How the Kepler Telescope Works

be honest on just the Kepler Telescope. And the reason why I'm doing this in the first place is because in May, researchers with the Kepler Mission at NASA held a press conference in which they announced the largest number of exo planets verified ever at a single event, and that was one thousand, two eight four verified exo planets. Previously, from two thousand nine, up to that point, the mission had identified and verified four planets, so this announcement was

more than doubling the number of exo planets verified. That's incredible. So an exo planet, just in case you don't know, is of course a planet that is orbiting another star, not the Sun, so it's planets in other star systems, solar systems that are light years away from US, and it was a really cool thing to hear about all

these different exo planets that had just been verified. Um, what I thought was hilarious was leading up to this announcement you had several news outlets that were, uh, guessing what was going to happen. It was just it was just complete throw stuff against the wall and see what sticks. And there were quite a few that had guessed that NASA was going to announce that the Kepler mission had

somehow discovered alien life. Now, once you hear how the Kepler telescope works and what it's meant to do, you will understand that's really not in the purview of the Kepler telescope. It is looking for planets that could potentially support life, but it doesn't have the capacity to actually detect life on those other planets unless someone sent aliens to Earth and they started messing with the Kepler telescope, in which case you could say it discovered life, but

not through its primary mission. That did not happen. As far as I know, no aliens have been messing with the Kepler telescope. So let's talk about how this telescope works and this new verification method that the research team used to identify so many exo planets. What was it that sped up the process so dramatically as to more than double the number of confirmed exoplanets. Plus I'll talk a little bit about the new candidates for earthlike plants

that might be capable of supporting life. So first, let's go way back. The Kepler telescope is named after Johann Kepler, and astronomer of the late sixteenth and early seventeen centuries. That's not the original name of the telescope, by the way, but more on that in a little bit. So Johann Kepler is most famous for discovering the three major laws of planetary motion, although at the time no one called them laws. It would take Newton and Newton's observations before

that really started to become a thing. But law number one is that the planets move in elliptical orbits around the Sun. Law number two, the time it takes to traverse any arc of a planetary orbit is proportional to the area of the sector between the central body, for example, the Sun and the arc. So you can think of the two points along the arc the starting point in the endpoint of the arc as your your barriers on

one side, the Sun being the third point. And what's not exactly a triangle because you have a herv line on a straight line on the ark side, the area within that that is proportional to the time it takes to traverse that arc. Essentially, what that tells you is that the further out you are from a star, the slower your orbit is going to be, and the closer and you are to a star, the faster your orbit

is going to be. Also, there's a relationship between the square of a planet's periodic time and the cube of the radius of its orbit, which is also known as the harmonic law. That's law number three. Now, the Kepler mission all started out with a question, which was how frequent are other Earth sized planets in our galaxy the Milky Way? How common is that? Is the Earth a strange outlier that is one in a billion or one in ten billion, or or even more rare than that.

We had no way of knowing. Now that particular space based tell us cope, the Kepler telescope tries to answer this question by looking for planets using what is called the transit method. Now, this method was proposed a few times leading up to nineteen one, when Frank Rosenblatt really went strong with the idea. He suggested the transit method

for detecting satellites orbiting other stars. And technically, the way this works is that you look at a star and you measure the amount of light coming to Earth from that star, and you look for any variations and that

any dimming of that light. Now, if a planet were to pass between that star and the Earth, you would expect the light from that star to dim a tiny amount, and that if you were able to detect that difference, and you were able to observe that this happens regularly over the course of a given amount of time, you could come to the conclusion that what you have seen is in fact a planet passing between its host star

and the Earth. This is what we call transit when we see from our perspective a planet crossing its star. Now we're looking at the planet making its progress across its star in the course of that planet's year. So if it's close to the same distance from its host star as the Earth is from the Sun, you have to wait a long time to verify that that in fact is what you saw, especially if you want to truly verify it and and get a few periodic uh

instances of that dimming. And if it has, if it's a star that has multiple planets alone that same orbital plane, then that's going to cause some confusion too. But by by seeing the amount of light that has been dimmed, and by detecting how long it takes the this dimming to change back to normal, you can start to make some conclusions like how big the planet must be, how quickly it travels tells you a bit about its orbit. It tells you also by that orbit, how close it

is to its home star. As long as you know information about the home star, then you can start to make guesses as to how hot the planet is or how cold it might be. And this is how you start to draw some conclusions about the nature of that planet.

