University Place

cc >> Good evening and welcome to UW Space Place. Tonight is our monthly guest speaker, and I would like to introduce Professor Rich Townsend of the UW-Madison astronomy department. Rich has a varied background. He has a PhD from University College London in astronomy, but then he went off and had some other career adventures teaching physics in Ghana and working for Reuters and other interesting stuff, deciding that astronomy, he says, is what he really wanted to do. And so he's here tonight to tell us about his research on the magnetospheres of massive luminous stars. Rich. >> Good evening. It's a great honor to be here. Thanks, Jim, for inviting me. I'm going to tell you today a little story about some of the work I do as an astronomer in the astronomy department at UW-Madison. And the tale I'm going to tell you is one of really trying to detect something that shouldn't be there, is almost not there, but nevertheless we can find a great deal out about it. And that's why I kind of draw this analogy to ghost hunting. So my talk really has three parts. I'm going to first start off by telling you about magnetic fields of the Earth and of the sun just to lay some groundwork to then go onto talk about magnetic fields in massive stars. And we're really going to start with the basics. What is a magnetic field? What does it look like on the Earth, the sun? And then what are these massive stars I'm talking about? And then I'm going to go and talk about the region around massive stars where the magnetic field has an important role to play. And these circumstellar regions controlled by a magnetic field are known as magnetospheres. So let's start off with a very basic definition of what a magnetic field is. And this is something that's reasonably easy to explain to somebody who's played with a magnet. And if you've played with a magnet with iron filings, you find that the magnet tends to organize the iron filings in these very sort of well defined arcs and loops. And the force aligning these iron filings is invisible, but we can certainly see its effects and we can measure its effects. And we call that force field a magnetic field. So it's an invisible sort of force that permeates the space around the magnet and that you can detect by putting another magnet near it or maybe putting something metal near it. So, often when we're describing magnetic fields, we introduce this sort of abstraction called magnetic field lines. And these lines indicate the orientation another little test magnet, maybe a compass would assume or an iron filing would assume, if we put it near our magnet. So magnetic field lines don't really exist, but they're a useful sort of abstraction that shows us the sphere of influence around the magnet. And here we see we have the same bar magnet as before and we see these field lines travel from the North Pole to the South Pole. So any point on one of these field lines, if I put a little compass there, then it would spin around and end up pointing along the field line in the direction of the North Pole. So if I put it here, for instance, it would actually point up towards the top of the page because that's the direction along the field line that would eventually take it back around to the North Pole. Now, the reason why compasses are used as a navigation aid is that compasses, which are little magnets themselves, interact with the magnetic field that the Earth has. So the Earth behaves as if it had a big bar magnet at its center, and this is a sort of schematic of what the Earth's magnetic field looks like. Here is this sort of imaginary bar magnetic. In fact, there's not really a bar magnet there. I'll explain where the field comes from in a moment. But we can see these field lines emanating out from the Earth's magnetic North Pole and looping around to the magnetic South Pole. In fact, they're going the other way. The Earth's magnetic North Pole is actually a South Pole. So what causes the Earth to have this magnetic field? As I said, there's not really a magnet at its center, but there are currents in the Earth's molten iron core which generate a magnetic field. And this sort of field generation, not by a permanent magnet, but by these currents and a conducting liquid are sometimes referred to as a dynamo process. So one way we can have a magnetic field is if we have a permanent magnet. Another way is if there's some sort of conducting fluid that's moving around and operating like a dynamo which generates a field. Now, if we measure the strengths of the Earth's magnetic field somewhere on the surface, the typical strengths we find are around one gauss. So gauss is just a unit with which we measure the strength of a field. If we have more gauss, that means that the magnetic pull or push on a magnet or a piece of metal will be stronger. To place the Earth's magnetic field into context, a typical bar magnet is around a field strength of a hundred gauss. So that's the sort of bar magnet I used to play around with as a kid. Kids nowadays, however, if they have very bad parents, get given these rare Earth magnets, these neodymium magnets, and these have strengths of five kilogauss or 5,000 gauss. And these are enough to really do quite a lot of damage if you hold a couple of rare Earth magnets too close together and you get your fingers in the way. And the internet is full of pictures of mangled hands that have come from people playing around with these neodymium magnets. Luckily, the Earth's magnetic field is nowhere near the strength of one of these neodymium magnets, nor is it even close to a normal bar magnet. Okay, so let's go now from the Earth to the sun. The sun, and I'll back a bit, but the sun itself has a magnetic field. But we first discovered the existence of this magnetic field not by wandering around on the sun's surface with a compass. Had we done so, we would have been burned to a crisp. Instead, we