>> Stephanie Marcus: Good morning, I'm Stephanie Marcus from the Science, Technology, and Business Division here at the library. Welcome to our lecture today. We're going to come back to the home planet. And today we have Dr. Tom Neumann, who's the deputy project scientist for the ICESat-2 Mission. So we got the right guy. And he's accompanied by Adriana, one of his assistants today, so welcome to her. Dr. Neumann taught at the University of Vermont and did some ice and snow projects that were funded by the National Science Foundation and NASA. And then he came to NASA in 2008. He got his BA in Geophysical Science from the University of Chicago and a Ph.D. in Geophysics from the University of Washington. He's spent a good deal of time in the cryosphere and somewhat recently, made a 470 mile trek across Antarctica with three others. I've read that he also plays guitar with other ice scientists, so that sounds very interesting. [laughter] And [laughs] I do hope he flavors his talk with some tidbits about the expeditions and the ice band. I'm also happy to announce that Dr. Neumann did some research here at the Library of Congress a while back, so please help me welcome him back to the library. Thank you. [ Applause ] >> Dr. Tom Neumann: Great. Thank you much. I did, in fact, come to do some research. When I was at Vermont, I was there from 2001 to about 2008, and their geology program there is pretty small. I was in the geology department, and I was the only ice person. [laughter] So I mean, the number of journals they had, things weren't online as readily as they are now, and so I would put together a list of articles I just couldn't find anywhere else. And I had a brother who lived in Alexandria. I'd come to visit a day early, and I'd spend the day -- spend the day getting articles. It was -- yeah, it was awesome. It's sort of funny that I had my little Library of Congress library card, and I'd tell people, I went to the Library of Congress. They're like, is that a real library? I'm like, it's an -- [laughter] It's an unbelievable library. You can get anything you want there. [laughter] Journals from anything. But mostly what I'm going to talk about today is not the ice band. I'll maybe touch on it at the end if people are interested. And I do have some field pictures. But most of what I'm going to talk about is ICESat-2. That's the project I work on. NASA has just recently launched two missions to study the cryosphere, ICESat-2, and the other one being GRACE-Follow-On. I'll talk just a bit about GRACE-Follow-On. But mostly you'll hear about ICESat-2, since that's what I know best. So the polar regions. We use words like the "polar regions of the cryosphere," so I wanted to just have a couple slides at the beginning to put us all in the same page about what we're talking about. There's two main pieces to the cryosphere. Once is glaciers and ice sheets, and they're formed by accumulating snow. It's so cold in Antarctica and Greenland that it rarely melts very much, and so snow just accumulated year after year, decade after decade, thousands of years at a time. It doesn't snow very much down there. It's a polar desert. But it doesn't melt, so it just builds up. These ice sheets can be 4,000 meters thick. You might have heard about ice coring projects that drill down through the ice sheets to collect ice that fell as snow tens of thousands or hundreds of thousands of years ago. That's the kind of place they come from. They move. They move really slowly though. They deform under gravity, just like pancake batter. When you pour it on the pan, it kind of slides out towards the edge. But they don't move very quickly. In the middle of the ice sheet, it might move just a few meters a year, sometimes less. They lose mass. So they gain mass by snow fall. They lose mass by generally calving icebergs at the edges of Antarctica. Whenever there's a big one, it always seems to make the news. Recently there was a square one somebody took a picture of, and so we got phone calls about, what do you think about this square iceberg. And I'm like, it's square? I mean, what do you want me to say? [laughter] It's a thing that happens. In Greenland, it's warmer, so actually about half of the mass lost there is due to melt. So the iceberg that sank the Titanic almost certainly came from Greenland, and it does calve, but it also melts quite a lot down there. The other main piece of the cryosphere we'll talk about is sea ice. Now, sea ice is different than glaciers or ice sheets. Sea ice is formed by freezing of ocean water. So the air temperature gets cold, that will freeze that top layer of water, just like you might see a frozen lake or a frozen pond. But in the ocean, it's much more dynamic. There's tides, there's wind, there's waves. And so the sea ice moves around quite a lot. It's much thinner than glaciers. It only, at its maximum, gets up to about 20 meters or so thick. In the upper left, you can see there's some really freshly formed sea ice and the wind has caused it to raft on top of each other. You see -- if I can get my pointer going here -- see areas where it's overlapping. You see places where there's open water between it. And if that sea ice is allowed to continue to grow, it'll start to collect snow fall on top of it, and so that surface will turn from a dark or translucent to a bright white, as it collects snow on top of it. Sea ice melts seasonally. When the weather gets warm in the Artic or Antarctic, sea ice extent shrinks and then it grows again the next fall and winter. And it pretty much forms -- well, anywhere it's cold enough to freeze the water, so places like Hudson Bay get sea ice. A few years ago, the Great Lakes froze over almost entirely. Of course, they're freshwater, so they have a different freezing point. But around the Arctic Ocean and around the Antarctic continent, you get sea ice annually. So those are the two terms I'm going to end up using that I wanted to make sure I defined. So why is NASA interested in these kind of areas? You might know that places like Greenland are losing mass. The Greenland ice sheet is getting smaller. And what we've learned over the last 10 years or so is that most of that mass lost from Greenland or from Antarctica is concentrated at the edges of the ice sheet. Now, it turns out these are the places where ocean interacts with the ice sheets. These glaciers tend to terminate in water, in the ocean water. So as ocean water gets just a little bit warmer, that's a large amount of heat that's available for melting more ice. So this is a map of Greenland and where we've seen changes there. In Antarctic, here's a time series from a mission called "GRACE." And what GRACE does is it measures the mass change of the whole ice sheet, and I'll talk about that in a little more detail. But the take-away here is that through time -- so you can see time marching forward here -- you can look at the record of mass loss measured by GRACE. Most of the loss is here in West Antarctica. You might have heard about places like Pine Island Glacier or Thwaites Glacier that are losing mass. And that's a place where the ocean water contacts the ice sheet. So as that ocean water warms just a little bit, it has a big impact