybe a supernova goes off and jostles on a little bit. Regions start to contract and form these star-forming regions. Smaller regions contract even more under their own gravity and as they do, they start to spin. When things spin really fast, you think about making some pizza. They tend to form discs and the same thing happens with stars. That material forms discs. Gravity causes material to flow down into the center of these discs and that's sort of where stars form. As that all contracts, the material heats up. This material is made primarily of hydrogen. When four hydrogen atoms love each other very, very much, they smash into each other and they form helium and this fusion process releases a bunch of energy. So once fusion starts within a star, all this energy is being generated, that energy is going to work itself out of the star and that process creates a bunch of pressure. Pressure holds the material up against further contraction from gravity. We reach this balance between the inward pole of gravity and the outward push, this pressure from the ongoing fusion process that is now happening inside of a proper star. The rest of this material that was in the disc will form planets and those planets will orbit a fairly stable star. For something like five billion years up until now, we have a stormy untap and here we are. That's the story up until now. But that's only half the story because the Sun needs hydrogen to continue to fuse in order to stay alive and it's halfway out of fuel at this point. In another five billion years, the first thing that happens when the Sun runs out of fuel is it's going to expand. It's going to expand so much it's going to consume Mercury and Venus and the best case scenario, it only reaches very close to the Earth, close enough to evaporate off all our oceans and turn the entire surface of the planet into molten lava. So nothing survives. This is why the Earth is doomed. I'm really interested in these final stages of what happens to planetary systems and what happens beyond this. You heard something about this in trivia already. And after this expansion, the outer layers that have puffed up to destroy the inner planets, they're going to get shed off in this beautiful display called a planetary nebula. I think some of the most gorgeous objects in space next to a few spiral galaxies and maybe Saturn are these planetary nebula. And then these inner objects, the rest of it compresses down under gravity, no longer held up by nuclear fusion pressure. They contract further until you just can't physically squeeze electrons together any closer and that halts this compression. You end up with about half the mass of the Sun squeezing an object about the size of the Earth. And we call this a white dwarf star. So that level of compression is like taking something like an elephant and compressing it into the size of a marble. This is the density of this material. White dwarf stars are called white dwarfs white and this compression really heats them up until they're white hot. And dwarf stars because they're so small. So that's the story of what will happen, a rough cartoon of what the future holds for us and our planet. I'm interested in what the actual situation is for specifically the future of our solar system and the types of planets that we know about. This is the general story that we have. There's a couple of facts for you. The vast majority of stars become white dwarfs. So I talked about this for the solar system. This is what's going to happen to most planetary systems and the vast majority of stars host exoplanets. That's something we found out in the recent decades. We've discovered thousands of planets around other stars and the future for them looks like this cartoon picture. So there's a couple of facts. You put them together and it sort of implies that most white dwarf stars should host exoplanets. The inner ones are maybe destroyed as I described but these outer planets are going to be fine as the stars lose some mass. Their gravity is a little less strong. Outer planets may drift out a little bit but they should otherwise be fine. Unlike normal stars where we know of thousands of planets now we've discovered no planets around white dwarf stars. We think they should be there. Logically they should be there but we've never observed any and I'm working to try to discover the first of this new type of planet planets that orbit white dwarfs. I'm doing this because if we find planets that orbit white dwarfs and we understand their properties, what that population of planets look like then we can try to connect them back to the population of planets that we know exist around normal stars that are going to become white dwarf stars and use the differences between these populations to understand what the complicated processes are, the complicated interactions between the final, late stages of stellar evolution and the planetary systems that orbit these stars. So that's my goal. The project I've just begun a few months ago at UW is to do exactly that. I'm going to run you through it. First let me dazzle you with some math. This is going to be fun for you. I want to take a look at how likely it could be whether it's really possible to find a planet around a white dwarf star. We have no examples of this and so I want to make sure this is a reasonable undertaking. So I can write an expression for like what's the number of planets that I expect to detect if I were to make an effort to search for planets around white dwarf stars. It could be expressed as a simple product of just four things. I'm going to tell you what those four things are and