learned about the sun's magnetic field by observing the light from the sun. And, in fact, observing the light from many astronomical objects is usually the only way we've got of finding something out about that object. So in the case of the sun, if we take light from the sun and pass it through a prism, that light gets separated out into its constituent components. We find that light is a mixture of red, orange, yellow, green, blue, indigo, and violet, all the colors of the rainbow. Now, this picture in the upper left is just a schematic of splitting up the sunlight into its constituent colors. But this sort of big panel here shows an actual detailed spectrum for the sun. The way this picture is being prepared is that a very, very long detailed spectrum of the sun has been taken, and then it's been sliced up into segments of equal length and stacked one on top of the other. Now, the interesting thing about this spectrum of the sun is that we see dark lines appearing in the spectrum. These lines are known as Fraunhofer lines. Here's an example just here. And Fraunhofer lines are produced by various different elements in the atmosphere of the sun blocking out radiation at certain frequencies of certain colors or certain wavelengths. So Fraunhofer lines are unique to each different element that's in the sun's atmosphere. These two dark lines here and here are the Fraunhofer lines due to sodium. So we know that sodium is in the atmosphere of the sun. This big, broad, dark line down here is a Fraunhofer line due to hydrogen. So by looking at all of these different Fraunhofer lines, we can figure out what elements are present in the sun's atmosphere. Now, with regards to detecting a magnetic field, in the early part of the 20th century, a Dutchman named Pieter Zeeman discovered that if he put a gas in a magnetic field and then observed the Fraunhofer lines associated with light passing through that gas, the lines were actually split into multiple lines by the magnetic field. When Fraunhofer first discovered this, it was doing an unauthorized experiment in the lab with which it was affiliated, and he was fired from his job. He then went on to get the Nobel Prize for this discovery. So I imagine in the long run he wasn't too bothered by this. So in the middle of this slide here, I have an image of part of the surface of the sun. And this big dark feature here is a sunspot, a region of the sun's surface that's darker than its surroundings. And sunspots on the surface of the sun are usually associated with strong magnetic fields. And if we take this spectrum of this sunspot and compare it against a spectrum from a normal part of the sun's surface up here, then we see that there's a difference in the Fraunhofer lines of the two regions. Without a magnetic field, we just see ordinary dark Fraunhofer lines. With a magnetic field, this dark line in the center has been split into three components. And this is a consequence of the effects that Pieter Zeeman discovered. And the strength of the splitting gives us a measure of the strength of the magnetic field at the location of the sunspot. If the splitting is large, if the individual components are more widely spaced, then we know that the field is stronger there. And using this technique of Zeeman splitting to measure magnetic fields, we can actually go ahead and take a spectrum at all different points on the surface of the sun and look at the amount of Zeeman splitting at each point and measure the strength of the magnetic field right across the sun's surface. So these two images here, which were taken with the SOHO spacecraft, a joint NASA/ESA mission, show the visible surface of the sun in normal light, in the left-hand panel, and we can see the various dark spots on the surface. These are sunspots. And on the right-hand image, we have exactly the same picture but showing us magnetic field strength rather than brightness. So this is what we call a magnetogram. And we can see that the sunspots, for instance this one here, are associated with regions of strong magnetic field. And we also see that we have a dark region next to a light region for almost all groups of sunspots. What this is telling us is that sunspots typically have a north magnetic pole and then nearby a south magnetic pole because in the magnetogram, white corresponds to magnetic north and black corresponds to a magnetic south. And if we look at the scale on the bottom, I'm not sure whether you can read this that well, but the typical strengths of the fields are on the order of a couple of hundred gauss. So the magnetic field on the surface of the sun, it's about a hundred times stronger than on the Earth, and it's kind of comparable to the field of a bar magnet. Now, as well as measuring the strength of the field on the surface of the sun using Zeeman splitting, we're very fortunate to be able to see the effects of the field above the surface of the sun. What I'm showing here is a movie obtained in extreme ultraviolet radiation of the sun over the past three days. I went onto the website of NASA's Solar Dynamics Observatory, SDO, which you can see on the left here, and the SDO spacecraft is basically continually looking at the sun at different wavelengths to see how both the surface of the sun and the outer parts of the sun's atmosphere, its corona are behaving. And observing at very, very extreme ultraviolet wavelengths, we're able to pick up very hot gas in the sun's corona that happens to be confined to closed magnetic loops that thread through the corona. So, ordinarily, magnetic field lines are invisible, but if, for a selected subset of field lines, we manage to inject very hot gas along that field line and that gas will emit like crazy in extreme ultraviolet, then we're essentially able to trace out the field lines. So if you look at the edge of the sun here, we can see along a group of field lines, hot gases appeared along