A conversation with Jon Jenkins, the science lead for the Science Processing Operations Center at NASA’s Ames Research Center in Silicon Valley. At the SPOC, the raw data from NASA’s newly launched planet hunter, TESS, is prepared for scientists to analyze.

Transcript

Host (Matthew Buffington): You’re listening to NASA in Silicon Valley episode 91. This week we’ve brought on Jon Jenkins. Jon’s the science lead for the Science Processing Operations Center, that’s where the raw data from NASA’s newly-launched planet-hunter, TESS, goes to be cleaned up for science. In case you forgot, TESS or the Transiting Exoplanet Survey Satellite just recently launched back in April. “From pixels to planets,” that’s the catch-phrase. Jon’s had a long history at Ames working on finding planets in other solar systems even when this topic was fringe science. He’s done his part to bring us into the golden age of exoplanets.

So, without further ado, let’s jump right into our chat with Jon Jenkins.

Music

Host: Tell us a little bit about yourself. We always start the podcast off in the same way. How did you get to NASA, how did you end up in the Bay Area, in Silicon Valley. Tell us about you.

Jon Jenkins: Okay. I was a graduate student at Georgia Tech, and I was a radio scientist in electrical engineering who was studying the atmosphere of Venus using Pioneer Venus Orbiter at the time, back in the late ’80s, early ’90s. And I graduated and needed something to do, and so I came to Ames with a Pioneer Venus guest investigator program grant that allowed me to continue my work on Venus.

And that’s how I met Bill Borucki, the PI [principal investigator] for Kepler, and that is that he was studying lightning in the atmosphere of Venus. He’s not shy of controversy, right? Lightning on Venus was actually controversial at the time. But he was just such an enthusiastic person and a great scientist, so after a few years I got involved with a group of international astronomers who wanted to find exoplanets, and the exoplanets they wanted to find —

Host: Wasn’t that kind of, I don’t know, controversial, but I remember reading there was a lot of false positives, or it wasn’t a guaranteed thing that there was exoplanets?

Jon Jenkins: No, not at all. In fact, when I was a graduate student going to meetings, professional meetings, every once in a while there would be a claim that somebody had found an exoplanet, and it would get retracted. There was a large amount of concern about the fact that if we were going to do a mission like Kepler, that we had to be very certain the discoveries we made were real planets. We did not want to get any egg on our faces and have to retract a claim. So, that was very important.

Alison Hawkes:What year was this? When did you start working on the exoplanet work?

Jon Jenkins: Circa 1994. So, this was before the first exoplanet around a normal star was discovered, 51-Peg-B. That didn’t happen until 1995. But a group of astronomers, one of them Lawrence Doyle, who was here, who also worked for SETI Institute at the time, wanted to look at, for circumbinary exoplanets, for planets orbiting binary stars —

Host: I was going to say,”circumbinary?”

Jon Jenkins: Yeah, circumbinary. This is when you have a double-star system.

Host:Oh, the Star Wars reference, Tatooine.

Jon Jenkins: That’s right, we were looking for Tatooines. And we had picked this one star system that was the smallest known set of stars in an eclipsing binary. So, they were eclipsing each other. The orbits were nearly edge-on from our point of view, which meant that if there were planets there, they’d almost be guaranteed to be crossing in front of the stars as they orbited.

Because the stars were really small, about a quarter of the size and mass of the Sun, that meant that we had a really good chance of finding planets as small as, say, twice Earth.

Host:Help me out on this one, because if you have two stars that are binary, are you thinking they’re so close that they’re almost like the touching and collapsing, or it’s like they’re circling each other, but then they also have planets circling them? Or are they circulating both planets? Or, how does —

Jon Jenkins: Right, so this is a very close circumbinary system, but they’re not contact-binary. So, they’re not touching each other because the stars are small enough, but the orbital period is 1.26 days. This is really short. And the neat thing about this is because they’re smaller and cooler than the Sun, if you put Earth in the system and said, at what orbital period would we have the same amount of flux or radiation from this system from the stars, would we be happy? Could we grow plants, would liquid water be on the surface? We would be in a 17-day period orbit. So, our year would only be 17 days for us to be as warm as we are orbiting the Sun in a 365-day period orbit.