I hope it makes sense. The first is the number of white dwarfs that you look for planets around. If you look for planets around more stars you're more likely to find them. You've got to reduce that by the fraction of white dwarfs that actually have stars. Sorry, that actually have planets. So only some fraction of the stars that you look for planets around that's maybe going to reduce how many planets you could find if you carried out a big search. The next factor, it's a little tricky I'll explain it later in more detail. It's the fraction of time that those planets could be discovered if you try to look for them. So maybe sometimes the planets are not very detectable if they're only detectable for a small fraction of the time that you could look then it's quite possible you could miss them if you looked at the wrong times. And then you also benefit if you try to look for these planets many times over. So maybe planets are only visible you know 10% of the time if you only look once you'll probably miss it if you look 100 times you'll probably find it. This is the idea. So let's walk through each of these terms in turn and explore sort of what these numbers might look like try to see whether I have sort of a good idea if it's a good idea to invest my time in looking for this or if I'm totally destroying my career on some like wild goose chase. So the first term here is the number of white dwarfs observed and for this we really benefit from measurements of stellar parallax. Stellar parallax is this effect where if we observe some object from different positions then it appears to move on the sky in particular if we look at a bunch of stars as the Earth orbits the Sun we're observing these fields of stars from different locations on the sky from different angles and compared to some distant background like this field of distant white stars here this red star appears to change its location based on where we're observing in Earth's orbit and if the star we're studying is really close to us that amount of change is much larger and if it moves further away and now the star only wobbles a little bit on the sky observing how much a star wobbles on the sky as we look from different locations can tell us how far away a star is and this enables us to detect white dwarfs because white dwarfs are incredibly faint objects they're faint because they're so small remember the size of the Earth they're faint because they have very little surface area or light to escape from and so for us to see a faint white dwarf it has to be really close to us so things can be faint because they're far away or they can be faint because they're close to us and they're small like white dwarfs so we can detect the white dwarfs because they're the objects that look faint but when we measure their parallax they're clearly very close to us there was a recent mission called Gaia that did this that took observations like this and measured the parallaxes of billions of stars and created this incredible map for us by the way, parallax is the same process that we use for a depth perception we have two eyes and they're slightly separated from each other and so they each get a slightly different view of what's going on and depending on how far away things are the message to each of our eyes is different Gaia actually orbited a little bit further away than the Earth from the Sun so it got extra leverage on parallax this would be like someone with eyes really far apart they'd look kind of silly but they would have great depth perception so this next term is really tricky this is the big unknown this is what I want to discover the first planets around white dwarf stars so having to know sorry, let me drop a number in here first about 200,000 that was the point in the last slide Gaia has revealed about 200,000 white dwarfs that I can observe from the facilities I have access to at the Apache Point Observatory in New Mexico for instance these are objects visible from the northern hemisphere so I'm going to survey some 200,000 white dwarf stars in the search for planets this next term is tricky I'm trying to discover the first planets around white dwarf stars so putting a number on this is not straightforward we don't know of any but we know that these planets should be out there the way we get put a number here is we can use some other evidence for how many planets there appear to be and this important, the most important bit of evidence is that some 25 or 50% of white dwarfs are what's called polluted and what this means is that they have stuff in their atmosphere that doesn't seem to come from the star itself it seems to come from outside of the star so white dwarfs are so compact that they have incredibly high gravity as the gravity on the white dwarf can be as much as a million times stronger than the gravity here on earth as you can imagine if you took a stone and you tossed it into a lake under such strong gravity that stone would sink very fast and the same thing happens on white dwarf stars we drop an asteroid on the surface of a white dwarf that heavy asteroid material sinks immediately below the surface the fact that we see material, heavy material on the surfaces of so many white dwarfs means that there is some process that's continuously feeding that material to the surface of the white dwarf star and when we look at what that material is made of we do a spectral analysis and see what the composition of this pollution is on the surface of white dwarfs it matches roughly the make up of sort of solar system objects asteroids and terrestrial life objects we even have evidence that water is accreting onto