them, and we can actually see the structure of those field lines traced out by the hot gas. Likewise, over at the other limb of the star, you can see bright regions on the surface. These are known as active regions. These are very bright parts of the sun's surface that cause hot gas to be fed up onto field lines, and then out in the limb of the sun, we can see those field lines traced out very nicely by the hot gas. So, being able to directly image the sun at these extreme ultraviolet wavelengths allows us to see how material from the sun's surface, hot material, actually fills up the magnetic field that extends beyond the sun's surface. In a sense, we're able to directly image the sun's magnetosphere. Now, where does the sun's magnetic field come from? Well, in the case of the Earth, I said that the magnetic field is generated in the molten core by sort of circulatory motions of the liquid iron. In the sun, we believe that the magnetic field originates in the outer layers, maybe the outer 30% of the sun. And these outer layers are in a state of convective motion where we have hot parcels of gas rising up and cold parcels of gas sinking. These are exactly the same sort of motions we see, let's say, in a tub full of hot water where we have plumes of hot water rising and plumes of cold water sinking. And these sort of turning over turbulent motions we think are responsible for generating the sun's magnetic field. So this is actually the sunspot plot I showed you as before, and I've zoomed in on a tiny region of this sort of surface mottling. The mottling you see here is actually the top of these convective cells, as we call them. The bright parts of this granulation pattern are the hot upwelling plumes, and the dark boundaries separating the cold downwelling plumes. And this plot to the side kind of gives us an idea of the circulatory motions that are associated with these upwellings and downwellings. So in the case of the sun, we can actually see by just looking at the surface of the sun, direct evidence of these convective motions. And, by the way, in this plot, this picture of the Earth is to scale. So a sunspot on the sun is about the same size as the Earth. Wow. Okay, so we've talked about magnetic fields on the Earth, and we've talked about magnetic fields on the sun. Let's talk about, now, magnetic fields on massive stars. So first, let me define what a massive star is. Massive stars are the stars that I like to study. They're the 800-pound gorillas in the galaxy. They're pretty rare, but they account for about half of the light that permeates our galaxy. Typically very massive, hence their name. They generate huge quantities of radiation, especially ultraviolet radiation. And they play a very important role in driving the evolution of a galaxy determining what new generations of stars will be born at what time. So typically, a loose definition of a massive star is something that has a mass of three times or more that of the sun. The masses can go up to maybe a hundred times that of the sun. The radius of a massive star is typically two or more times that of the sun. But again, that radii can get up to tens or a hundred times larger than the sun. And the surface temperatures are very high. On the order of between 10 and 50,000 Kelvin. So, in Fahrenheit, that's probably about 50,000 to 100,000 Fahrenheit. So they're quite hot. Because of these high temperatures, they're blue in color. And this image to the right shows a collection of massive stars. In fact, the collection of massive stars that's closest to the Earth, those of you who look up at the sky regularly will recognize this as the constellation of Orion. And almost all of the bright blue stars in Orion, especially the ones making up Orion's belt and then Rigel down here, these are massive stars. Orion hosts a number of massive star forming regions, and it's the one that's closest to the Earth. So when we want to study nearby massive stars, Orion is where we usually start. In fact, a lot of the stars I'm going to end up talking about later on in this presentation will be in Orion. So, massive stars typically have very short lifetimes because even though they have more fuel to burn than the sun, they burn it much more quickly than the sun. So in the most extreme cases, massive stars will only last for a few million years. Some of the stars, I apologize, some of the stars that you see here, they were born while the hominid ancestors of humans were still developing. Whereas our own sun was born billions of years ago. So these stars are kind of contemporaneous with humanity. The massive stars, even though their lifetimes are short, have a big impact on their surroundings, both through the winds, during their lives they shed a lot of mass and wind outflows, but also at the end of their lives, if they're above about nine times the mass of the sun when they begin their lives, at the end of their lives they won't fizzle out but rather explode as a core collapsed supernova which is some of the most energetic phenomena in the universe. So let me say a little bit more about the winds of massive stars because it turns out when we throw magnetic fields into the picture, the interplay between them becomes quite interesting. So because massive stars have such hot surfaces, the radiation field at their surface is incredibly intense. In fact, so much so that the radiation itself is able to drive mass off the surface. So these are radiation driven winds. Just the pressure of photons hitting gas at the surface of these stars is enough to strip that gas off and launch it out into a hypersonic wind. So the wind velocities reach up to about 2,000 kilometers a second. So that's two-thirds of one percent of the speed of light, or, in US units, 4.5 million miles per hour. So these are some of the fastest outflows in the galaxy. And the mass loss rates can reach up to a millionth of the mass of