Ultimately looking for planets that are similar to size, uh similar to Earth size, I should say, and similar to distance from its host star as the Earth is to the Sun. The reason for that is we know that if the planet is about two times the size of Earth or smaller, it's probably going to have gravity that is uh amenable to life as we know it. It's gonna probably be a rocky planet as opposed to a

gas giant. That's also important. It's probably gonna be at a temperature that will allow liquid water to be on the planet. And since life as we know it depends upon the presence of liquid water, that's what we're looking for. Keeping in mind that there is the possibility there could be life out there in the galaxy that doesn't depend upon liquid water. But we have a sample size of one planet with life on it, so we have to draw our conclusions based upon the limited information we have.

So assuming that water is in fact necessary for life, we need to find other planets that could potentially have water on them. So that's kind of guiding the principles behind looking for planets through the transit method. But it's really really hard to do. Now, let's get back to to Frank Rosenblatt for a second. He wasn't just famous for suggesting this astronomical approach. He wasn't known as a

great astronomer. He was actually better known as a leading expert in the field of artificial intelligence, particularly in the areas of recognizing visual patterns and speech recognition, So that was his specific forte. He was really working with computers so that they could recognize visual patterns, they could recognize when you what you are saying when you talk to them.

And these are fields that today are really coming into their own with stuff like Google's Deep Dream, where it starts to recognize patterns even if patterns aren't really in the picture. It really enhances that and looks for patterns in in the in ways that are really interesting and trippy. And speech recognition, which we're all using to some degree these days, um often with the personal digital assistance that are popping up all over the place. Now here's the problem.

Rosenblatt's suggested approach as not practical at the time in the early seventies. We just lacked the technological sophistication necessary to detect and analyze such a very tiny change in

a star's brightness. If we were looking for an Earth sized planet at a distance similar to what Earth is from our sun, we're talking about a reduction of one ten the brightness of a star, and that dimming lasts between two and sixteen hours, so that's not a lot of time, and it's certainly not a big difference in brightness. You have to have a very sensitive instrument in order to be able to pick that up, and that just

didn't exist in nineteen seventy one. Now, we also have to keep in mind that stars tend to be much much bigger than planets. For example, the Sun's diameter is a hundred and nine times greater than the Earth's diameter, and because of that, that's why, with the distances involved and the size is involved, it's such a tiny change in the brightness of the star. However, NASA began to explore how they might be able to use the transit

method to detect exoplanets in a practical way. Back in they were essentially laying out the requirements necessary to detect planets with a reasonable amount of confidence. A conference that they held on High precision photometry acted as the launching ground figuratively speaking for discussions about a space based telescope designed to detect a transitting planet. The idea being that in space there would be less uh noise in the signal to noise ratio you could get it outside the atmosphere.

The effects of the atmosphere would not be an impediment to a space telescope, and it would be more likely to pick up something as tiny as this change in brightness.

So in two NASA proposed new missions to look into the possibility of life in our galaxy, and the first concept that came up with was called the Frequency of Earth Size Inner Planets or FRESIP f R E s I P. But that proposal was rejected largely because there was doubt at the time that our technological sophistication had actually reached a level sufficient to detect any transitting planets

of Earth like size. So, in other words, if we had gone forward with it, we would have built a tool that was not up to the task of actually doing what was supposed to do, and we would have wasted millions of dollars in the process, not something NASA

could easily afford to do now. Two years later, in a team proposed FRESIP again with a space based telescope in lagrange orbit, but again a committee determined the price would be similar to that of the Hubble, which was incredibly expensive and also had been a black eye on NASA because when the Hubble launched, it launched with a defect that later had to be corrected in space. A team had to be sent up to make some tweaks

to the Hubble. Space telescope so that it would be perform more closely to what it was supposed to do because one of its mirrors was not right at any rate, they didn't want to involve a you know, I didn't want to invest in a big time project that was unproven um, especially after the Hubble issue, so they rejected

the proposal. By the way, in case you're wondering what a lagrange orbit is, that refers to five specific orbits around the Sun. In two of the orbits L one and L two, a spacecraft would orbit the Sun either

