 On Monday evening, we already had a view into the theory of detecting exoplanets. Today, we are going to search for them ourselves. A meteorology PhD Michael Teuchner is fascinated by astronomy and can fascinate people for astronomy. And since his encounter with Halley's comet, he has been on the search for extrasolar planets. And 14 years ago, he was successful for the first time using fairly simple methods. He was able to prove the existence of an exoplanet. Michael, please help us find more. And it's the stage three for Michael Teuchner. Thank you and thank you very warm welcome to this talk on a topic that I find is very interesting and one that hasn't been practically current in science for a long time. Well, it's been about 30 years, but until 30 years ago, nobody knew if other stars would actually have planets around them. And of course, a lot has happened since then. And I've been, I had, I was following this from a distance for a while. And around 2007, for the first time, I wondered whether I as an amateur would have an opportunity to prove, detect such an exoplanet. Well, not, not by chance. That is almost impossible. But stars that are known to have an exoplanet, whether I was able to prove its existence and with using amateur means. So to start with what actually is an exoplanet, a very easy description. Basically, it's just a planet outside the solar system, not one of the eight planets we have, but one that circles around another star. And by now, we know quite a lot of those just this morning. I checked and we have 3,628 planetary systems that have been confirmed. And 808 of these systems have more than one planet. And when there are some very exotic things among them, because of course, initially you assume that a planetary system should look very much like ours. Everything is stable and small planets inside large ones outside. But that's all been thrown out. And that has changed the way we explain the generation of planets. And that applies especially to those that I as an amateur can detect easily. And by this morning, there are 4,904 exoplanets that have been found. And if you imagine that the first one was found in 1995. Well, there was a fairly exotic case in 1992, a so-called Palsa. So that was 20 years ago, kind of was found around a star corpse, as you might say. But you can find this around all kinds of systems these days. And in 1995, the first one around a regular star was discovered. And how do you find those? I think I remember that there was a talk about this. And of course, it's not so simple. If you look at a star, how do you actually know that another planet is circling around it and orbiting around it? And of course, other astronomers found all kinds of intricate ways of finding them through various methods. So it means to actually find such an exoplanet and establish its existence. These are the methods that have been successful so far. You can broadly divide them into three categories. The first one is called dynamic detection. So you look at things that change over time, such orbital changes, the trajectory being changed. Then there is a very exotic method called microlensing. And that means that you have two stars in our light of sight, one very far away, and one closer to us. Stars move, of course. They are not fixed, although the name sometimes you shouldn't suggest that. And when one star passes another, there is a grational lensing effect that happens. The gravitational field of the closest star amplifies the light of the star behind it. And if that star has planets behind it, there are some very interesting peaks in brightness that you can observe. And the light of the star behind it is thus amplified. And there are these interesting outliers that you can use. But the disadvantage of this method is that this only works once, because this constellation of one star being in front of another is there for a limited amount of time, just once. And photometry is the other category. So that means direct imaging for one thing. You have to have very sophisticated, such as the 8-meter mirror in Chile that you have to use. You can actually image planets directly that orbit other stars. But you need specific conditions for that. And the third one that we'll be looking for are the so-called transits. And that, I think, is the only method, more or less, that amateurs, where amateurs have an opportunity to find exoplanets orbiting other stars. All the others, the dynamics, for example, methods are too specific. They will be reserved to professional astronomers for a long time. And the microlensing, too, you would have to know in advance that this is happening and have to be looking at the right time. And we will be learning about transits, what they are. And with that, you can repeatedly observe the systems. And what you can also see in this diagram is up to which size you can detect planets with all these methods. So you can see the PASA method, what was used in the one used in 1992. You can discover planets until less than one Earth mass. The other methods don't reach as far down. Microlensing can take it down to one Earth size. But the space telescopes we have these days, and some of you may have noticed that the James Webb telescope was launched on Christmas Day. That will be an opportunity of observing much smaller planets. But these transit methods, they take you down