Host:But I’m thinking, but if you’re orbiting a star that is doing a do-si-do with another star, I’m wondering, are you going to slingshot into that star, are you going to get too close to that other one? You’re going to have a bad time, or am I just completely nuts and crazy on this?

Jon Jenkins: Right. Well, no. It turns out that if you are closer than about five times the distance, five times the distance between the stars, then bad things can happen. So, but if you’re outside of that range — so, for this particular star system, once you’re outside of five times 1.26 days for the orbital period — then you were okay. So, we could have habitable-zone planets there.

Alison Hawkes:This was before a single planet had been found?

Jon Jenkins: That’s right.

Alison Hawkes: So, I want to just take you back for a second, because this was crazy-town back then on exoplanets. There was nobody, there was very few people working on exoplanets. And did you, so did you feel like you were on the fringe of science back then, and what was it like working on exoplanets when we hadn’t even found a single one?

Jon Jenkins: Well, at the time, almost nobody thought it could be done. In fact, there was a lot of skepticism, even though we were looking at a very small star system and it was quite favorable, because the stars were small enough that we could hope to reach the precision from the ground to be able to detect planets as small as twice Earth-size from the ground. We got lots of observations over six years, but we never got any funding from NASA or NSF [National Science Foundation]. And some of the reasons we got back in the debriefs were, “Planets don’t form around M-stars,” “Planets don’t form around binary systems.” And guess what, the people who had predicted that was the case — because nobody knew at the time — were wrong. So, Kepler-16b was discovered in 2011 and was the first case where we discovered a Saturn-sized planet orbiting a pair of stars.

Now, it turns out that we should not have expected to find planets around it because it’s a close binary system. So, when we find planets in circumbinary orbits about binary star systems, they’re much more widely spaced, it turns out.

Alison Hawkes:So, how did you pivot from working on this binary star system, and trying to find a planet from the ground to the idea of, we’re going to go off and put a telescope into space and see if we can see it there? How did that idea come to you guys?

Jon Jenkins: Well, Bill Borucki was working on this all the time. And so, I heard him give a lecture here at Ames about Kepler. It was like, “I’d love to work on that mission. That would be so exciting.” This is truly level-zero science. It’s not like working on Venus, which was really exciting to me at the time, but that was in essence learning more about Venus, about something we already knew a lot about, and here was a question we’d been pondering for thousands of years as humans — are we alone, are there other planets out there when you look up into the night sky? And I know I did this as a child, lying in the grass in the summertime, looking up, wondering are there planets orbiting those stars, and are there beings like me looking up into their night sky in my direction wondering the same thing?

At the time, when I was in grad school in the early ’90s, almost nobody believed it could be done, because the technology was not ready for it to happen. And there were a lot of — there was a series of key technical challenges that needed to be overcome for the community to recognize that this was a credible thing, and that we should be spending the funds to do this, because it’s a real investment by NASA to fly a Discovery-class mission like Kepler.

Host:And didn’t, as I understand it, Bill Borucki, he had pitched this a handful of times. There was like, “Let’s do this,” and they’re like, “No, go back, get more info, or modify it slightly.” It’s not like it was just thrown out there once and it’s like, “This is genius!” It was trial and error, editing, getting back to the drawing board, back and forth. That seems to be, that’s how things function. That’s how the system works.

Jon Jenkins: That’s right. Today, we think of Kepler as a fait accompli, of course it’s successful, of course we’ve detected 2,600 planets. Back then, almost nobody believed it could be done. In fact, I joined the team right after we’d gotten rejected by the first Discovery selection process in 1994. So, I joined the team in 1995, just shortly before I got married, and when Kepler was still called FRESIP, Frequency of Earth Sized Inner Planets. And the key questions at that time were, “Can we operate CCD sensors?” So, these are the image sensors in your cellphones — so, they’re in your digital cameras — can we operate these devices and reach a level of precision of 10 parts per million in six hours? That’s the precision we needed to achieve in order to be able to find Earth-sized planets transiting or crossing in front of Sun-sized stars.