on the ice sheet. So this place over here is called West Antarctica. This little thumb is the peninsula. And the bigger piece here is East Antarctica. East Antarctica is probably three-quarters or so of the total amount of ice in Antarctica. And it doesn't change nearly as quickly as West Antarctica does. So that's one reason why NASA is interested, because these ice sheets are changing and they're losing mass, and that mass, that water, when it leaves the ice sheet, it just goes back into the world's oceans. And that's one of the main contributors to global sea level rise. You took ice that used to be sitting on rock in Antarctica or Greenland, you put the ice cubes back in the water, and they're going to raise sea level. So that's one of the main reasons NASA's interested. For sea ice, if I can get this guy to play. We talked about how sea ice is pretty dynamic. And in this animation, you see time marching past up here in the upper left, and these are images of the sea ice extent in the Arctic Ocean. You can see Greenland over here, Alaska over here, Canada here. And this should give you a sense for just how dynamic that sea ice is. It grows and shrinks annually, and when storms come through, it moves it quite a lot. You can see it generally tends to form over in this part of the basin and is lost right down the coast of Greenland over here. You also lose some over in this area. And we've been able to take pictures of sea ice extent from imagers, since about 1979. So that's how we put together this time series of how the extent has been changing. Now, one of the things we're really interested in is how the thickness of sea ice changes. Really thick sea ice, it's hard to break up. It lasts much longer than very thin sea ice. If it's first-year sea ice, you have a storm come through, it can break up that ice. And what happens when you lose sea ice is that it increases the transfer of heat from the ocean to the atmosphere. And you say, all right, well, why would we care about that? It turns out that the exchange of heat from the ocean to the atmosphere in the Arctic is one of the big things that drives climate patterns and weather patterns around the world. Things like the jet stream position is influenced by the heat exchange up here. Whether we have a snowier winter here in D.C. is determined, in part, by the state of the sea ice in the Arctic. So that's really a driver for weather patterns around the world. So the more we can understand about how it grows in shrinks, including the thickness, the better. Oop -- and I'm going to go to the next one. There we go. So if we just look at the area of Arctic sea ice in September of every year, we can make a time series for how the extent has changed. We've picked September because that's basically the end of the summer season in the Arctic. You get into October, it gets cold, sea ice starts to regrow. So what we're looking at here is a graph of the minimum sea ice extent in the Arctic. It's usually about the second week of September. And what you can see is that since 1979 here, it was kind of flat for a most of the '80s, but you get into the '90s and you get bigger and bigger changes. And now we're -- oop. I should put the new one in. Now we're out here in 2016, there's actually two more data points on there. It's a lot smaller than it was back in the '80s. And remember, this is the extent. This doesn't tell us anything about the thickness or the stability of it. It's just how much was there. The mission I work on is called ICESat-2 and it's designed to measure that thickness of the sea ice. In the Antarctic, the seas around the Antarctic, the changes are much more subtle. The graph up here is the change in the Arctic, which we just saw a moment ago. For the Antarctic, it's pretty flat. It might be increasing a little bit. You can see there's been this change in about the last four or five years. It's hard to know if that's something permanent or short-term. But at least since the early '80s, since 1980 or so, Antarctic extent has been pretty flat. The Arctic has been losing sea ice area over time. So that's why NASA's interested. One of their mandates is to understand the planet we live on, Earth, and the cryosphere is a big piece of that. So as I mentioned, we have two new missions to study the cryosphere. GRACE-Follow-On, its launch is up here in the upper left, launched in May of this year. And the mission I work on, ICESat-2, just launched about -- not even two months ago. We launched in the early morning hours, so we had a really nice view of the rocket. GRACE-Follow-On launched in the afternoon. Both from the West Coast, Vandenberg Air Force Base out in California. So GRACE-Follow-On. As you might imagine from the name, there was an original mission called GRACE. We saw some of its data earlier. This is the Follow-On mission to that. And GRACE and GRACE-Follow-On are composed of two satellites. You can see the one in the foreground, and there's another one way off here in the background. And what the mission does is it measures the distance between those two satellites very accurately. It uses microwave ranging between them. So it can range to better a millimeter how far apart these two satellites are. And the way it works is that as the lead satellite goes over a place in the earth that has relatively more mass or less mass, it experiences a slight acceleration and it moves a little farther away from the trailing satellite. Now, when the trailing satellite gets up to that place where there's a little more mass, it also accelerates, and so the distance between the two is a measure of how much mass difference there is in the earth. So over Antarctica, when the Antarctic ice sheet is losing mass, we compile data from many passes of GRACE over the same places in Antarctica to determine, is it gaining mass or is it losing mass? Is the front satellite accelerating just a little bit more than it was last time, or maybe a little less than it was last time? GRACE was mainly designed to study the hydrologic cycle. It can do things like measure aquifers. Are we losing water out of aquifers or are we gaining it? Is there more mass here, is there less mass here? It can measure changes in seasonal snow, which is totally cool, because you get a big snowfall here and you could measure that in GRACE data. So what GRACE is doing is measuring mass changes primarily. Given that, it's on orbit 350 kilometers above the surface of the earth and these two satellites are hundreds of kilometers apart. You end up with a pretty low spatial resolution measurement. And what I mean by that is you can measure mass changes really accurately, but only an average over hundreds of square kilometers. So it's an average over a big area. However, given its orbit, you get a lot of data through time. So you can get a data point over a place about once a month, which is great. We saw that time series earlier of mass change in Antarctica from GRACE and that was a data point about once every month. GRACE and GRACE-Follow-On are by NASA's Jet Propulsion Lab out in Pasadena, California, and that's all I'm going to say about GRACE, because I work at one of the other NASA centers, right out here in Maryland, NASA's Goddard Space