the sun a year. Now, a millionth of the mass of the sun, that doesn't sound like very much. But let's put that into grams. That's about 10 to the 27 grams a year. That's a huge amount of mass. In fact, the mass loss rates this high mean that even for a massive star that's got a lifetime of only a few million years, they're going to loss and appreciable amount of their mass as they burn their way through to becoming a supernova. And these winds, as they travel out into the interstellar medium, sweep up the gas between stars and blow these beautiful bubbles that lead to gorgeous emission nebulae such as this one, the bubble nebula. So where my mouse pointer is now, just about there, there's a massive star and it's blown this bubble around itself through the wind that's coming off its surface. Okay, now let's talk about magnetic fields in massive stars. Now, all stars, apart from the sun, are too distant for us to obtain direct images of their surfaces. So that gorgeous picture of the sun's magnetosphere taken by the Solar Dynamics Observatory over the past few days, that one where you saw these magnetic loops lighting up, we're just not going to be able to do that for other stars. They're just too distant. And likewise, we're not going to be able to measure the Zeeman splitting individually at different points on the surface of a star. Nevertheless, if a star is magnetic, then we can still hope to see some kind of Zeeman splitting in the sort of average spectrum that comes from all the different parts of the star's visible surface. And certainly, in this upper panel here, this is a for a few different stars. These are somewhat cooler than the massive stars I'm talking about, but we can see the clear signatures of Zeeman splitting in the Fraunhofer lines of these stars. So, at least for certain types of stars, it doesn't matter that they're too distant from us that we can actually resolve their surfaces, we can still see the Zeeman splitting in their spectrum. However, for massive stars, they typically rotate quite fast, quite rapidly. The equatorial rotation velocities of these stars are on the order of maybe a hundred kilometers a second. So that's probably on the order of 200,000 miles an hour. This is how fast they're spinning at their equators. And an effect of this very rapid rotation is to smear out the Fraunhofer lines via the Doppler shift. When a star is rapidly rotating, half of the star is coming towards the Earth, the other half is going away, and this has the effect of actually blurring out the spectral lines. So this little schematic here shows what the Fraunhofer lines look like for a slowly rotating star, that's the sort of narrow one, and for a rapidly rotating star, that's the broad one. And in fact, almost all massive stars rotate so rapidly that we have no chance in practice of actually ever measuring the Zeeman splitting coming from a magnetic field. So it looks like massive stars are off the cards in terms of finding magnetic fields on them. And for a long time, people were happy with this because massive stars don't have the same convective envelop that the sun has. So they don't have this large outer region where we have these turbulent motions that can generate a field. So for a long time it was assumed that, well, we can't measure fields in these stars, but it doesn't really matter because they're not going to be magnetic anyway. However, things changed about three decades ago for a small subset of stars. And then in the past decade, the lid has really been blown off. This idea that massive stars are nonmagnetic and that we can't measure their fields even if they were magnetic. And the key to unlocking the magnetic mysteries of massive stars comes from a sort of secondary effect that Zeeman pointed out, which is not only when you place a gas in a magnetic field are the Fraunhofer lines split, but the light associated with those lines is also polarized. And some of it is linearly polarized. I think those of you who've played around with Polaroid glasses are familiar with linear polarization. Basically, a linearly polarized light is light in which the electromagnetic wave which makes up the light is oscillating in a plain. So in the left-hand animation, we see linear polarization. But it turns out a magnetic field not only introduces linear polarization but it also introduces circular polarization where the electric fields of an electromagnetic wave actually describes, at any given point, a circle as the wave goes past. And it turns out the circularly polarized light, it's possible to detect it in massive stars for quite modest field strengths even though the stars are rapidly rotating. And so being able to take spectra of stars in circularly polarized light was the technological breakthrough that allowed us to start detecting magnetic fields on massive stars. So in the past decade, in fact shorter than that, since 2007, I was fortunate to be involved in a survey known as the MiMeS Survey. It's a bad acronym that stands for magnetism in massive stars, where a group of 50 researchers from institutions around the globe, including the University of Wisconsin-Madison, used world class telescopes combined with circular polarization spectrographs to look for magnetic fields in a large number of these massive stars. There were a few massive stars that were already known, but this was the first systematic attempt to get proper statistics on how common these fields are. So the MiMeS survey involved nearly 5,000 circular polarization spectra of around 500 of these massive stars. And the big result of the survey was around 10% of the stars surveyed were discovered to have magnetic fields. So, for those of you wondering, this is the main telescope we used for the MiMeS survey. This is the 3.4-meter Canada-France-Hawaii Telescope located on Mauna Kea in Hawaii. And this is the ESPaDOnS instrument. It's the world's most