white dwarf stars in some cases so there is some process that is able to move planetary material onto white dwarf surfaces in at least 25% of white dwarfs so this may be a good proxy number for how many planetary systems we might hope to detect in a search for planets around white dwarf stars so I'll drop that number in there 25% a quarter perhaps may have planets that we could hope to detect that's the big unknown the next one, fraction of time planets can be detected now we have to talk about detection methods a little bit the method that I'm using to try to detect planets before this is the way that we've discovered most planets around other stars it's the method that the Kepler spacecraft or the current test mission is using to find new planets and the way this works if you've got a planet orbiting a star and that planets orbit happens to be aligned with our line of sight so that that planet passes in front of the disk of the star that could block out a very small amount of the light from the star from reaching our telescopes if we measure the brightness of a star continuously and with great precision we can hope to measure that slight temporary dimming of light and identify that there's a planet there and we could use the shape of this dip that we measure to characterize that planet in this case I've used rough sizes of the sun and earth to represent how big of a dip that would cause it's about 1% of 1% depth there's a very small change in brightness and to measure something like this would require very precise instruments in space above the atmosphere and a lot of data very bright targets but it's possible let's replace this star with the white dwarf that it'll become and see what that signature looks like so this is the problem this is why we don't know any planets around white dwarf stars white dwarfs are tiny they're the size of planets white dwarfs may have planets that are bigger than the stars themselves and they're just such small targets to transit in order for a planet to transit a white dwarf star it has to be so perfectly aligned with our line of sight across this tiny tiny star let's zoom in on a configuration that would work here you see the planet in black moving around the blue white dwarf star it's aligned perfectly this is a simulation and the other problem is not only is this a small target to transit in terms of the alignment but the transit doesn't last very long the thing passes right past the star very quickly and so you also have to capture this precise moment when there is a transit you have to be lucky in this way the upside is that the signature is huge unlike the 1% of 1% dip the earth crossing in front of the sun would give us here in this example it's like a 70% dip and if the planet were bigger than the star it could be 100% dip if this white dwarf were bright enough for us to see in the night sky we could see it just disappearing just winking out completely on the night sky this is something that is easy to measure from small ground based telescopes it's not a sensitivity issue it's a timing issue it's the luck of the orbital configuration kind of issue if we look consider just nearby planets or close in planets close to their stars there's maybe a 1 in 100,000 chance that we're going to detect such a planet to pass in front of the disk of the star this is the rough this is the tough part of this okay but we'll hope to overcome that let's look at how many observations we could get and for this project I'm utilizing data that we're getting from this new survey that UW is involved in called the Zwicky Transient Facility I think some of you who come here regularly may have heard of this in previous talks it's a great source of data using this 48 inch telescopic palomar observatory in California the Zwicky Transient Facility well this telescope was initially built to do very large area surveys of the sky so it focuses light from the sky where it's pointed down to a place where originally images were recorded on photographic plates photographic plates were about 14 inches across these large pieces of glass with a photoemulsion on them and very long exposures would be taken of the sky and that light would come down to record an image on these photographic plates it would record an image of a fairly large area of the sky an area of the sky about the size of your hand on the sky which is a pretty big image for us in astronomy this was meant to survey large areas of the sky the problem with photographic places is that there really weren't very sensitive for all the light that made its way through the telescope only one or two percent of that light actually became embedded in that photoemulsion and got recorded on those plates and so when more efficient electronic detectors like TV cameras were invented this is the same kind of digital camera technology you all have in your pockets in your cell phones then these became preferred for astronomy because they could record like 99 percent of the light that made it through the telescope very efficient at recording images the problem, the downside was that the chips were very small usually so instead of recording these large images on 14 inch plates instead you're recording an image on like a centimeter sized CCD detector so you're giving up this large image for a very zoomed in image of a field that maybe just has a few stars instead of this big part of the galaxy the ingenuity of ZTF is to create a camera that is able to regain this capability to do very large area surveys it uses an array of large CCDs to fill the focal plane that area that would have held a photographic plate now we have