Alison Hawkes:And you were working on this before, you know, digital cameras had not even really come out yet. This was not part of the consumer market even, right?

Jon Jenkins: It was certainly not widespread at the time. And in fact, what we needed is, we needed a breakthrough in the technology. We needed backside-illuminated CCDs to become commercially available. At the time, it was a research project. We had some research-grade backside CCDs on loan from JPL [NASA’s Jet Propulsion Laboratory]. We put these in the lab, we tested them, we demonstrated that we could reach this exquisite level of precision necessary to do the job.

But there were a whole series of questions like stellar variability, okay? So, what about star spots? If you look at the Sun, it has Sun spots, and they cross the face of the Sun. It takes about a couple weeks for them to do so. And if you measure the brightness of the Sun over that time period, the signatures of the spots in terms of the amount of light they block is five times deeper than that you would get from an Earth-sized planet.

Host:So you could end up mistaking a star spot for a planet?

Jon Jenkins: That’s right. So, first big major analytical question he had for me, besides doing this lab work, demonstrating that CCDs could reach this level of precision, was what about star spots, Jon? I want you to work with the SoHo [Solar and Heliospheric Observatory] data on the Sun to demonstrate whether or not we can actually make this happen. And it was a challenging task. It required sophisticated signal processing techniques.

Fortunately, what’s great about coming to NASA is you can bring a wide skillset with you. So, I had a background in digital signal processing, and so I took some of those tools and brought them to bear on this problem, and developed an algorithm, a computer software program that allows us to analyze the signals from stars like the Sun to characterize the observation noise from star spots so we can selectively ignore them and focus on the transits.

And we were talking about music earlier. It turns out that the algorithm is very much like a soundboard, or like a graphic equalizer on a stereo. So imagine, if you will, that the Sun has these star spots, and those are the bases. And you’ve got contra-bases that might be a little deeper, and you’ve got your baritones and your altos and your sopranos. When a planet crosses the face of a star in an orbit like Earth, it takes only 10 hours on average for it to cross the face of the star. But when you’re talking about the spots, it takes them a couple weeks. So, it’s kind of like listening for the piccolos. If you’re out on the marching band field and you’re listening to The Star-Spangled Banner, you have no problem whatsoever listening to the piccolos playing their solo during the interlude, no matter how loud those tubas are playing or the trombones are playing.

But the way we do it with software is we analyze the power in the noise as a function of frequency or pitch, and that allows us to then fold in and take account of the details of the noise process to formulate an optimal detector, so we can find these very small planets.

Alison Hawkes:I think it’s so fascinating that you really — the data comes down and it’s got — it’s just raw data, and there’s anything and everything in there. There’s instrumentation glitches, there’s star spots, there’s other astrophysical noise in that data. And you’re looking for something very specific out of that, right? And that’s what your work with the data processing really is, the science data processing pipeline. Could you say a little bit about what that pipeline is, and why that’s so important to what we get out the other end, which is the usable data that scientists can go out and discover planets?

Jon Jenkins: Right, so you’re absolutely correct that it’s a very challenging task. And what we have to do is go from what we say, pixels to planets. And that is that we’re taking down this stream of image data. So, imagine if you will we’ve got Kepler up there. It’s taking images of the stars every half hour. We can’t afford to store all the data onboard, so what we do is we cut out the images of the stars that we want to find planets. It’s kind of like taking your high school yearbook, cutting out the pictures of your friends and pasting them on a new page and throwing the rest of the book away.

But then, you get the data down, and there are artifacts due to the instrumentation that you have to remove, and you have to put the digital numbers, because it’s just a stream of numbers, into physical units so you can properly interpret them. And essentially, you’re looking at a set of pixels with an image of a star moving around slightly on it. And you add up the pixel values, the numbers, to estimate the brightness of the star. And you have to do things like, you have to identify cosmic rays, because those happen.