just inside the Earth's orbit or just outside the Earth's orbit. So, in other words, if you were to look top down, you would have a spacecraft that would be just inside of the Earth orbit moving at the same speed, or one just outside the Earth orbit, moving at the same speed as the Earth around the Sun. And you might think, well, how is that possible? Because earlier you mentioned if you're closer in to the star you move faster, and if you're further out from the star, you move slower. So

how would spacecraft keep up with the Earth. The answer is gravity. Earth's gravitational pull would be enough to hold the spacecraft in that orbit. That Lagrange orbit, and you could do this. You might want to do this for lots of reasons. If it's on the inside of the Earth's orbit, you could do it to study the Sun as it faces Earth. If you did outside the RS orbit, you could look away from the Earth out into the outer Solar System and not have the Earth in the

way when you're studying those planets. Um There's also another one. L three Lagrange orbit, is on the opposite side of the Sun from the Earth, essentially following more or less the same orbital path as the Earth. This is a great way to see the far side of the Sun while the Earth is in its normal location. So if you wanted to study the far side of the Sun, you could do that and look for solar activity. Then there's L four and L five, which are sixty degrees

separated either before or after the Earth's orbit. But enough about that. Those are the Lagrange orbits, the idea being that when you play something in there, it tends to be pretty stable. But NASA had determined that putting something into one of those orbits would be really expensive. Uh So eventually they came to the conclusion that perhaps they would want to just put the Kepler telescope in an orbit around the Sun in its own orbit, not a

lagrange orbit. Meanwhile, engineers started to experiment with charge coupled device sensors to see if they could be made to detect tiny changes in light, and lab experiments with c c ds proved that they were a pretty good candidate for this. So let's talk about c c ds for

a second. They're designed to move an electrical charge, typically in a way that allows a device to convert the electrical charge into something else, like a digital value, and C c D image sensors are important in digital imaging, particularly for highly sensitive imaging, such as with very low levels of light. Now you can find digital cameras with c c ds, but many also use or rather instead they'll use active pixel sensors or the seamost c MOSS

c m O S sensors. And it used to be that there was a noticeable gap in quality, that c c D s were demonstrably much higher quality than c MOSS sensors. These days, that gap is much more narrow, it's not as uh as blatant as it used to be. So we've seen the technology of one start to catch up to the technology of the other. Within the c c D, you have millions of tiny light sensitive squares called photo sites, and each photosite corresponds to an individual

pixel in the final image. It uses the photoelectric effect. It turned photons into electrons. That's actually an oversimplification. It really uses photons to energize electrons push them into higher

energy bands, and that is the key to how CCTs work. Essentially, photons raise the energy level of electrons from low energy valence bands to high energy conduction bands, and each photo site has a positively charged capacitor when the photon converts that electron when it adds that energy to an electron, the electron is then attracted to the positively charged capacitor and the number of photons that penetrated the CCD affects the voltage that this creates, and that voltage is then

converted into a digital signal. The whole array is actually cooled through a series of heat pipes that run through an external radiator, so they're actually rading the heat directly out into space. So as this generates heat, they just vent that off into space and it keeps the whole thing cool enough for it to operate without overheating and

causing any problems. Now in six two years later. So remember this was first proposed in ninety two and rejected, ninety four, rejected, nine proposed again, and at this point they started to make some changes. One of those was decided to put the telescope into a solar orbit rather than a lagrange orbit. It was also the point where they renamed the project Kepler, after the astronomer we talked

about earlier. But this proposal was also rejected. This time is because no one at that point had proven that a telescope could simultaneously observe thousands of stars. One of the big selling points of the Kepler telescope is that has a very wide field of view and can keep an eye on a hundred thousand stars simultaneously. But NASA budget oh overseers were saying, no one's proved that you could do this yet, so researchers went to work on

a prototype photometer to prove it could be done. In nine seven, they had finished building that prototype photometer and in they demonstrated that it could observe six thousand stars in a single field of view and generate data that could then be analyzed. The results of this project were published in a paper in n SO nine, seven years after the initial proposal. It's proposed yet again, and it