both strengths we have an incredibly efficient camera that's able to record most of the light that comes through the telescope and it's also able to record these very large images so in every image that it takes it's a huge image it records thousands, tens of thousands of stars in an individual image in cases and it's a great dataset over three years this ongoing survey is going to record on average a thousand images of every part of the northern sky and this is the dataset that I hope will make the discovery of the first planet's orbit white dwarfs possible and it gives us a number here about a thousand times we'll be able to observe each of these white dwarfs and we plug those numbers in and we get a pretty reasonable number in the ZTF dataset there could be as many as five hundred detections of transits of white dwarf stars and I think that's pretty optimistic this isn't the most thorough like five minutes of math I've ever done I can simulate this whole dataset more thoroughly and I get numbers that are more like 15 detections it depends a lot on your assumption of what this number is and I think this is definitely an overestimate but I could be off by a factor of a hundred and still hope to find these planets in the ZTF dataset so that's what I'm working on I want to run you through this because I think it's interesting to see how to come up with a scientific idea and to test whether it's feasible I wrote this up as a proposal to the NSF for a project that I wanted to do and it got funded and so I have three years to now find these planets and this is the hard part right because now it's really a big needle in a haystack problem right I'm looking for what could be just a handful of detections of moments when white dwarf stars are temporarily fainter than they are most of the time and I'm looking for those handful of detections in tens of billions of individual measurements of the brightness of individual stars that ZTF is proposing so that's the big haystack and on top of that like a bunch of the hay is really sharp too there's a lot of stuff in that dataset that looks like it could be transits but that aren't and so I'm trying to pick out what are actual detections what are real promising candidates is my goal luckily at UW we have a lot of experts on like how to very efficiently search through and analyze large astronomical datasets so I'm benefiting from a lot of our in-house expertise there we also have access to these facilities these observing facilities like the telescopes at Apache Point Observatory that I'm supposed to be using right now and those will allow me to follow the most compelling candidates and confirm their nature and so that's what I've been doing this week I want to show you one first results in this area from the ZTF data it's not the intact planet that I've been looking for but it's a very interesting system that a friend and colleague of mine named Zach Vanderbosch at the University of Texas at Austin found recently we've submitted a paper on this this is a white dwarf system that in the ZTF data shows these very large dips in brightness of 50% and this recurs on a time scale of sort of 110 days this is really different than anything that we expected to find around a white dwarf star we have this picture that planets or planetary material is being disrupted around white dwarfs material making it on to the atmosphere as a white dwarf so we see this debris we know things are being broken up and orbiting around white dwarfs planetary debris material but we expected that to happen for objects that are orbiting really close to white dwarfs with orbital periods of just a few hours when this material is swarming around the star in time scales of a few hours this is 110 days between these events individual events individual intact planets should transit a white dwarf for maybe a minute these transit for 10, 20 days something very different than anything that we expected and so what we think this is something broke up around this white dwarf it seems to be a large swarm of material and it's orbiting perhaps in a very elliptical way something got nudged in close to the star got ripped apart and then this cloud of debris material orbit way far away from the star and what we see here this interesting shape is indicative of sort of the leading edge of this cloud of a destroyed planets material first moving in front of the star and then the trailing edge of that is really like like a tail of debris that more slowly passes by that white dwarf star so that's the model for what we're seeing it's really nothing like anything we expected to find around white dwarfs and it's it's only the I would say like the third example of real planetary material like really concrete detections of planetary material around a white dwarf star it's filling out our understanding of what the whole diversity of planetary systems around white dwarfs could look like and I'm hoping to add to that sorry that's sort of the picture of that I'm hoping to add to that picture of what different kind of real configurations of planets around white dwarfs are in the coming months with hopefully more discoveries and with that I will take your questions thank you yeah yeah the question is about like how would how would Jupiter's transit look like I think things that are that far out are actually very unlikely to transit it's less about the duration but it's about how close to edge on you have to be if you're really far away you have to be very close to edge on so precisely edge on because just the angular tolerance there a very small change in angle longer pass in front of the star so