And then, when you look at your, what we call “light curves,” the measurements of brightness of the stars over time, you have instrumental effects. For Kepler, one of the biggest instrumental effects is the fact that the telescope changes its state in response to the changing thermal state or the temperature over the spacecraft and the instrument. And it turns out there was a heater on the reaction wheel assembly that kept the reaction wheels nice and happy. But if it turned on, it would swing its temperature by five degrees Celsius, and that would cause the shape of the telescope to change by .1 microns. So, it would change the distance between the 1.4-meter primary mirror and the focal plane that was 1.4 meter away. So, you could see a .1, a tenth, of a micrometer change in the distance between those. And you could easily see the signature of this triangle wave imposed by this heater in the pixel data. So, it’s crazy-sensitive.

A friend of mine said once you invent an instrument that is an order of magnitude better than anything in its class, you’ve invented the world’s best thermometer at the same time. That’s certainly true with Kepler. To answer your question, the next stage of the processing is to identify these instrumental signatures and remove them. And we were surprised. This was a big surprise on orbit. We thought based on pre-flight testing and models that we would be mainly dealing with pointing errors. So, if the pointing got a little bit worse, got a little bit better, that would modulate or change the brightness measurement. But it turned out that we’d underestimated and underappreciated how important the thermal state of the telescope was, and how the focus would change over time. So, we had to go back to square one and come up with a new method —

Host:And filter that out, basically.

Jon Jenkins: Yeah, come up with a completely new methodology for how to deal with that situation.

Host:And the thing that I find fascinating, because this is — going back to the pipeline — there’s so much information packed in there. And right now, there’s people who are writing papers, they’ve published in journals and stuff, about the information they found there. But what about, there’s so much information still hiding in there. And so, I’m thinking long after the telescope is retired, long after it’s everybody’s — we’re no longer operating the telescope, there’s still going to be science papers and analyses and stuff coming long after we’ve all packed up and moved along.

Jon Jenkins: That’s absolutely right. And in fact, people may be able to come up with better algorithms, better software for identifying and removing these instrumental signatures. After we do that, then the goal is to have the brightness changes over time, and each star only represent what’s happening on that star. And we get a lot of information from that, not just about transiting planets, but about the star itself. We talked about star spots before. You can actually see the signatures of the star spots, measure how long it takes for them to go around the star, and get an estimate for what the rotation period of the star is. And that, together with the temperature of the star, can give you an estimate for the age of the star. So, that’s telling you more about the system.

It’s also true that one of the most fascinating things we did with Kepler was to measure acoustic oscillations in a large number of these stars. They’re just like big bells, and they ring. And they ring at a set of frequencies that is connected.

Alison Hawkes:And what does that mean exactly? What does it mean for a star to “ring?” Is it like the movement of the plasma, or something?

Jon Jenkins: Yeah, so it’s this big ball of fluid, of gas, and it’s kind of like think about a big bowl of Jell-O without the bowl. And now, if you perturb it in any way, it’s going to jiggle, right? It’s going to jiggle, and communicate that jiggle across it, and that might bounce back, and so you might set up sort of like a coherent jiggling that has a period to it. And so, the same thing happens with a star. If you have a disturbance on the surface, which can happen because you have convection, and you can have turbulence, then you will have acoustic waves, sound waves, travel down through the plasma. They will get refracted, or they will bend back up to the surface at another point. They will bounce off the surface, go down, and it’s kind of like a daisy chain where, if they come back to the same point they started, they’ll keep on going.

But, you have all kinds of things going on, so you setup lots of these modes. And so, stars ring in different ways, they oscillate in different ways, but you can take the frequencies of the oscillations you measure and infer the mass and the size of the star to within a couple percent. And that’s amazing, because these stars are thousands of light years away. And we can’t measure the disk. It’s not like you’re measuring the disk of the star, we can’t see that. We can’t resolve it.

Host:And I guess piggybacking on the pipeline thing, I’m just curious because you say pipeline and automatically I’m thinking of an oil pipeline. You’re taking something from one place to another. Maybe walk through that for folks who may not be aware. I’m guessing it’s telescope down to you guys on the ground, then that’s like process it. At that point, are you piping it out to the scientific community? I’m just guessing here.