Flight Center. And we recently launched ICESat-2. Now, at ICESat-2, as you might imagine, is also a follow-on. There was a first ICESat mission, although it'd be pretty bold to name your first mission ICESat-2, just to mix it up and confuse people. [laughter] We didn't do that. ICESat-2 is a laser altimeter, so it has a laser on board. And what it does is it sends out a small pulse of laser light, and it times, super accurately, how long it takes that light to go from the space craft to the earth and back again. So that's essentially a range. And what it does with that is we reconstruct elevation of the earth's surface. We do that by taking that time, how long did it take to go from space craft to the earth and back again, figuring out where we were pointing or am I pointing at the wall over here? Am I pointing at the wall over there? And where are we in the orbit? You put those three pieces together and you can measure the height of anything you fly over, just by bouncing these laser beams off. The footprint size of the laser beam on the ground is about 17 meters, so sort of the size of this room. So we get really detailed information. We can measure the height of something about this size. Whereas GRACE was over hundreds of square kilometers, we're measuring meter scale. However, we have fairly low resolution temporally. We've got these little spots on the ground, and our orbit repeats about every 91 days. So we can see about times a year how a particular place is changing. Together, these two missions, GRACE-Follow-On and ICESat-2 are really complimentary. One gives you great time resolution, the other gives you great spatial resolution. You put those two things together, you we should be able to learn more about how ice sheets and sea ice are changing than we've ever been able to before. So I'm going to walk you through a little bit more about how ICESat-2 works. The main instrument on here is called the lidar. That's the thing that generates the laser light. And the time it takes to go from space craft to ground and back again, we call a range measurement. It's essentially a distance. The other piece we need is the pointing angle. Are we pointing straight down? Are we a little left? Are we a little right? And then the last piece is where are we in the orbit? Are we over Maryland or are we over Canada, or are we over Greenland? And we figure that out with GPS, just like you would in your phone. Certainly, these are much fancier GPSs and give us much more accurate data than you would get from a phone. But basically, you take these three pieces: pointing angle, range measurement, space craft altitude, and you figure out the height of the earth. If you measure that height at the same place through time, you can tell, is it getting higher or is it getting lower? From the vantage point of space, you can see how ice sheets are changing, how sea ice is changing, oceans, deserts, forests, all of it, which is totally cool. A couple of slides about technical details, in case anybody is interested. The laser on ICESat-2 fires 10,000 shots a second, which is an incredibly fast repetition rate. And it sends out really small pulses of light, so micropulses, we would call them, in each of six beams simultaneously. That's the multibeam part of it there. And the way it does its detection and its measurement, it sends out these light pulses of light 10,000 times a second, and it's actually detecting individual photons as they come back. So we have super sensitive detectors on board for each of those six spots. So people say, oh, you know, could I see the laser light from the space craft? Probably not, because they're so low energy and the number of photons that we're detecting on the way back from each of these shots is kind of 1, 2, 3, 4. It's really small number of photons coming back in, but that's the basis of our measurement. With this super fast firing laser and our altitude, we get a footprint that's about 17 meters in diameter every 70 centimeters along each of those 6 beams. So the footprint from one shot to the next overlaps the previous one by about 95%. So we end up with continuous coverage in the along-track direction. It's a tremendous amount of data. And it's a very NASA way of measuring things, you know? It's a laser in space, 10,000 times a second, counting individual photons. It's kind of crazy. And the cool thing is, is that later in the presentation, I'll be able to show you what some of our early data looks like. We've been on orbit now for about six weeks. We're still in calibration phase, but it's astounding what you can see with this kind of instrument. The last slide about technical details is here. If you were able to see each of these six spots on the ground, you'd see that we've arranged them in a rectangular pattern, so that as the beams are described on the surface of the earth, we have these beam pairs. We have two beams right next to each other, separated by about 90 meters. And what that does for us is it lets us measure the elevation in the along-track direction, sure. But it also lets us measure the slope in the across-track direction. Is this one higher than that one, or vice versa? And it turns out that's a key piece in figuring out how ice sheet elevations are changing. So we have six beams arranged in three beam pairs. And if anybody's interested in what wavelength we're using, it's a bright green light, 532 nanometers. We split it into six beams, and yeah, we're detecting individual photons as they come in. I realize that for folks who don't work on this kind of thing, it all sounds a little bit esoteric. So we actually have a little animation that -- that really brings it home, what this is all about. So let's see if I can get this thing to play. Doo-doo-doo. [music] This is perfect for management level. [murmuring animation sounds] >> No, no, no, no, no! >> Attention, photons. Prepare to measure Earth's changing ice. [music] [inaudible] pulse 22A4. In 5...4... [inaudible] [music] 3...2... ONE! [whirring sound] [music] >> Dr. Tom Neumann: There you go. >> [video] WEEE! WEE! [shouts of excitement] >> Dr. Tom Neumann: Laser light comes out, tons of photons, down towards the earth. [video] [photons murmuring excitedly] [music] >> Dr. Tom Neumann: Couple of them make it down to the earth's surface and bounce back towards the space craft. >> [video] Weeeee! [ Music ] >> Dr. Tom Neumann: Turns out photons get discouraged, you know. [laughter] It's a long way. [laughter] [photon gasps] [ Triumphant Music ] [ Laughter ] >> Whoa! [laughter] [ Triumphant Music ] [ Laughter ] [ Music ] >> [video] Excellent job, pho. Your travel time has been recorded. [music] Let's see the data elevation dataset. Ready the next group! [music] [laughter] >> Dr. Tom Neumann: You know, it's a little tongue-in-cheek, but it really is the basis of how this works. We send out a whole bunch of photons, we're detecting just one or two on the way back and recording their travel time. And Adriana's here, and she was the lead -- one of the leads on putting this one together, of coming up with a way to explain how a laser altimeter in space lets you measure height. That's actually how it happens. [music] Except photons don't get discouraged. [laughter] They