I've taken a look and just said let's just simulate a population of planets around a white dwarf star and just assume that planets just sort of uniformly populate the region around white dwarfs if we do detect a transit you know what's the likelihood of that planet you know what kind of orbital periods might we have some hope of detecting transits for and the truth is that like 90% of planets that we do detect a transit of have to have an orbital period of less than a day so things that have multiple year time scales there's just no chance for them to be that well aligned it makes it difficult to detect those types of planets that we know those planets should be there that story about Mercury, Venus, possibly the Earth being destroyed that doesn't affect Jupiter will be fine it will move out into a further orbit which makes things even harder but one way that we could probe the existence of those planets is actually if we detect a stable transiting planet that's really close into the white dwarf then this regular transiting behavior that we have this very regular mark that sort of tells time in a way it gives us a clock around the white dwarf star and this clock will give us a way to detect far out especially very very massive planets like Jupiter because these planets have a gravitational effect on the star and they'll cause the star as well as the planets that orbit that star and transit that star very frequently to have a reflex motion as these more distant massive planets orbit around them and that reflex motion changes the light travel time distance between us and the star that star may move a little bit further away from us a little bit closer and this very regular clock we have now the timings of the transits of the interplanets those transits will appear to be a little delayed when the star is further from us and those transits will appear to happen a little bit sooner when the star is nearer to us in a way that will reveal the presence of the more distant massive planets of the stellar system so I think that detecting the transiting planets close into white dwarfs is going to be the key to uncovering the non-transiting planets that are more massive that are more distant from white dwarfs in the back here how much data do you collect? sorry? how much data do you collect? that's a good question I don't know, there's a lot of data the question is how much data do we collect CTF is collecting I'm not a computer person really a ton of data there's a ton of data there's 10 million there's 10 million measurements 10 billion CTF measurements and that was sort of the amount of numbers that I have to sort through how do you collect data? how do you collect data? the way that I'm searching for for transits and this data set the first thing is I'm using the Gaia data I already know which stars are white dwarfs so I'm only looking at the stars that I know are white dwarf stars so that throws out most of the stuff and then for all of the repeated measurements of the white dwarf stars I take a look at all the measurements I have I come up with a baseline measurement what the average brightness of that star appears to be and I look for measurements when the star was significantly fainter that's pretty much it the problem is that there's a lot of image artifacts and other sorts of things that cause problems I'm looking at sort of a hundred a hundred million observations of white dwarf stars and when you're looking at that many individual measurements like very rare noise artifacts become quite as common as actual detections question here so how do we get rid of the false positives the first thing I do is I go to the image data so there's this catalog of all of the measurements where some software has gone through the images and measured the brightness of each of the individual stars I'm looking through those measurements of brightness and if I find something that looks compelling I have to go back to the image and see that that star's brightness was measured correctly that the appearance of the star in the image looks like what I expect a star to look like that everything checks out and if everything looks good in the data then I go to a telescope and I try to detect that planet like I said if there's a detection then the orbit of that planet should be and so what I do is I go to a telescope where we have a high speed camera that can just take a continuous stream of images and I watch that star sort of as it moves from horizon to horizon I get a continuous record of what its brightness is so that if it does transit I know I don't miss it and then I can either confirm that planet or find that it doesn't seem to really get validated in the follow up data one more question right there so the atmosphere of a white dwarf so I mentioned that the gravity is extremely high and so heavy things sink in these white dwarf atmospheres so in some ways white dwarf atmospheres are very simple the lightest stuff floats to the top most white dwarfs when we look at them they only show hydrogen in their atmospheres they're really stratified they've got a thin layer of hydrogen in the outer region and then below that there may be a thin layer of helium and then most of what's inside is carbon and oxygen those are sort of the byproducts of most stages of stellar evolution so most white dwarfs have just pure hydrogen atmospheres and then some of them some 25 or 50% when we look in detail there's a layer of heavier elements that we think must be coming from outside of the star thanks 25th next month and you don't want to miss it because our very own trivia is going to be giving a talk so we'll see