Jon Jenkins: So, for Kepler, Kepler would turn after a month of collecting data. And it would send down the image data through the Deep Space Network to the ground, and then it would be piped through the Ethernet to the data management center at Space Telescope Science Institute. They would reformat it into files that they would then send to us. And these files would basically contain the image data as a function of times. We’d have one file per half-hour for each and every star that you were observing.

So, we would put all those files on a server, and then we would — we have a pipeline infrastructure that’s a control system that basically controls all this. So when the files arrived, part of the system would ingest the data into our database, and then it would fire off the first stage of the processing, which is to do these image-level calibrations we talked about. For example, one of them is the fact that we don’t have a shutter on Kepler, but it takes .5 seconds to read the data out. That means the uncalibrated images have these vertical smear trails from the fact that as you clock the charge, the image down the CCD in each row, it picks up a little bit of light from the stars that are under it. And so, that’s one of the things that we make special measurements to correct out.

But we do that processing on the supercomputer now. In the early days of Kepler, we didn’t need the supercomputer, because we could keep up with the data. But in the out years, we moved everything to the supercomputer, so instead of processing the data on say 400 computer cores we had available, we could process it on 10,000 computer cores available at the supercomputer. This allowed us to keep up with the data, but also it allowed us to go back and reprocess the data as we learned about the sky, as we learned about our instrument, and we could evolve our algorithms and have better science products.

Alison Hawkes:You essentially started from scratch with this, right? There was no pipeline to start with. You put this all together, you and your team?

Jon Jenkins: We had some heritage, so yes. We had the data processing software we developed for these laboratory-based experiments, so we had some ideas of what we had to do with the data. And then, one of the things that we did before we were finally selected in the fifth Discovery proposal process that Bill competed in was that we built a camera that we put up on Mount Hamilton at Lick Observatory. And we used that to observe a seven-degree-by-seven-degree swath of the sky. And we called it the Vulcan camera. And so, I developed the pipeline for that with a small handful of people. And so, we had that software as a basis to start Kepler. But Kepler was so much more data —instead of observing 10,000 stars, we were observing 150,000 stars.

Alison Hawkes:It was a brand-new instrument, right? Yeah.

Jon Jenkins: That’s right. We needed professional software engineers to come and help us scale up the system so we could actually do it with a lot of confidence and that it would run every time, and you didn’t have to have someone go in there and change things.

Alison Hawkes:How important is having a pipeline like this? I know some members of the scientific community develop their own pipelines, right? So, what’s the value added by having our own in-house pipeline?

Jon Jenkins: Well, the main advantage for the science community is that we’re set up to process and deliver the data on a regular basis, and we’re also set up and funded to document what we do to the data, to document the characteristics of the data. And so —

Alison Hawkes:And it’s all publicly-available, right?

Jon Jenkins: It’s all publicly-available.

Alison Hawkes:This is the pipeline, right Jon, that’s going to be used for TESS, a reworked version of it?

Jon Jenkins: Yes. So, we took the Kepler pipeline and we ported it and reformulated it for the TESS mission. And the biggest challenge there is that TESS is collecting 13 times as much image data as we did for Kepler. And so, Kepler basically opened the door for the TESS mission and for other exoplanetary missions.

Host:JWST [James Webb Space Telescope] is like — everybody loves exoplanets now. They’re falling over each other.

Jon Jenkins: That’s right. 25 years ago, it would’ve been hard to imagine all of the excitement, and all of the interest, and all of the work that’s going into these future missions. So, we’ve got TESS launching. TESS is going to do an all-sky survey. Yes, we get to use our pipeline, but we’ve souped it up. We’ve found some ways to simplify it. We process a month’s worth of data. It takes us somewhere around a week or so to process a month’s worth of data.

Alison Hawkes:And you’ve been able to do that with TESS because of advances in computing power, right? The sheer amount of data coming from TESS is just enormous, right?