do wear sunglasses though. I've seen them up close. [ Laughter ] So ICESat-2 is in orbit now. Our orbit takes us up to 88 degrees north and down to 88 degrees south, and we do about 15 laps of the planet every day. Can you guys see this okay? From my perspective, it's a little bit washed out, but you can see it cruising around the planet there. You'll notice there's lots of clouds on Earth, right? You go outside, maybe it's a sunny day, maybe it's a cloudy day. We're using really low energy visible green light. So on a cloudy day, we're mostly going to measure the tops of the clouds. The light will bounce off the clouds. You might have seen that in the last video, where some of the photons got scattered by the clouds or bounced back. They didn't make it all the way to the ground. But places where we have a clear sky view, we'll detect photons coming back from the surface of the earth. We're called ICESat-2 so the instrument really is optimized to measure changes in sea ice and in glaciers and ice sheets. But from space, as I mentioned earlier, we really do get a measurement of everything. We'll get lots of data over the oceans. We'll get lots of data over land, over cities. And we have data products that cover each one of those things. The data product I work on, the goal of that product is to generate unique latitude, longitude, and elevation for each photon that we detect. So with a laser firing 10,000 times a second, six beams at the same time, order of about five photons coming back, something like that. So one way to think about it is that in the time it takes to blink about half a second, ICESat-2 will do on the order of about 15 or 20,000 individual measurements of photons. So it's a tremendous amount of data. The laser is on 24/7. It's on all the time, collecting data. So these data products are big. If you're interested in ice sheet elevation and ice sheet elevation exchange, you probably don't want to deal with, you know, billions of individual photons. You're going to average those together. Or similarly, if you're interested in tree height, you don't want to deal with things at the photon level. The data sets are simply too big. So what the higher level products do is to aggregate those photons together. Maybe they'll take 40 meters or so worth of photons or 100 meters worth of photons, average those heights together to come up with one elevation measurement that has much better precision than a single photon. So I'm going to -- we watched that one already. Oops. So what does the data actually look like, if you're going to look at it? So we talked about how it's detecting individual photons, right? And we're assigning an individual latitude, longitude, and height. This turns out what you would see if you plot it. We have elevation along the y-axis, and all of these plots, and either time or distance along the x-axis. And this is for data for a single beam. You see each one of these little blue dots? That's a photon that ICESat-2 detected and sent the information back to the earth, and we assign that latitude, longitude, and height. Now you see there's a lot of speckle in these random dots. We use green laser light, but it turns out the sun also puts out light in the green part of the spectrum. So that's background light from the sun. At nighttime, you don't see any of that. In the daytime, this is about what you see. And the photons that are reflected off the surface are this bright blue line that jumps out at you. That line is actually composed of millions of tiny -- of individual photons that we detected and located. So your problem then becomes, if you're interested in the ice sheet's surface, is figuring out exactly what the elevation of this line is. So that's why they're doing averages and combining lots of photon data. With just a single photon, let's say this one, I can't tell if it's a background photon or is it a signal photon that I'm interested in off the surface? But your eye just naturally aggregates photons together and it's just drawn to this bright blue line. It turns out this is real data from space collected about two weeks ago over Antarctica. And the next couple slides will show some of the data over different surface types to show you what the data looks like. But each one of the slides will be basically the same. Elevation along one axis and time or distance along the other. You can think of it as, like, a line across the stage here, and you're looking at it in side view. So up above is higher elevation. This is the sky and the space craft up here. Down below is the center of the earth. Does that view make sense to folks? Each of these dots? Yeah, okay. Cool. So we take data over ocean, right? Here's a track that went over the Atlantic, just off of the coast of the U.S. You can see Maine over here, Canada over there. And we have elevation along this axis. Here's about four and a half kilometers on that axis. And it looks really kind of lumpy through there. But it turns out if you zoom in a little bit, what we're doing is we're detecting individual ocean swells. [audience murmurs] The height of this is -- these ticks are two meters apart. So these things are kind of four meters high are so, about 700 meters along-track. So that's four meters of relief in each one of these waves or swells that are tens or hundreds of meters long. So wavelengths in this picture are about on the order of 140 meters. You see there's relatively little background light, so the sun was pretty low in the sky when we took this pass. But that's pretty cool. But yeah. This is all data from space. We had simulated data ahead of time to work on this, that, or the other. This is real data. It's mind boggling to me in some ways how well it's working and the cool things you can see. Here's some data over Russia, a forest in Siberia. You see all this background light. This is daytime data. Certainly, you can see these blurry things here at the surface. Those are returns off of trees. What you see is this bright line underneath, that's the ground beneath the trees. And some photons are being reflected off the top of the canopy, some off the middle of the canopy. So we can use ICESat-2 data to figure out tree heights around the world. We take this height, compare it with the height of the earth, and there's your tree height. Here's a little blow up of that scene right there, and you can see these trees are kind of on the order of, oh, 15, 20 meters high. So they're pretty big trees. They're kind of 60 foot tall trees. You can also see a break in the canopy right here, either a road cut or a stream or that sort of thing. Here's an image from Google Earth with our tracks on here. And then here's the little road we're crossing right down there at the bottom. So that's pretty astounding. And we haven't even got to the good stuff yet. [laughter] Here's a track coming off of Mexico into the Pacific Ocean from left to right. Again, elevation on this axis, distance on that axis. We're coming down the forested hillside. You can see the trees, just like we saw in the last picture. And then we cross over into this lagoon. You get a bright return off the surface, but some of that green light is actually getting down into the water, and