got rejected yet again. So why was it rejected this time? Well, it was rejected on the grounds that there was no evidence the photometer would be precise enough to find Earth sized planets that could also operate in orbit in the presence of noise. So now the argument was, all right, you've shown that it's precise enough to detect a planet, but maybe un Earth sized one, because those are particularly small,

they're not the size of a gas giant. Uh So we need to prove that, and we aren't sure that if you're in space, you will be able to differentiate a planet passing between Earth and its host star or just some random piece of space debris that happens to pass between a star and Earth, until you can prove

that we're not giving you any moneys. So the engineers built another test bed and they proved that the Kepler telescope could operate satisfactorily even within noise, that their analysis would be able to differentiate between false positives and the real thing. So two thousand rolls around and the Kepler gets proposed one more time, and this time it's selected as one of three proposals out of a total of twenty six to compete ver NASSA approval, So it then

goes on to compete with the other two projects. And essentially this is the way NASA works. They have teams proposed different UH potential missions and then they whittle that down to a group of finalists and then they say fight it out, convince us to fund your project. And sometimes only one project gets funded, and that was the case for Kepler, and in two thousand one it won the right to be funded. It became Discovery Mission number ten.

So the Kepler Telescope is a discovery spacecraft by that definition. The actual work on the mission began in two thousand two, and that started with orders placed for the detectors for those c c D s and the telescope was completed and launched UH and it was launched on March six, two thousand nine, and went into space around its solar orbit. So here's some stats about the Kepler tell lescope and the Kepler spacecraft. The diameter of the photometer is just

shy of a meter. It's point nine five meters in diameter. Which is about three ft. The camera has a ninety five megapixel array. So your typical smartphone has an eight to maybe thirteen megapixel camera on it. This one is a nine five megapixel camera, and it can continuously monitor the brightness of more than one hundred thousand stars simultaneously. The field of view is thirty three thousand times greater than that of the Hubble Space telescope, so it's looking

at a pretty wide range of space. Keep in mind, the Milky Way galaxy has a hundred billion stars in it, so a hundred thousand is nothing. It's it's a tiny little drop in an enormous bucket. Let's talk about the spacecraft though. The Kepler spacecraft is two point seven meters

in diameter, that's about nine ft. It's four point seven meters tall, that's about fifteen point three feet, and it weighed one thousand, fifty two point four kilograms or two thousand, three hundred twenty point one pounds at the time of launch. Why is it just the time of launch, Well, part of that weight was taken up by fuel hydrazine propellant, which it has used some of since it was launched. There was eleven point seven kilograms of fuel at that point,

so that makes a difference. Also the fact that it's in space hard to weigh things in space. You can talk about mass, but weight not as relevant. It generates electricity with one hundred nine point eight square feet or about ten point two square meters of solar panels. Now those solar panels can provide one thousand, one hundred watts

of electrical current. The space cars also has a twenty amp hour lithium ion battery that's a rechargeable battery, so when it's generating excess electri city, it charges the battery and UH and everything can continue to be powered. Once it launched, it became part of the Exoplanet Exploration Program Office, part of the Jet Propulsion Laboratory, so it's been shifted from one group in NASA to another one to actually

manage the mission. So basically, the way the Kepler works, it has more than a hundred thousand stars in its view and it can detect these very tiny fluctuations in the light from those stars. Typically, up until recently, we just had to keep those stars under observation for a really long time to see if that fluctuation would repeat at regular intervals, and it wasn't just something passing between us and the star. And that was how we would go from a signal that was a potential planet to

a verified planet. It also explains why, over the course of several years, we were only able to verify n exo planets with a whole bunch of candidates that maybe or exo planets, but we're not sure, so we can't call them that. But then you had this May tenth announcement of one thousand two four exo planets. So what changed?

How could we potentially do that? They actually the researchers had analyzed four thousand, three hundred two potential signals, these candidate planets, and out of those four thousand, three hundred two they decided that one thousand two Night four should be verified as actual exo planets because they had a greater than certainty that they were in fact planets and not some anomaly or impostor as they called them. This is pretty phenomenal, right, this is Night the fact that

they could more than double them. Uh. They also had said that there were another one thousand, three hundred seven signals that have a better than fifty chance of actually being a planet, but those would require more research and observation before ASSA would go so far as to say, yeah, here are some. These will also join the list of verified planets and are very high threshold. To call a signal a planet, it had to be greater than certainty.