Jon Jenkins: That’s right. The supercomputer keeps on getting more super, our data storage devices get faster and bigger, the Kepler spacecraft, for example, could only store about 10 gigabytes of data onboard, and that represented 66 days of data collection. With TESS, we’re able —

Host:Modern technology.

Jon Jenkins: Modern technology. The data storage is 185 gigabytes, so yeah. It’s about 18 times more data storage onboard, and we can send all that data down. So of course, we’re going to collect a lot more data and send it down.

Alison Hawkes:How excited do you think the astronomy community is that’s about to come down from TESS? This must be a great deal of excitement.

Jon Jenkins: There is. They’re terribly excited. I think one of the fortuitous things is that we had Kepler, and then the best/worst thing happened to Kepler, K2. We lost our second reaction wheel in May of 2013, after four years of observing. A Ball engineer named Doug Weimer came up with an idea of how we could continue to point the spacecraft stably at a given point in space for a couple months at a time, and that became the K2 mission.

The community has then rallied around K2, because we can do a lot more different kinds of science on the fields of view we can observe with K2. And so, the K2 community now is starting to cut their teeth on synthetic science data sets that we produced in the TESS project, and that we here at Ames generated via a high-fidelity science simulator that we developed specifically for TESS that was modeled on the one that we did for Kepler.

And so, that data is up there on MAST [Mikulski Archive for Space Telescopes] now. It’s called the end-to-end six data set. And there was a workshop that happened called the TESS Ninja Workshop in March, and they went to town on that data set. In fact, within a couple of days, the first paper appeared on astro-ph, the science server, about making light curves from the synthetic full-frame image data that we put out there.

Alison Hawkes:Are you feeling ready for TESS?

Jon Jenkins: I’m ready. We — we’re not twiddling our thumbs. We’re tuning things up. But, we delivered the functionality back in December, and so we’ve been participating in simulations and rehearsals, and we’ve been stress-testing the system to make sure everything’s just right.

Alison Hawkes:You’re like one of the first people who gets to see the data, which is pretty exciting. Same with Kepler. And you have a great story about first seeing the Kepler data. And maybe you could rehash that and tell us a little bit about your anticipation for TESS.

Jon Jenkins: Okay, well, it took us seven years on Kepler to get selected for flight, and then it took another eight years before we could build and launch it. So, I had a little trepidation the first morning I arrived at work. I was the first one at work. We’d gotten the message the night before that the first science-like data set had come down, 10 days’ worth of observations on 53,000 stars. I wasn’t quite ready to sit down at my work station and look at it. My heart was thumping, my stomach was in my throat.

So, I decided to wash the coffeepot. Now, we never washed the coffee pot, but I decided that day I had to wash the coffee pot. That gave me an extra five minutes to compose myself, to sit down, to face the music if it was bad. But it wasn’t. I pulled over the data. I started displaying and visualizing the light curves and the eclipsing binaries. And the giant transiting planets just fell in our laps. And in fact, we were looking at stars that were variable stars. And my friend, Doug Caldwell, had the Encyclopedia of Variable Stars — it was a graphical encyclopedia. And I’d say, “Doug, what kind of star is this?” And he’d flip through the encyclopedia and say, “I have no idea, Jon.”

And so, that was one of the biggest Kepler discoveries, was that we found that some of the variable stars actually are hybrids. We found —

Alison Hawkes: So, you were seeing stuff almost immediately in the data?

Jon Jenkins: We were. In fact, one of those things was that there were three previously discovered exoplanets that were in our field of view that were discovered before we launched, and of course they weren’t on our target list. One of those was Hap P7B. Now, this is an inflated Jupiter in a 2.2-day period orbit about its star. And we knew that when we looked at this light curve, we would see 1.6 percent drops in brightness every 2.2 days corresponding to the planet.