we're picking up the bottom of the lagoon, as well. You can measure depth of water up to some distance with this data. Compare the surface with the bottom and you can say, all right, well that's about four meters deep, which is totally cool. You go across a little shawl right here and you can see these individual waves in the ocean. The water in the lagoon is nice and still, as you'd expect. And you get ocean waves, as we saw earlier. Here's another scene from Mexico. We have the Gulf of Mexico over here. We're getting a lot more reflections in the water column here. It's because the water is turbid. There's lots of stuff in the water reflecting laser light. A stand of mangrove forests, and they're interpreting this as mangroves because you're getting a really bright return beneath them, and it's very flat, which is what you'd expect from still water under these kind of trees. As you move farther to the right, you see the ground elevation starts to vary. You get hills, you get little ridges. So that's more of a non-flooded forest. But that's totally cool. Here's another scene. This one is off the coast of Australia. You see the land, there's some trees at the coast. And then the coastline's right here. We have the surface of the water, some ocean waves. But you look faintly, you might be able to see the bathymetry through here. So you can tell how deep the water is. You can see that there's a mount right here. There's another ridge over here. And that's nine meters worth of water that we're seeing through. Further off to the right, you kind of get some hints of some other structure down here, which is a good 25 meters of water. That's 75 feet of water. That's a lot of water that we're seeing through. Super excited about it. So here's one more. This is the Bikini Atoll, some really clear Pacific water. Similar to picture to what we've seen before. You've got some ocean waves. You've got the land right here. So that's this peak showing up. And you can see the edge of the lagoon reef. This is 30 meters of water. That's almost 100 feet of water we're ranging through, from space, using individual photons. That's insane! It's just crazy. So when we started showing this data to some of the oceanography community, and they're super excited. Because measuring bathymetry close to coast is a pretty hard thing to do. You can't bring ships into each of these little lagoons and you can't bring them close to the coast so they can map deeper water more easily than they can map shallow water bathymetry. This is going to be an awesome tool for them. So for the cryosphere, we're designed to do cryosphere. So all of this is like gravy. This is all bonus. We're going to do cool things for forests, we're going to do cool things for oceans. What we're really interested in is sea ice and glaciers; glaciers and ice sheets. So here's a graphic of how this works for sea ice. So we talked about how we can measure sea ice extent pretty readily, but we can't measure the thickness very well. And this graphic didn't want to play. I'm going to try backing it up again. Boop! Let's see if it goes this time. But yeah, you can measure the extent with a camera, but measuring the thickness is harder. Getting this graphic to play is harder yet. [laughter] Hmm. I don't know why that doesn't want to go. That's goofy. But actually, we can just pause it right there and I can explain it from here. So we're going along with our six beams, we're measuring the height of everything, right? So here, you see some open water. And here's some sea ice that's sitting in that water. Ooh, here you go. I wonder if I can pause it right there. Ooh. Rats. [laughter] I'm going to get it yet. If this is my biggest problem today, I'm doing pretty well. So we'll take those two measurements. We have the height of the sea ice and the height of the water that it's sitting in. And somehow it got to a really good shot. I'm going to be patient. Okay, I'm not that patient. [laughter] But what we do to measure the sea ice thickness is just compare those two heights. Trying to get a stable picture that I can actually talk to. There we go. Cool. That's perfect. All right. So here's a crack in the sea ice, where you can see the ocean water sticking through. We're getting heights and elevations of the sea ice on either side of it, and you have the ocean water in between. So for sea ice science, what we want to do is compare those two heights. That tells us how high the sea ice is sticking up out of the water. The term for that is "free board," same as a free board on a boat, if you guys are sailors at all. But what it does for sea ice is since we know it's ice, we know the density of it, we know the density of the water, it lets us estimate the thickness of that sea ice. About 10% of sea ice sticks up out of the ocean water. So in this case, let's say the free board of the sea ice was about 30 centimeters. That means the total thickness of that sea ice is about 10 times that. It's about 3 meters. That lets us get at the third dimension of sea ice. We can measure the extent really readily from cameras, but getting the thickness has been a long-standing challenge for sea ice science. So with ICESat-2 data combined with those images, we really can put together a three-dimensional picture of the sea ice in the Arctic, in the Antarctic, and how it's changing through time. So here's an example of what data over sea ice look like. It's a little messier, but what you end up seeing are all of these ridges on sea ice. That's the sea ice elevation. But can you see these little flat spots through here? Those are cracks in the sea ice. That's where we're ranging down to the ocean in between them. So these are really small leads. Here's a scale bar for reference. It's about 70 kilometers, so this is a long stretch of data through the Arctic Ocean. But there's some larger leads over here and another big one over there. So in this case, we have the ocean water here, right about zero meters. And the sea ice here is about -- sticking up by about one meters. So that's actually really thick, probably old sea ice. It's on the order of ten meters thick, about ten times what you're measuring for the free board. So from a data processing standpoint, the challenge of sea ice data is to separate ocean elevation from sea ice elevation so you can do that difference. Over ice sheets, here's a stretch from Antarctica. Here's the Ross Sea down here, Ross Ice Shelf going up onto the plateau. Actually, this one's going the other way. It's going from the plateau down onto the ice sheet. Elevation-wise, ice sheets are kind of boring. They're pretty flat. They do have some typography to them, but not like the coast of Mexico and not like Australia, and not like sea ice. So this data often looks pretty flat at a short scale, but at a larger scale, you start to see mountains through there that we're ranging to. The blue line is ICESat-2 data. And this black line that's on top of it is data from another lesser satellite that we won't talk about. And you can see how this black line is far inferior to the super detail we're getting from ICESat-2. The data from along the black curve are totally missing this value right here, while we're picking up