to planets actually smaller than our moon, if you observe from space. And that is from the ground. It's not that easy. You come down to around, well, just under 10 Earth masses. And sadly, that is the fact that is determined by our Earth's atmosphere, always distorting observations. Some of you may know this, if you're out on the road in the summer, you can see the Earth turning, you can see the air turning and moving. And that makes conditions on the ground worse. It's so easy to compensate that. And that gives, so professionals have a great advantage. OK, so the transit method requires a certain constellation here. So we have a star. And from the line of sight of the Earth, the planet needs to transit in front of the star. So we need to lie in the plane of the star. And if the planet transits in front of the star, the brightness of the stars gets reduced a slightly bit. So by this reduction of brightness with today's methods, and even me as an amateur, can measure the transit here. So we know this from our solar system as well. Maybe a couple of you have also observed transits like that. So for example, here's the transit of Venus from June 2012. And this is how it looks from Earth. And so doing this event for a couple of hours, the amount of light is much smaller. So we know, of course, that in our solar system, that only works for the inner planet, Venus and Mercury. And since Mercury is much, much smaller than Venus, you see that the effect is very, very small as well. So it's possible today with our modern methods, but the effect is very small. So if you look at the profile of these transits here at the relationship between the size of the star and the size of the corresponding planet, so if you have this effect of 0.12, we see an attraction in the brightness of just 2.7%. So the whole thing tells us that we as hobbyists could only be able to measure the transit of very large planets, relatively pertinent. And also the duration of the transit depends on how far away the planet is from the star. And so those transiting planets we can observe as amateurs right now, so we see something like one to four days this time here. And so for some of these planets, we can weather one year for this planet is something like one or two days on Earth. So for this observation, I need to observe the whole transit. So if the transit takes something like four hours, I need to observe at least six hours during the night, the time before the time after. And for that, the sky has to be clear all the time. I have to be ready all the time, so that's it. So here, an example, this is how the Jupiter would look like transiting past sun from very far away. So you would see this lowering of the brightness for a few hours, and that's it. And this is how this would look from really far away. You see this dot here, which gets a little less bright over some time. That's all. You don't see any changes in brightness here. So if you do the actual measurement, you're doing lots and lots of photos here with a certain exposure and with a certain software packet that I'm going to present later on. We then analyze the data here on whether the star got less bright over the time or not. And so let's have a look at a couple of wheel transit of brightness spots here. This is from Hubble, from the Space Telescope. You see lots and lots of dots for individual measurements, individual samples here. And you see how the brightness goes down here. So minimum is here. And then when the planet passes away from that away, the brightness goes up again. So that it doesn't look at a corner here, depends on where the planet passes. The planets are here. So the further we are at the exterior areas of the star, the less square-like this curve looks like. And what also might notice is that the stars seem to be less bright in the outer areas than right in the center here. So we also see this here. So this gives us the brightness curve here. So we see HD 209459B is what the name of the planet. So this was just the ID of the star here, counting to some star catalog, because we know a lot of them. And most of them are so dark that it's impossible to watch them with your bare eye. And so the first exoplanet you see on a star gets the letter B. And this animation then gets extended to the other. So you have B, C, and D in the time of their detection. So this has nothing to do with the distance between the star and the planet. But it's just the order in which we detected those. So you don't see a very big effect here. And the reason for that is that this planet here is very, very small. So the whole difference is why this is also very small and noisy, whereas here on a very large planet, you see that there's a lot less noise in the signal here. And the effect is much bigger, much easier to see. So I ask myself, can I do this as an amateur for myself, as an hobbyist? So I don't have any space telescope, unfortunately. And I don't even have 35 meter mirror. So no amateur, no hobbyist has this kind of equipment here. So as an hobbyist, I depend on the weather condition. I only have a small one here. So that's what I had. I had a small reflector with 60 millimeters diameter, 370 millimeters of focus. And so back then when I tried this, the digital cameras were not as good as they were today. So back then I had a DMK camera, an 8-bit industrial webcam, back and wide, 