sophisticated spectropolarimeter, as they're known. And it was with this instrument that these fields were able to be measured. So this is a break down of the general results from the MiMeS survey by spectral type, which is a sort of jargony quantity. It's a way of characterizing the Fraunhofer lines of stars. And it gives us an indication of how hot a star is. So on this scale along the bottom, I've actually got the scale back to front. So let's fix it right now. There we go. At the hottest ends, surface temperatures are around 40,000 or 50,000 Kelvin. That's around 100,000 Fahrenheit. And at the coolest end, maybe around 10,000 Kelvin, 20,000 Fahrenheit. So, as I said before, the incidence rate is around 10%. About one in 10 massive stars has a magnetic field, and these fields are strong. In the weakest case that we can detect, the fields are around 100 gauss. That's around the same as the sun's magnetic field. It's about the same as a bar magnet. In the strongest case, we see fields of around 10 kilogauss, 10,000 gauss. That's stronger than one of these rare Earth magnets that will mangle your fingers. So you really wouldn't want to be stood next to one of these strong field stars because my glasses would fly off, for instance. All sorts of mayhem would ensue. What's interesting about these fields is there's no real obvious dependence on any of the stars properties. No dependence on their spectral type, their temperature, their rotation rate, whether they're undergoing some kind of oscillation seismic-like activity. And in fact, this lack of dependence on anything we can measure makes us think that these magnetic fields are not being generated by a dynamo, as they are on the sun, but instead they're fossil remnants from the star's formation process. The protostellar nebula that eventually collapsed to form these stars was itself threaded with a magnetic field. And as it collapsed, that magnetic field got trapped inside the star and led to the fields that we measure today. Now, one of the really important spinoffs of the MiMeS survey was the realization that many of these stars harbor magnetospheres. So these are structures around the stars and the circumstellar environment where gas is confined and controlled by the star's magnetic field. And this gas emits radiation that we can in principle detect. So, for the remaining third of my talk, I'm going to focus on these massive star magnetospheres and talk about how we can detect them and how we can model them. But before I do that, let me just talk a little bit more in detail about the sort of data that the MiMeS program gives us on an individual star. So, on this page, we see observations of the star HD37776, which is actually a designator from the bright star catalog. It would be nice if all stars had proper names. This star happens not to. But nevertheless, it's a very interesting star. In fact, I'm going to come back to it later on in my talk. Right now, though, I just want to show you what typical results from a spectropolarimeter look like. So in the right-hand panel, we have a sequence of spectra that were taken across a few nights. The star is rotating, so by observing it at different times, we can catch different sides of the star and see what the magnetic field looks like all the way around the star. So, in the left-hand column, we have the normal Fraunhofer line. This is actually an amalgam of a number of different lines. We blended them together. We have the Fraunhofer lines for the star in the left-hand column, and they don't really change much as the star turns around. But in the right-hand column, we have the Fraunhofer line as measured in circularly polarized light. And this changes quite significantly as the star rotates. And in fact, by looking in detail at the variations of the circularly polarized light from the star, we can actually build up a map of what the magnetic field strength looks like across the surface of the star. This is a remarkable technique known as magnetic Doppler imaging. So it's relying, one, on the fact that the Zeeman effect polarizes light, and, secondly, on the fact that as the star rotates, one side of it's approaching as the other side is receding. So there's a Doppler shift on the Fraunhofer lines. And combined together, the Zeeman effect and the Doppler effect allow us to map out the magnetic field strength across the surface of the star. And this is what the star's field strength looks like. So, in the top row is just an indicator of what the field strength is at each point on the star, and these are looking at the star for five equally spaced phases during one rotation cycle. And then down on the bottom, there's an indication of which way the field is pointing at each point on the surface of the star. And, in fact, if you look at the scale of this, for this particular star, the surface field strength goes up to 30 kilogauss. That's 30,000 gauss, which is absurdly high. It's really quite remarkable. Now, magnetic Doppler imaging is great because it can give us an idea of what the magnetic field is like on the surface of the star. But can we ever hope to know what the field structures above the surface look like? Remember, in the case of the sun, we've got those wonderful animations from the Solar Dynamics Observatory where hot gas is illuminating those closed magnetic loops. We can't do that in this case. But it turns out in certain situations we can actually extrapolate the magnetic field from the surface of the star up into its circumstellar environment. And, actually, we can get an idea of what the magnetic topology looks like around the star just from the measured field on the surface that we get from this magnetic Doppler imaging. So this animation I'm going to show you now, this is for HD37776 again. So we're going to start off just with the bare surface magnetic map, and then I'm going to show you what happens if we extrapolate field