the bottom of the valley and the little peaks in the valley, as well. So that's totally cool. Here's another stretch where there's some crevasses or cracks in the ice sheet. These are pretty steep sides on these things, and we're picking up the slopes from the top all the way down to the bottom, some data off the bottom and back again. So we can actually measure crevasse depth, as well, which is cool. What we really want to do though is take data from ice sheets from the same place and compare it through time to see how it's getting bigger and how it's getting smaller. We think with the data quality we're seeing so far, we should be able to resolve elevation changes of about a centimeter, from space, using photons. It's -- I've worked on the project for about ten years, and you go through lots of hoops and reviews and things go well and things don't go well. But it's so gratifying to actually see it on orbit and collecting data like that. It's really amazing. And the data quality has been just outstanding. Knock on wood [knocking] [laughter] Cool. So ICESat-2 is in orbit now. Here's the first known sighting of ICESat-2. It might be a little bit hard to see. One of our colleagues in the lab, Jeremy, was down in Chile. He took a look at where we were supposed to be through time and he noticed that on October 25th at about 3:00 in the morning, we were going to be very close to where he was in Chile. And so he set up his camera on a tripod on about a 20 second exposure. And I can't even see it from my angle, but there should be a line right through here and that's ICESat-2. That's how you know it's not fake. [ Laughter ] You can go outside and see it go past, right? >> Somebody could draw a line of it. [ Laughter ] >> Dr. Tom Neumann: I encourage you to take a look at our website. You can see when we're going to be in your neighborhood, and you can look up at the right time. And if the sun catches the solar array, this is exactly what you'd see. >> The website is the heavens-above, you can see any satellite going over. >> Dr. Tom Neumann: That's right. And you can just type in ICESat-2 and it'll tell you when your next pass is. Absolutely. Cool. So the last section I'm going to talk about is, okay, great, we're measuring all these elevations from space. It's outstanding. Yada, yada, yada. How do you know they're right? How do we compare with something else to say that we're getting the right answer? So you might have noticed in the animation of the orbits that we go up to 88 north and 88 south. And all of our tracks converge right at 88 north and 88 south. It turns out that 88 north, of course, is in the middle of the Arctic Ocean, kind of a tough place to get to. 88 south, ironically enough, is much easier to get to. It's pretty close to the South Pole. It's about 220 kilometers away. And you can fly down to the South Pole. The U.S. has a research station right at the center there, which is what I did last year in December with three other colleagues. And we went out to this area to measure the elevation of the ice sheet with GPS. We flew into South Pole Station, we took two tract vehicles and drove out to this 88 south line of latitude and surveyed the elevation along this part of the arc right here. We had two high precision GPS instruments with us to create a reference dataset for ICESat-2. So when we're getting data from space, we can compare it with our GPS data to say, well, how are we doing? And having that independent dataset is a great check on the data. All of our tracks converge around this circle, so as it turns out, ICESat-2 will survey this little section at least once a day and sometimes twice a day, depending on the day. The whole circle was too far around for us to go. We were cold and wanted to go home. [laughter] So we only did this 300 kilometer stretch. Actually, that's not true. It was all part of the plan. But still. [laughter] We'll intersect 277 of the tracts, it's about 20% of the ICESat-2 tracks. And yeah, we're getting about one centimeter accuracy with our GPS survey, so it's a really solid dataset to compare with. There's a campaign, an airborne campaign, called Oppression IceBridge that flies a laser altimeter from an aircraft, and they've actually flown this circle several times. So we're able to take our ground data from the pink line, especially around this 88 south arc, and compare it with their data from an airplane, and then compare the airplane data to the data from space. So it's sort of a nested calibration approach. We just surveyed this little 300 kilometers, but if we use our 300 kilometers to prove out the aircraft data, then we can use that aircraft data to look at an even larger scale, all to assess ICESat-2. So here's what it looks like when you're cruising around at 88 south with 3 of your closest friends. We had tract vehicle here. We had a big long sled behind us that had pretty much everything we needed for the next couple weeks we were out there. We had two of these vehicles. You guys might be able to see these yellow blobs on here. Those are our tents. We just left them set up right on the surface. And this is pretty typical of what it looks like down there. It's pretty flat, right? You see some little bumps, little ridges here. These are little snowdrifts called sastrugi. But big picture, it's pretty flat. If you take a look at what the -- I think I got to go forward two -- what the data actually look like, here's a similar plot to what we saw before. Here's elevation on this axis, distance on that one. We went over this giant hill. It was like 300 feet high. But it was like 100 miles long. I mean, you would have no idea that there's a hill there. It's so subtle. [laughter] But when you look at the whole arc, all the way around 300 kilometers, this is the kind of typography you see. It's one of the flattest places on the ice sheet down there. The other experiment we left out there are called corner cubes. So comparing against our reference GPS data gives you a sense of elevation, how good are we doing in our elevation measurement. What these do is let us test out how we're doing in terms of assigning a latitude and a longitude. There's a little plastic cap we bought at Home Depot, drilled a hole in it, and there's a little piece of glass in there. Now, the facets and the edges on that piece of glass are optimized to reflect green light back to the space craft. When we put these poles in, we surveyed the location of these poles super accurately with our GPS to be good to about a centimeter. And these little pieces of glass in here, corner cube retro reflectors, bounces that light back to the space craft, and we should be able to pick these up in our data. So we should be able to uniquely say, oh, this is pole 209, and our data processing assigns it a latitude, a longitude, and an elevation. But we have an individual check, an independent check, on the latitude, the longitude, and the elevation of about 40 of these little poles that we left out in the middle of the ice sheet. So the GPS data by itself lets you check elevation, but surveying specific known targets lets you check the latitude and the longitude, as well. Another way you might think of doing it is hunting through our data