What we didn’t recognize at the time was that when you looked at what the light curve did in between the transits, you saw it was bow-shaped, kind of like an upside-down smile. So, the light increased in between the transits and reached its peak just in between. And then there was a notch cut out at the very middle, at the very top. And it turned out that when we read the Discovery paper, because we pulled it right down that day, it represented a really great candidate to look for and detect thermal emission from this planet, because it’s so hot, it’s so close to its star, it’s kind of like a burning ember. It’s glowing on the side that’s facing the star. And it’s kind of like the phases of the Moon. When it transits, it’s in between us and the star, but then its orbit carries it around the star, so we see more and more of its Sun-kissed face. So, it’s like the Moon going to full Moon, this went to full planet. Except then it went behind the star and the light from the planet winked out.

Now, the fascinating thing here was that the depth of the notch which corresponded to how much light we lost because the planet was going behind the star —

Host:You were seeing the other side of it.

Jon Jenkins: —yeah, was 100 parts per million. That’s very special, because that’s the same size you would expect for the transit of an Earth-sized planet going in front of a Sun-sized star. So, whoa, that was such a relief. I mean, we demonstrated that Kepler right out of the box had the sensitivity to allow us to find Earth-sized planets transiting or crossing in front of Sun-sized stars.

Alison Hawkes:So, you’re going to be doing this with TESS. You’re going to be the first one to look at some of the TESS data and see what it shows us. How you feeling about that moment? You going to clean the coffee pot again?

Host:I know, right? Everybody, bring your coffee pots to Jon’s office. It’s the only chance you’ve got to get them cleaned.

Jon Jenkins: I’ve got the Brillo pads, I’m ready. I think it’s going to be a lot of fun, but Kepler, there were some real nail-biters right off the bat. We went into safe mode during launch, we had some safe modes during commissioning. Every once in a while, Kepler got seasonal affectional disorder and decided around Christmas it was too lonely and it would shut down for a while. And after a while, I got inured. I got used to the fact that every so often something might happen, but we always were able to resurrect it. Even when we lost the second reaction wheel that was a real blow. We were all real glum for a while. But then it resurrected just like Phoenix from the ashes with something that in my mind is, in many ways, almost as good as Kepler, and in some ways better than Kepler.

With Kepler, I just wish we’d had that extra two to four years, before if we’d had three reaction wheels, we could’ve done a four and a half year K2 mission or longer just as well.

With TESS, it’s going to have some surprises for us, and we haven’t even talked about this, but we collect a full-frame image with TESS every half-hour. This is 15 times the sky area we collected with Kepler.

Alison Hawkes:It’s 1/26th of the sky, I think you said that.

Jon Jenkins: 1/26th of the sky, so that we’re going to change our field of view every month, and after 12 months, 13th pointings, we’re going to be able to map the entire southern hemisphere, then we’re turning the spacecraft upside down and we’re going to observe the northern hemisphere.

But what’s so fascinating to me with Kepler is going to be 15 times more amazing for TESS. And that is that with Kepler we discovered things and objects and systems we had no idea were there. We found a small sub-Mercury-sized disintegrating planet. Nobody thought we’d find something like that. We found signatures of exo-comets, we found Tabby’s Star, we found circumbinary planets, we found heartbeat stars. There were so many things that we found with Kepler that we had no idea were out there to find.

Think of what it’s going to be like to have a movie of the sky, of the full sky after two years. That’s going to be so totally amazing. I can’t wait for all the science that comes out of that that we have no idea what it is, but it’s going to be great.

Host:So, for folks who are listening, if you have any questions or comments, we are @NASAAmes. We’re using the hashtag #NASASiliconValley. Also, we also say we’re a NASA podcast, but we’re not the only NASA podcast. You can always check out our friends over at Houston, We Have a Podcast over in Texas. There’s also Gravity Assist over in D.C., and also This Week at NASA. But for music fans, there’s also Third Rock Radio.

The best way to grab all this stuff is to hop on over to the NASA app, or we even have an omnibus RSS feed called NASACast, where it takes all the NASA podcasts and puts them all into one place.

But Alison, thanks for tag-teaming this.

Alison Hawkes:Yeah, my pleasure.

Host:But Jon, thank you so much for coming over, this has been fascinating.

Jon Jenkins:Well, thank you Matt, and thank you Alison. It’s my pleasure, and stay tuned.

END