lines from that map and actually build up a picture of what the field lines look like around the star. So this is the same surface magnetic field strength map that I showed you on the previous slide, and now we're going to extrapolate field lines out. And so we're able to actually reconstruct what the field looks like around the star. And does it look like a dipole? That's the word we use to describe a sort of bar magnet-like field. Well, certainly not. This field is particular non-dipolar. It doesn't look like a bar magnet. And that's why this star is so interesting. So the magnetic field, in fact, is quite a mess for this star. It's got all of these tangled loops. Okay. Let's talk about magnetospheres now. So the reason massive stars have magnetospheres is that they have these strong radiation winds driven off their surface. And these winds are not made of ordinary gas. Massive stars are so hot that their winds ionize. They're, in fact, a plasma. So for atoms in the wind, let's say hydrogen atoms, temperatures are so high that the nucleus of the atom, in the case of hydrogen it's just a proton, can't hang onto its electron, and the two get separated. So the winds of massive stars are plasmas. These are fluids or gas-like fluids that are electrically neutral, but they're nevertheless made up by free charged particles. A mixture of positively charged nuclei and negatively charged electrons. And the numbers balance out exactly so that the overall charge of the wind is zero. But nevertheless, the wind is made up from charged particles. And when we put a charge particle in a magnetic field, it interacts strongly with that magnetic field via a mechanism known as the Lorentz force. And the Lorentz force stops it from going in the direction it wants it to go and actually forces it to spiral around the field lines. So if we have field lines going in this horizontal direction, if I try and fire an electron through that magnetic field, it will actually end up spiraling around the field lines in a helical motion. And because the parts are forced to helical by the magnetic field, the net effect is that the plasma making up the wind is forced to follow the field lines. Essentially, the magnetic field acts as a conduit for the wind, and wherever the magnetic field goes, the wind has to go. Now, this raising the possibility of a very interesting effect. For a lot of the magnetic fields we've seen so far, a prominent feature is there are closed magnetic loops near the star. Points where magnetic field lines go up above the star and then come back down to the stellar surface. Now, if we have closed magnetic loops in a massive star, we can expect there to be wind streams launched from either foot point of the loop. And these wind streams will be channeled by the magnetic field until they actually collide with each other. So along closed magnetic loops, we are going to have wind collisions, and these collisions will produce shockwaves that heat the wind plasma up to temperatures of millions and millions of degrees. The reason for this heating is that the kinetic energy of the wind, which we know will be large because these winds are moving very fast, the kinetic energy gets converted into thermal energy, into heat. So once the wind streams have collided and shock heated, then we've got a region containing a lot of very, very hot gas that's trapped on the magnetic field lines, and it's not really moving anywhere. And this gas will radiate like crazy, initially, in X-rays, material at millions of degrees radiates in X-rays. And as it cools down, it will then radiate in the ultraviolet, and eventually, once it gets cool enough, it will radiate in the visible. So this circumstellar distribution of hot gas that's formed by the channeling and confining of wind streams on closed magnetic loops is what forms a magnetosphere around one of these massive stars. So let's go back to the constellation of Orion to look at some specific cases now. So I've already mentioned that Orion is the closest collection of massive stars to the Earth. The distance to the stars in Orion is between about 350 and 500 parsecs. So just as a reminder, a parsec is not a unit of time, as it is in the Star Wars universe, it's a unit of distance. A parsec is around three and a quarter light years. So these stars, let's say 500 parsecs, that translates maybe a little over 1,500 light years. Okay, so Orion is particularly interesting though, not just from the fact that it's the nearest collection of massive stars to the Earth, but also it has a lot of magnetic massive stars in it. So two out of the three in Orion's belt turn out to be magnetic, and there are three stars in the sword, which is actually the Orion nebula cluster, which are magnetic. So it's perhaps surprising that Orion should have so many magnetic stars, but this is really a reflection of the fact that magnetic fields are difficult to detect in stars, and, therefore, the place where we're naturally going to find a whole lot of them at first with present day instrumentation will be the ones that are closest, and those will be the ones that are in Orion. So the stars I'm going to talk about right now Theta 1 Ori C, which is the brightest star in the Orion nebula, which is magnetic. Then Sigma Orionis E, which is a star just up here that's magnetic. And then HD37776, which we've already encountered, which is also in Orion. So let's start off with Theta 1 Orionis C. This stars is the brightest star is the trapezium. Now the trapezium is a group of four stars at the center of the Orion nebula cluster that illuminate the cluster. In fact, Theta 1 Orionis C, which is in this picture, the lower star in the trapezium, the trapezium is stars there, there, there, and there, Theta 1 Orionis C is in fact the star that's responsible for lighting up the whole of the Orion nebula and making it such a beautiful object to look at. Back in 2002, I