until you find data that passes over this building. We know exactly where this building is. We know exactly how high it is. You can go into the ICESat-2 data and say, well, what did ICESat-2 think about the height of this building and the location of the edges of the building. This is a finer measurement than using a building, because this corner cube is about the size of your fingernail and we know its position to better than a centimeter. I'm not sure if you could get positions of GSA buildings to a centimeter. If they have it, they probably don't give it out to people, I would guess. Anyway, so yeah. That was a nice day to be out on the ice sheet, blue sky. Blue on the top. White on the bottom. Pretty typical of what it looks like down there. Yeah, so for NASA studying the cryosphere, the future really is here. We've got two new missions. ICESat-2 has a required lifetime, design lifetime of three years. We have fuel on board to go a whole lot longer than that. And so far, the data look really, really good. GRACE-Follow-On had some initial hiccoughs in getting their systems turned on. They had to switch from their A-side instruments to some of their B-side instruments, but they started collecting science data again about two weeks ago, about the same time that we did. And hopefully, they'll get their processing sorted out and start putting out some good data, as well. Our data from NASA, from both GRACE-Follow-On and ICESat-2, freely available to the public. We have a period where we do checkout and calibration, but after that, it goes to data centers that are open to the world. Our data center's called the National Snow and Ice Data Center. You didn't know that we had one of those, but we do. [laughter] It's in Boulder, Colorado, and if you're interested in cryospheric data, like that from ICESat or ICESat-2, you can get it there. Anybody can. Typically the time from date of data collection to the date that you can get it at the data center is about 40 days. We're still in our beginning phases. We just turned the laser on for the first time 35 days ago. So but by January, we should be into pretty normal operations. And all of you now are super excited about ICESat-2 and looking at your house from space. [laughter] You can do it there. National Snow and Ice Data Center. That's all I had, and I'm happy to take whatever questions people have. Thanks very much for coming. [ Applause ] >> Stephanie Marcus: Okay, the house rule is you ask the question, Tom will repeat it and answer it. >> Dr. Tom Neumann: Ooh. >> Stephanie Marcus: So we get it, you know, none of this, what did they say? What did they say? >> Dr. Tom Neumann: What did they say, yeah, right. Please. >> Audience Member: Thanks. Thank you very much for that presentation. I'm interested also in other ways of acquiring ground [inaudible] or -- and the work that you actually did are available to use earlier data from known points in the Antarctic and [inaudible] to check other places? >> Dr. Tom Neumann: Right, so the question was, if we're -- if I can paraphrase -- are we able to use data collected earlier from places like Greenland or Antarctica or other places to check the quality of the ICESat-2 data? And the answer is, sometimes. Greenland and Antarctica do change through time. We know that they're changing especially close to the edges. If you had data collected from the middle of the ice sheet, which is changing more slowly, then yes, we probably can use that. There are places on the earth that the rock elevation changes either slowly or in a known way. You get these steady uplift or subsidence. We can use that data as well. The piece I didn't mention about our 88 south traverse is that we'll repeat that survey every year that ICESat-2 is on orbit. So even if that middle part of the ice sheet is changing, and it probably is, we should be able to correct for that. That part of the ice sheet is one of the more stable parts. The elevation change there should be less than a centimeter a year. But we should be able to correct out for that. So you could use and you can use prior data, but you have to make sure you account for how the earth has changed between when the data was first collected and today. Yeah, in the back. >> Audience Member: So this is ICESat-2. What's the difference between it and ICESat-1? >> Dr. Tom Neumann: Question was, this is ICESat-2. What's the difference between it and ICESat-1. Whole bunch of good stuff. So ICESat-1 was on orbit from 2003 to 2009. It was a single beam altimeter, so it just had one spot. And its spot was pretty big. It was about 70 meters in diameter. That laser fired about 40 times a second, so you'd get a footprint, 70 meter footprint about every 150 meters along track. So you'd get one giant spot. Another 100 meters later, you'd get another giant spot. And so for sea ice that was really difficult to use because we couldn't detect those individual leads that are sometimes a lot smaller than 70 meters. And for ice sheets, what would happen is with the big footprints and the single track, we had a hard time correcting for slope effects. So imagine if this room wasn't flat, but it was tilted, and you're ranging to this tilted surface, and you're getting an average measurement over that slope. It was pretty hard to tease out what's just an effect of the surface slope and what's an effect of elevation change. That map of Greenland I showed at the outset that had the color codes along the edges, showing that there were changes at the edges, we were able to detect those changes because they were on the order of like, 10 meters a year. They were pretty big differences from each time ICESat would go pass. But with ICESat-2, we'll have these six beams that will let us take care of those slope effects. They'll let us find those leads much easier in sea ice, and should give us a way more refined picture of how things are changing. But we learned a ton from the first ICESat mission that really went into designing ICESat-2. Yeah? >> Audience Member: [inaudible] the photons that are transmitted and make it back? >> Dr. Tom Neumann: The question was what percent of the photons that are transmitted make it back? To round it off to the first ten significant figures, zero. [laughter] So we send out about 300 trillion photons in each pulse, and we get about one back. So it's about 1 part in 300 trillion. So it's -- yeah, you might say, oh! Well, all those wasted photons are going for nothing. [laughter] Which is true. But what happens is the beam spreads. There's some beam divergence to it. And so the beam gets larger and larger, it comes down. It's 17 meters when it hits the earth. The earth doesn't reflect all of the light back towards you. Some of the light's going to go left or right or forward or backward. Clouds scatter a lot of it, as well. And our field of view up in space is an 80 centimeter diameter telescope. So the field of view on the ground is about 45 meters. So we're only detecting photons that essentially go straight down and straight back again. A little to the left, a little to the right, and we'll never see them. But yeah, most of those are lost. Yeah, way in the back. >> Audience Member: If you're typically getting one or two or