8-bit resolution. So what you could also do with that, I think you need a little bit of equipment, of course. You need the astronomical mount for your reflector, which points to the polar star and compensates for the rotation of Earth so that the stars always appear at the same position on your picture. And of course, a computer with a software to analyze the pictures there. And today most amateurs, most hobbyists, have this kind of equipment. And it's not expensive actually. It's just maybe a few hundred euros is what gets you equipment to actually measure the first transit. So this is the first one that I did was the transit of the exoplanet HD18973B. I observed this on October 2008. You see this here. And this is, and the whole transit took about two hours here. And so as I said before, the planets that can be detected by amateurs needs to be very, very large. They need to occur large parts of the stars. So what we call them hot Jupiters. A Jupiter, so planets of the size of Jupiter are larger in a very small, very close to their stars. So what I'm saying here is that I'm doing lots and lots of photos here. So yeah, you do not over-exposure that one. So if you see this, this is how the brightest profile should look like. And this looks like a bit of normal distribution here. And if you over-exposure, then the sensor goes into saturation. You see some flatness here. In my example, with 8 bits, that would be 256. And you would see any change at all, even during the transit here. So I couldn't really detect the transit here. You may see some changes there, but that's definitely not enough to measure. So it's OK to get slightly out of focus. This might even make things simpler here. And so this is my equipment from 14 years ago. So you could think that I just have to observe the brightness of the stars, but that's not enough. And it doesn't even work because we have turbulences in our atmosphere. And by that, the brightness of the star varies so much. It's so noisy that I couldn't really observe the transit of the stars. So what we call a relative photometry, so we compare the brightness of our star and some neighboring stars and compare the relative brightness between those two. And so we can just measure the brightness of one star from where we are, from Earth. And there's some dedicated software for that. So the optimal exposure would be something like two or three minutes because this averages out the turbulences and other noise from the shorter noise. And even if the star is not very wide, I'm getting enough signal from that one. What I also need to know is when does these kinds of transits happen? And for that, you have the Exoplanet Transit Database, great website that you can find for each evening. What transits there are going to be at your location as well. I've ended Hamburg as a location here. And just by tonight, there are 30 transits that you could observe in theory. And there's one here where the transit will take about 97 minutes. Star brightness is 12.8 magnitudes. That's 300 times darker than you can see with the naked eye. But there are some examples where 2% to 3% of the light are actually blocked by the transit, given whether that would be a good observation target. And it's high enough above the horizon to pick one for yourself in there, maybe after the talk. And what you also need to detect an Exoplanet like this is, of course, you need to evaluate the data and look at the planets. And you have this extra solar planets in cycle P, giving you all the details on the known exoplanets, their orbital periods and things like that. And I'll just show you how it looks like at my home. So I've got this. It looks fairly technical in my setup at home here. You don't even have to be next to the telescope anymore. I use Team Hero or something to control it. It's all become a bit more comfortable. This is not the 6, but it's 10 centimeter telescope, a specific astronomy camera attached to it. There's a huge selection here, starting from, say, around 100 euros. And you need, of course, a tool to take the actual images. APT is a soft suit. And of course, you have to follow the movement and control that. As you observe a star, you have to follow its movement across the night sky to actually keep its constant moving less than, say, two, three pixels. And of course, then, to process it, you need something like Astro Image J. That is the free software that the professionals also use. And I'll show that to you right now with a actually real data series I took. And you can find tutorials on this on YouTube. Links will be coming up on a later slide. This is what the software looks like as you open it. I've prepared this for you already. This is a run of the software. And it first has to load the images I took. I prepared a few of those. And these have to be in the astronomical format called fit. But the software you use to store these images will always save it in that format. So that's the format you should choose. So I'll open this. I only need one image. It tells me that there are 143 images in that file. And you should always click an option that is called use virtual stack. Because otherwise, if you don't do that, your computing power will be overused. We'll be seeing a few windows opening up. And that gives you the actual image. You can see several stars. So I chose the resolution. I chose the