So that's pretty incredible, much higher standards than I have. I'd be cool with now. As a throwback, the first Earth size planet that Kepler telescope found in a potential habitable zone, also known as the Goldilocks zone, was Kepler one eight six f UH, and the Goldilocks zone that's what we think. That's the band of orbits we think would be UH would be amenable for life to exist,

for water to exist in liquid form. The Goldilocks zone is dependent upon lots of stuff like the not just how close you are to the host star, but how old is that host star. You know, if it's an older star that's burned out a lot of its energy, it's a cooler star. So you have to be closer to the to the star in order to get enough energy to support life as we know it. So it's depend upon a lot of factors. In the case of Kepler one six f the host star is older than

our son, it is more red, and it's cooler. And this also means that if there is in fact life on Kepler six F, it probably looks different from life on our planet. It's receiving a lot of red wavelength photons coming in, which could mean that the plants themselves might look very different. They might be big red plants all over F or it could be a barren waste land. We don't know. We have no way of telling yet.

We can just make some guesses based upon the age of the star, the size of the star, the distance that we estimate the planet is from that star, those sort of things. Those are the kind of things that we can kind of start to draw some basic guesses around. But there's still guesses until we can develop some other

means of really looking at these distant planets. Now, out of all the one thousand two four planets announced by this research team, nine of those are considered potentially habitable, meaning that they are relatively the same size as Earth or no greater than two times the size of Earth,

and within their host stars Goldilocks zone. Now, what's really cool is to look at how they determined this, Like what was the way that they verified these planets and they used a probabilistic approach, meaning what is the probability that any given signal is in fact a planet? Essentially,

they were looking at two main factors. How much does a single transit signal resemble what we would expect from a transitting planet, so does it look like what a planet would look like when passing in front of a star. And then also what is the likelihood that that particular signal could have been caused by an impostor? And you take these two ideas into account and you try to figure out what is the likelihood that we have and

a real legitimate hit here. One of the people associated with this mission, Timothy Morton, who's an associate research scholar at Princeton University, calculate the probability that any given transit signal is actually a planet. Uh by using this and and essentially you've got numbers between zero and one, and only the results that were as close to one as possible better than in fact were kept and verified as a planet. Now, the big advantage of this approach is

that it can be applied to many signals simultaneously. Instead of having to continuously review the data of a single signal and look for those replicable results. You could take this approach and apply it across multiple planets all at the same time and see which one's come out at greater than certainty. Or as Morton said, if you drop a few large crumbs on the floor, you can pick those up one by one, but if you spell a whole bag of tiny crumbs, you're gonna need a broom.

And the statistical analysis approach is their broom. So we've got out of all the different planets found, about five fifty of the one hundred and eighty four were announced on May, about five fifty of them might be rocky planets like Earth, and out of those only nine are considered to occupy the habitable zone. And he might think, well, that's a really small number. Nine how many were there before? The answer was twelve, So there were a dozen discovered

before this announcement. Nine more added to it, for a total of twenty one. There are several others that are possible candidates for rocky like planets that could be in the Goldilocks zone, but they don't meet that criteria of yet, so they have not been verified. There has just been twenty one verified planets that are of rocky most likely anyway rocky consistency and in that Goldilocks zone. So this is also we gotta remember based on that assumption that

liquid water is ne necessary prerequisite. If it's not, then obviously we could be looking at lots of different plants that could potentially support life, might not be life that we would recognize. However, so the kepler, as awesome as it is, cannot detect all the exoplanets orbiting stars. If the planet orbit isn't at the right angle from our perspective,

from the kepler's perspective, it won't detect any dimming. In other words, if there's a planet crossing that star, but it's at an angle that does not go in front of the star from our perspective, the star doesn't dim, we don't see any change in that, and the kepler can't detect it. So how many plants are actually passing at the correct angle for kepler to detect them well.

The probability of such a thing is determined by the diameter of the star divided by the diameter of the orbit, which for a planet the size of Earth orbiting a star similar to the Sun eventually gives you a point five pc chance of detecting that signal. Being at the right angle to detect that signal point five half a percent chance of detecting the signal in the first place.