three photons counted, how many do you get on the corner cubes? >> Dr. Tom Neumann: Yeah. So those corner cubes, we don't know yet. I've poked around a little bit in the data looking for them, but it's kind of a needle in a haystack. Our geolocation, if you will, how precisely we can assign a latitude, a longitude, and a height, we're sort of on version 1.1 of that algorithm, and we keep tuning it. So I haven't found any data for it yet. We think they'll be on the order of about 20 or 25. So it should look, like, really bright, compared to everything around it. And I put them up on poles that were anywhere between a meter to two meters above the surface. So if you're zoomed in far enough, you should see the flat ground, and then this super bright thing two meters above the surface. And that'll be the corner cube. Yeah. Yeah? >> Audience Member: You made a very brief allusion to forestry. >> Dr. Tom Neumann: Yes. >> Audience Member: And from a climate change perspective, forest loss is probably pretty -- in the same order of magnitude as ice loss. >> Dr. Tom Neumann: Yeah. >> Audience Member: Is there any formal process put in place to have [inaudible] forests management people look at this data? >> Dr. Tom Neumann: Yeah, we actually do have a data product design that's the -- we call it vegetation canopy height. So we have algorithms in place from simulated data to detect the ground below the canopy, if we can, and then the top of the canopy and take that difference. And that should let terrestrial ecology folks get at biomass of forests in different places. We're not designed to do super well at that measurement. And the limitation for us is two-fold. One, if the canopy is really, really dense, like you might get in Brazil, some triple canopy rainforests, we just won't get any photons down through to the ground and back again. So we'll measure the top super well, but we won't know where the ground is. On the other end of the spectrum is a lone tree out in a field somewhere, you know? And the canopy isn't dense enough, you don't get enough data to say, all right, here's a reliable measurement of what the tree height is. We'll get the ground super well in that case, but not the tree height. So there's a sweet spot for us of forests like a deciduous forests, we should do just fine. And then building up data through time, leaf on, leaf off, maybe get a good measurement of when the leaves are on, of the top. You get a better measurement of the ground when leaves are off. You put those two together, now you're in business. >> Audience Member: Yeah, and then also, from a climate change perspective, is the aggregate loss of course, is crucially important and it looks like that data ought to be able to help them. >> Dr. Tom Neumann: Yeah. It -- [inaudible] absolutely. And we do have a couple of terrestrial ecology folks on our science team that are working towards exactly that. We didn't know how well we would do over forests for them, because we're not really optimized for it. It's the wrong wavelength and a few other differences that they'd prefer. But it's there and it's not going to be perfect, but it'll be helpful. Yeah? >> Audience Member: Are you able to get any data on the changes of the permafrost? >> Dr. Tom Neumann: The question was, are we able to get data on changes in the permafrost? Yeah, we certainly should. Lots of areas, permafrost is melting. And what you see when that permafrost is two things, the surface goes down because the ice that's in there melts and returns to the water table or to the ocean. And two, you get ponding. So we'll be able to measure the heights of those water bodies pretty well. We'll be able to tell when something has gone from wet to dry or dry to wet, in the case of melting. We don't have a data product specifically for permafrost, but we have one person on the team who's interested in that aspect, and that's kind of his main focus, is trying to measure permafrost changes. So we think we should be able to do something for them, as well. Yeah, in the back? >> Audience Member: Why are you guys using green light and not a different -- >> Dr. Tom Neumann: Different wavelength? >> Audience Member: What's the reasoning? >> Dr. Tom Neumann: Yeah. So the question is, why do we use green light as opposed to any other wavelength. The first ICESat used infrared light. It turns out infrared lasers are more efficient at taking electrical energy and turning it into light, which is a consideration. We ended up choosing green to optimize the overall energy efficiency of the whole design. So when you're making green laser light, you're essentially taking infrared light and doubling its frequency and you have a loss there, as well, a pretty big one. But on the detection side, the detectors in the green are about 15% efficient, whereas the detectors in infrared, A, weren't space qualified yet, and B, were about a half of 1% efficient. So we got about an order of magnitude or two better on the detection side, even though we had this penalty on the transmit side. But you put those two together, and it made green a better choice for us. Yeah? >> Audience Member: The National Weather Service uses about 8,000 volunteers to measure snow depth. >> Dr. Tom Neumann: Yeah. >> Audience Member: What's the implications of this for that measurement of snow depth? >> Dr. Tom Neumann: Yeah. So the question was the National -- National Weather Service? Yeah. Uses volunteers to measure snow depth all over the place. What can ICESat-2 do for them? It's a good question. So we have our six beams. They repeat every 91 days, so we should go through a place like Wisconsin, where I'm from, in the summer, and you get a good measurement of the surface without snow on it. And then we come through 91 days later, maybe it's October yes, it already started snowing there. So if you're there after a snowfall, you should be able to take those two elevation measurements, take the difference, and get the snow depth. It's just a height, and as I understand it, what a lot of their interested in is snow water equivalent, how much water is in there. But at least getting that height difference between snow on and snow off should be useful for them. There's a group called -- ooh, their mission is called SnowX and they're working to come up with a way to measure snow water equivalent from space for about a decade from now. So they're just at the beginning. So they're super interested in our data to get a map of what snow depth is like. A place like Wisconsin, it's relatively easy because it's farm fields, it's relatively flat. A place like the Rocky Mountains or Sierra Nevada, where there's steep slopes, is going to be a lot harder because you're pointing and your geolocation have to be that much better to know if you were -- on some steep slope, to know if you were here or you were here, because you'd say, okay, this snow pack is one meter or it's ten meters. Because just because that slope of that ground is so severe. Yeah. We're not optimized for it, but we should be able to help there, as well. At least in some circumstances. Cool. Anything else? >> Stephanie Marcus: Okay. Well, thank you for coming. >> Dr. Tom Neumann: Awesome. Thanks very much. [ Applause ]