zoom. So the next question is, what is the actual star I'm looking for? And that's the one. I'm clicking on that. And you see this kind of target icon here. At the end, the data within that circle are what will be evaluated. And everything else will become the background signal that will be subtracted, as it were. And to do the averaging process, now I'm picking the analyze item and choose the profile item. And with that profile, the software will tell me the optimal radius for distinguishing the actual data from the background. Close this. And now I need to do this relative photometry, meaning click on other stars for comparison to detect the actual transit. So I'm using the symbol with this here. And I've reset a few things. You can learn from the tutorials what all these settings mean. So I'm going to now click on place apertures. Purchase. And of course, and first I need to click on the star. And then click on a similar star for comparison. And I'll pick these. And this one too, maybe that's the one that's a bit too bright. And I can see the brightnesses. And maybe I can deselect one of them too. And now I need to click the right mouse key. And that will open several windows at once. And the processing will then take a bit. And I can watch it as it happens. And I will be explaining what the tutorials explain it as well. And there are forums that you can use for understanding the software too. So I selected for all these plot settings, I can make several settings. I can say when the transit started. And I just have sorted out all the data is in the file already. So I know which time, which image belongs to which time. I can tell when the transit started. That's just after 11 PM. And it's then transformed into a decimal representation. Also the time it ended. It was a 77 minute event. And the exoplanet is called 3B. That's another hot Jupiter. And you can see these lines are getting drawn. And the blue dots are data coming from the actual images. And I already told this after when the transit started, which I could take from the exoplanet transit database. That's this red line. So that's the theoretical, the predicted start of the transit and the predicted end of it. And I can also tell it which of the data items to choose for the transit. Because I'm telling it not to use the actual transit because certain fits of trends are also applied and taken out. So I'm clicking with the left button on one of these items here. And I can mark the dots to use. So on the ones on the left here. And we'll be doing it the very same on the other side. And what I then can say with these fit parameters, the software is built in such a way that it will automatically apply transit profiles to the measurements. And for that, I have to know what the over two period of the planet is, which I took from the exoplanet database. And in that case, it's 1.3 days, which I've pasted into this dialogue here. And what you also have to specify is what the physical properties of this planet are. Or to actually detect the physical properties, sorry, I can tell it the size of the star, which is also found on the website. So this circle here representing hot Jupiter, that's 0.924 solar masses, something I also input there. And then there are several data models that I used to actually determine the size of the star. And that can be reduced. And you can see the transit very well now. And you can see that the observed data actually fits the prediction, brightness reduction of 2% to 3%. And you can see that the measured items have some variance, some statistical variations. And you can see that there are some items now coming up where the brightness is increasing again. And as you look down, you can actually see various parameters on the planets that are determined. So you can see various fits that are applied. You can actually, you have to learn what that means by looking at the documentation or tutorials. And the orbital data determined from that fit are now becoming visible. So distance to the star. So 6.4 star diameters is the minimum distance of that planet to compare that to the solar system. And what you see also is the size of the planet. And you see 1.4 Jupiter diameters is what was calculated from my measurements. And what you also see is the orbital inclination, the angle at which we are looking at the orbit. It's not 90 degrees. So we're not looking at the orbit from the top. It's about 87 degrees, actually. And actually, with the mark coming from the control room, the screen share has stopped. No, actually, I stopped the screen share. But I finished with the screen share. So you see that the actual view would have been 1.3 Jupiter radii, and the inclination would have been 81.4 degrees. So actually, my measurements are quite close to the expected values. So going back to my slides, even as an amateur, I can actually find physical properties of exoplanets. And if you want to take a screenshot, I'm happy to. Thank you. First enthusiastic feedback, more than 300 people in the chat. Please imagine a room of 300 people who are giving you resounding applause. I'm trying to communicate this. Sadly, virtually, this is not very easy to convey, but imagine it should be possible to imagine this. So huge thank you. And we have many questions. Did the end still was the end in the video? The control room knows, and we'll see. I hope so. But