Bigger plants have a better probability because they are more likely to at least pass over a star partially to you know, they have fewer angles where you won't see anything at all, So a much bigger planet like something like Jupiter could be closer to a ten pc chance. So it's entirely possible, and even probable in fact, that what we have detected is just a tiny fraction of what is actually out there, even just with the one hundred thousand or so stars we've looked at, more like

a hundred fifty thousand. But even though we've looked at a hundred fifty thousand stars and we've detected so many of these exoplanets so far, there could be a lot more that we just can't see because of the angle. And then you take into account we're looking at a hundred thousand out of a hundred billion, and the mind really starts to boggle. We realize that the frequency of planets around other stars is much greater than we anticipated, and that we even maybe looking at more planets in

the Milky Way than there are stars. So if you have a hundred billion stars and there's more in a hundred billion planets, you think, wow, the odds of you know, the chances that there is another planet within our galaxy that could potentially support life are pretty good. It may

not be anywhere close to us. It may be on the other side of the Milky Way galaxy from where we are, but the chances are pretty good that there's at least some other planets within our own galaxy, let alone the universe, which is filled with billions of galaxies. And suddenly you think, there's no way that we're the only life forms in the universe. That's just not statistically plausible. Is it possible? Well, I mean, technically I guess so,

but it's certainly not not plausible. It's more likely that there's life on lots of other planets. Whether it's evolved life that is intelligent, that's another matter. Whether it's life that is anywhere remotely close to us where we would ever have an opportunity to discover it through communication, that's

highly debatable. Uh, it's quite possible that any any life that's that advanced is so far away from us that we haven't had enough time to pass for any communication generated by that civilization to get to us, because I remember that stuff has to travel at the speed of light. That's as fast as you can go barring some huge change in physics, and so if you are thousands of light years away, it's going to take thousands of years from the generation of a signal for it to get

to its destination. And by then the milk has gone bad, and that shopping list that the aliens gave us is not really going to do anyone any good. The party is over, but still it's really exciting to think about what the Kepler telescope has done. Now, keep in mind this is very different from other types of telescopes. There

are other ways of detecting the potential for exoplanets. One of those is to look at stars and look to see if they are moving at all, like if there's a little jiggle, which could indicate that there is a gravitational pull upon that star, and that in turn can indicate that there is a planet in orbit around the start and the planet's gravitational pull in the star is causing it to move just a little bit, and that

we can detect that. That's another way of detecting at least the potential of an exo planet in that star's orbit,

but it's different obviously from the transit method. And then there are ways where we can look at planets to try and determine what are they made of, and we usually use a spectroscopy for that, where we we take the light reflected off of a planet and we analyze that light and we break it down into the various wavelengths, and then we start to make very educated guesses about the types of elements that are present on that particular planet.

Even so, this is still largely working from very educated guesses, uh so educated that you could argue they are they are as good as fact, at least in some cases. But you have to keep in mind there's still there's still a tiny room for error. So that about wraps it up for the Kepler Telescope. It has served us well. It's primary mission is over um. We will continue to look at the data from the Kepler Telescope for many more years, but the actual data gathering portion of the

Kepler's life is over. There are other telescopes that we're planning on launching that will continue this work. It will either be looking for similar planets to what Kepler was looking for or different style planets. But we're just getting started, and the hope is that we will eventually be able to draw some very firm conclusions about the presence of life within our galaxy, and this could just be the

first step toward that. So while a lot of those news outlets were perhaps being a bit optimistic about the announcement of the discovery of alien life, it is true that this is a step toward making such a discovery, and it may be many, many, many more decades before we're able to say, yes, we've detected the presence of life on another planet. But it's through work like the Kepler mission that will get there. So this is a really cool science and technology story, and I just wanted

to touch on that because I loved that announcement. I actually listened to it live while they were talking about and it was just really cool to hear a group of engineers and scientists talk about their life's work with such passion. So, guys, if you have suggestions for future episodes of Tech Stuff or you've got questions or comments or anything like that. Maybe you've got a suggest stution for a future guest host or an interview subject. Let

me know. Send me an email. The address is tech stuff at how stuff works dot com, or drop me a line on Facebook or Twitter. The handle at both of those is tech stuff hs W and I will talk to you again really soon. For more on this and bousands of other topics. Is it how stuff works dot com