think it was, a kilogauss magnetic field, kind of dipole-like, so kind of like the fields you see on the Earth, was detected in Theta 1 Orionis C. But even before the field was detected, astronomers kind of suspected that there might be a magnetic field there, and the reason was Theta 1 Orionis C is a strong X-ray source. Now remember a few slides ago I talked about when we channel these magnetic streams to collide with each other, they're going to generate X-rays as the wind streams are shock heated to millions of degrees. Well, that's exactly what we think is going on in Theta 1 Orionis C. This is an image of the Orion nebula cluster obtained using NASA's Chandra spacecraft. So each of these points in this image indicate some kind of an X-ray source, and at the very center of the image, we have the brightest X-ray source in the Orion nebula cluster, which turns out to be Theta 1 Orionis C. And this was known about in the '90s, and it was suggested that Theta 1 Orinis C may be magnetic as a mechanism by which it produces these strong X-rays. And then a few years later, using new generations of spectropolarimeters, we were able to detect the surface field on Theta 1 Orionis C for the first time. So Theta 1 Orionis C kind of led the way for using other ways of detecting magnetic fields. For instance, using X-rays to find potential targets for followup surveys. Now let me talk about Sigma Orionis C, and this is my favorite star. So I'm actually a theorist. I work on computer simulations and pencil and paper, models for things, so for me to have a favorite star is probably not something that I should do. But nevertheless, I find this star absolutely fascinating. And when I arrived at UW-Madison in 2008, one of the emeritus professors there, Bob Bless, asked me what I like to work on. And I said, well, I've got this really interesting star called Sigma Ori E, and I think it's fascinating. He says Sigma Orionis E, hmm, I used to study that 30 years ago. Any friend of Sigma Orionis E is a friend of mine. So that was a very nice welcome knowing that there were like-minded people in Madison. So to place Sigma Orionis E on the sky, if you look at the lower left-hand star of Orion's belt, this is Zeta Ori, otherwise known as Alnitak, just down from Zeta Ori and to the right is the Sigma Ori cluster. Now, just down and to the left is a very famous dark nebula, the Horsehead Nebula. It's turned on its side in this picture. And the Horsehead Nebula is actually just a dark nebula in front of an emission nebula behind it. That emission nebula is lit up by the bright stars in the Sigma Ori cluster. But off to the side from the bright stars is this intermediate brightness massive companion known as Sigma Orionis E. And this star was actually first detected in the 1970s to have a 10 kilogauss dipole magnetic field. But back then it was thought as a real oddball. People weren't quite sure what to make of it. We didn't realize that magnetic fields in massive stars were really not that uncommon. So this is a surface field map of Sigma Ori E obtained as part of the MiMeS survey. These are just different ways of representing the same data. And again, we're showing the star at five different rotation phases. And just by looking at the direction of the magnetic field on the stellar surface at these different phases, we can see that it's kind of consistent with the field we would get if the star had a bar magnet at its center. That's why we classify this star as having a dipole magnetic field. However, this virtual bar magnet isn't aligned with the star's rotation axis. It's tilted at an angle. So as the star rotates, the sort of bar magnet associated with the star goes in a sort of procession motion. And you'll see that the field sort of processes around as well. So in this animation here, so this isn't based on direct observation, this is based on indirect observation plus a whole lot of science, we see Sigma Ori E, the star, with its surface colored according to the magnetic field strength, we see a whole load of field lines that I've reconstructed using this magnetic extrapolation, and then using my understanding of how the wind gets channeled along these field lines, I've also been able to figure out where wind plasma will tend to accumulate in the circumstellar environment. And this wind plasma is the stuff that's emitting light that we can detect here on Earth. So, potentially, this is a model that we can compare against observations of the star. And in fact, this is just sort of one facet of our model for the magnetosphere of Sigma Ori E. The next slide shows a slightly more detailed picture. This is using slightly older graphics. I've updated them recently. But the distribution of gas in the star's magnetosphere plasma is shown in this top left panel here. And then all of the other panels show quantities that we can measure for the star. For instance, where soft and hard X-rays, that's low energy and high energy X-rays, might be emitted. These maps in the top right corner, these four panels in the top right corner, show us polarized light coming from the system. In this bottom panel with the yellow curve shows us how the brightness of the star should change over time. And you'll see that the brightness of the star undergoes a couple of dips each rotation cycle, and these dips occur, if you look into the top left panel again, you'll see the dips in brightness occur when the material in the magnetosphere passes in front of the star and blocks out some of its light. So using a model for the distribution of plasma in the magnetosphere, we're able to build up a picture for how the magnetosphere should look, whether we're studying X-ray wavelengths, optical wavelengths and polarized light. We can even build up simulations that tell us how the magnetosphere will appear if we observe it using radio waves. And when we compare these predictions against actual