the slides, sadly, were not visible in the end. I've just been told. Anyway, the first question is one you asked at the very end here. Did actually Amateurs discover a new exoplanet? I just can't just say that, right? Yes, Amateurs did build a searching system using several telescopes. And yes, they did find an exoplanet. And you yourself, 14 years ago, no, I didn't discover that one. I only confirm known transits. And that is important, too, because the professional astronomers cannot track all those exoplanets that they discovered. Because the time allotted from these professional telescopes is very limited for all these various research programs that there are. And you can check when the transit happened and confirm whether it happened at the expected time. And you can actually see things changing, shifting slightly. Transits happening 10 minutes earlier than predicted from the first observations that the professionals did. So that way, you can actually correct the orbital data. But if you had been the very first one. If you, by chance, would have been the first one to find that, then, but there's really a huge chance there. So a very small chance of getting that. So taking a photo at exactly the right time and really look at. So if I do a lot of photos here and analyze all the stars in that one for exoplanets. But the likelihood is very, very small for that. Because you'd only see transits coming from the very, very large transits. Otherwise, I couldn't observe this as an amateur. But I mean, the phenomenon is there. As unlikely as winning the lottery is, there's always someone who wins the lottery. So I think it would be possible. Another question. Can you differentiate between a planet being very large and close to its star or very large and further than that? Or further away? So can you differentiate that? Yes, because of the duration of the transit. Because the further the planet is away, the longer it takes for the next transit to happen. So if you, for example, look at the Earth, you could only observe a transit of Earth just once every year. So these search programs, most of what they found, would be planets very close to the star. So that's also the question. How often are solar systems like ours? And this is very hard to observe. Because Neptune, the orbit time, is something like, I think, 800 years. So we couldn't really observe that. So would it be helpful for exoplanet hunters if they would be distributed all over Earth and looking at a transit? So with the idea that you would get a higher resolution of that one, similar with what happens with radio telescope. And so with those visual observations, you need to connect them and analyze them absolutely live in these very difficult methods used by professional professionals here. And that's not even possible for amateurs here. So this is way too complex. So next question. How do you make a difference between a transit and a star that just changes its brightness on its own? So that's an important question. So of the variable stars, we know their profiles for quite a long time. So the brightness profiles of these transits are very typical, very specific. And in addition, you just can't observe one planet for just one planet. So you would have to observe several transits to be sure that you have observed an actual transit here. So it would be at least one year of this specific planet for you can be sure here. So this is a very, very exciting topic here. So thank you for this bridge to the philosophical question. So you, as a exoplanet hunter, how do you feel the, sorry, the anthroposophical humiliation? So feeling just as a small, tiny human in a huge universe. So first of all, even 30 years ago, we weren't sure whether there are any exoplanets at all. So 1992 was the first time to actually observe that. And what we know now that is that having a planet is not that unique in the versatures. Having a planetary system is not that one. So well, so far we know that we are still the only known inhabited planet. But so the whole thing is very exciting. So we learn more and more about planets everywhere about solar systems everywhere. And so all of the research, all of these observations there even increased this feeling of humanity being something special, of Earth being something special. So also, I mean, the tech always improves there. China, for example, has a telescope with a 39-metre mirror. So the absorption of observations we can make are getting better and better. So maybe we get a chance to find other inhabited planets or planets live on that, looking at chlorophyll, looking at the spectrum of these planets. So the professionals will look at places where all these amateurs would be guessing, that there might be something. So there have been lots and lots of stars have been looked at with the Goldilocks zone. So we know lots of planets in that zone. And this would be candidates for actually searching for life. But on the other hand, we wouldn't even know how life could look like on other planets. We have a sample size of one here. So we have a very, very limited point of view. So we got the first batch of questions from the batch. So we got some answers for that. Those, I'll do a short bridge to the next one. But I am inviting everyone here to go to the question and answers room here so we can extend the discussion. So thank you very much.