observations of the star, we find that the agreement is very good. So I'm only going to give you one example of the observations versus model for Sigma Ori E, and this is perhaps cherry picking because this is one of the best matches between the two. So what I'm showing you on this slide is something we call a dynamical emission spectrum. Each of these panels shows us where emission from the star can be found, what wavelength it can be found at what phase in the star's rotation cycle. So, in fact, we're expressing wavelengths as a Doppler shift velocity. Material that's at rest relative to the star will emit radiation at zero velocity. Material that's coming towards us relative to the star will emit a negative velocity and be blue-shifted, and material that's going away from us will emit a positive velocity. And the red in this diagram shows where there's lots of emission, and the blue shows where there's a relative deficit of emission. And you can see that the star shows these sort of two emission peaks, one on the blue side and one on the red side, that sort of cross backwards and forwards as the star rotates. And, in fact, if we go back to the previous slide, this diagram here actually kind of shows you slices through that dynamical spectrum. It shows you these twin emission peaks that kind of go backwards and forwards. And there's a very well defined sort of S wave pattern to the emission that you can see in the observations. And if we look at the model for this variability, and this model is really just based on the magnetic field maps that we've measured for the star and extrapolating the field out, building a model for the magnetosphere, we see we get the same sort of double S wave emission as we do in the observations. And that gives us a degree of confidence that our model for the star's magnetosphere is essentially correct. Now, there are some fine details that don't match. The shape of this emission here is not the same as this here. But given that this star is at one and a half thousand light years away and it's just a single point of light, I think we're doing pretty good. Okay, let's finish up by talking about HD37776. So this is the star that's got the 30, I guess that shouldn't be 30,000 kilogauss. That really would be impressive. This is the star that's got the 30 kilogauss, or 30,000 gauss, magnetic field. This is actually located above the bright lower star in Orion's belt. Just here. And this star has a very non-dipolar magnetic field. Its field can't be approximated as the field from a bar magnet. It's got this sort of crazy field. We saw the field reconstruction earlier on. This is what the surface magnetic field map looks like. Again, we've seen this picture before, and just like the other field maps we've seen, this was determined using magnetic Doppler imaging. And this is what the star's magnetosphere looks like. So you can see the really tangled magnetic field, and you can also see that the material trapped on this magnetic field, the magnetosphere, is really kind of strange. It's this kind of warped disc-like structure with spurs sticking out, and it's really very dissimilar from the magnetosphere that we saw for Sigma Ori E. And we can compare the observables, the emission at different wavelengths predicted by this model against the observations. And this is what we get when we do that comparison. So these are dynamical spectra before. In this case, the degree of agreement between theory and observation is not really that good. Certainly, we see the same sort of backwards and forwards emission patterns, but the phasing between the observations and the model doesn't seem to be quite right. In fact, in places it's pretty far off. Now, does this cause us to pack up our bags and go home? Have we failed? No, of course we haven't. This tells us that there's something else going on. That there's something in our model that's not quite right and we need to refine it, and that's what we're working on at the moment. So this is the star that we're working on right now and hoping to be able to bring it into the same level of agreement with observations as we saw for Sigma Ori E. And when we try and sort of get it to agree with the observations, we're not really allowed to cheat. We can't do things to the model that aren't physically justified. We've got to try and figure out what physics is missing from our models because these aren't ad hoc models. These models are all based on well established physics. The trick is knowing which physics to put in, which physics is important, and which physics is unimportant. So let me finish with a summary. So although massive stars are really too distant for direct imaging, they rotate too rapidly for Zeeman splitting to be measurable, and they're not expected to be magnetic anyhow, we, nevertheless, find magnetic fields in about 10% with strengths that go from the strength of the sun to maybe a hundred times or more the strength of the sun's magnetic field. This is a pretty remarkable result. By now you must have noticed the ghost. I had to put a ghost in my talk somewhere because of the title. But I do see this, this process of mapping out these magnetospheres, as kind of a ghost hunting because I'm using very, very indirect evidence for something that's not really supposed to be there in the first place, and it's kind of like a detective puzzle. So by mapping the magnetic fields of massive stars on their surfaces using magnetic Doppler imaging, we can then extrapolate to what the fields look like around the stars, we can figure out what the distribution of material in their magnetospheres will be, and we can predict what the emission from these magnetospheres will be in such a way that we can compare against actual observations and say is our model going in the right direction or does it need refining. And that's all I have to say. Thank you very much for listening.

APPLAUSE