Continued space airborne science nerdery under the cut because it got kinda long. Spoiler: Here be lasers So. Laser altimeters. Fundamentally, they work on the same principle as sonar or radar: send out a pulse, measure the time it takes the reflection of the pulse to come back, calculate a distance to the target. Then, if you know where the laser is and where it's pointing to very high precision, you can calculate the position of the thing you hit with the laser (ideally the ground). We fly at an altitude of 9-12km, depending on the aircraft, and the laser sweeps back and forth across the ground to scan a ~2km wide swath. Like a lot of laser altimeters, LVIS is a 1064nm laser (near-IR). It fires 1500 times a second and normally has a footprint on the ground that's about 25m in diameter. The reflected light comes back in through a telescope and goes through a photodiode detector and a digitizer that samples the electrical signal from the detector at 1GHz (1 billion samples per second), which translates to a distance resolution of about 1.5cm at the altitudes we usually fly at. The LVIS website has some diagrams and pretty pictures. LVIS can be flown on just about any plane that's been modified to install a window in the floor. In Chile, we had it on the Gulfstream V owned by the National Center for Atmospheric Research. The GV is a long-range jet, which we needed for the 10-11 hour flights over Antarctica. (It's also pretty comfortable.) Here in Gabon, the flights are going to be a max of 5 hours, so we're on a smaller prop plane (a King Air B200 owned by Langley Research Center). Prepping for a trip starts a couple months before we leave. The engineers who work on the laser and the detectors are constantly changing things and upgrading bits of the instrument, so there are extensive lab tests that take place over a few weeks to make sure we iron out any errors in the pointing of the laser, characterize the performance of the detectors, etc. Then the instrument gets packed up and shipped to wherever the plane is for installation. Our heroic lead engineer and usually at least one other person fly out and install the instrument a couple weeks before the mission. Then they do a couple test flights. Usually they fly over the flattest water they can find and roll and pitch the plane all over the place. Back at GSFC, we analyze the test flight data - the roll and pitch maneuvers over the flat water allow us to calibrate and calculate all the correction terms we need to compensate for the distance between the instrument center and the plane center, or the fact that the instrument might not be mounted quite straight. It's also a last chance to catch any problems before they show up in the field. Then the engineers come home and get drunk, and it's transit time! The GV transit to Chile took two days. Two of our team went along with the pilots and they collected data the entire way. The B200 has a short range - too short to make the Atlantic crossing. It took the two pilots 5 days to work their way up through Canada and Greenland and then down overland to Gabon. All the scientists took commercial flights and met the plane here. Our team consists of the five people who work with the laser and the data (the project lead, the project scientist, the lead engineer, the data tech, and the data analyst (me)), the two pilots, and the 3-man plane crew. There's also a logistics person who deals with the planning and the diplomatic stuff and the Gabonese contacts. Next time: what I do all day during a deployment! And probably more than you wanted or needed to know about waveforms and inertial navigation! Flight day tomorrow, sleep now.
Flight today was cancelled because thunderstorms. While I'm waiting for food, you lovely people get an extended ramble about the weather. Spoiler: Sorry about this Gabon has a short wet season from (roughly) October to November, a short dry season from December to January, a long wet season from February to April, and then a long dry season from May to September. During the wet seasons, Gabon is in the Intertropical Convergence Zone (known as the doldrums at sea). The ITCZ is at the boundary of the two huge convection cells called Hadley cells that wrap around the Earth, one in the northern hemisphere and one in the southern. Hadley cells give rise to the trade winds near the equator and the subtropical jet streams. The cells circulate in opposite directions, and the ITCZ is where both of them are rising. If you have a lot of warm moist air rising, you should expect some thunderstorms, and indeed the ITCZ is visible from space as a band of clouds with heavy thunderstorm activity. Like so. Unfortunately for us, the inherent instability of the ITCZ means that the weather is damn near impossible to predict more than a few hours in advance. Thunderstorms just...appear. We have a whole bunch of data from three of the best weather models available and they're not much better than a guess. Satellite imagery is much more useful - if you want to see the kind of thing we're working with, here's the EUMETSAT imagery viewer - play with it a bit, it's neat! Under "RGB Composites", the Convection tab traces thunderstorms in yellow (times are GMT). But because we're reliant on satellite imagery to make the decision about whether to fly or not, we can't make that call and give the pilots the flight plan so they can file it with ATC until the morning of the flight, and even then we may have to cancel at the last minute. The pilots are very tolerant of all this nonsense. Low non-thunderstorm clouds are a problem too, because our laser doesn't penetrate them. Those are a little more scattered on the good days and we don't lose too much data to them. In addition to the clouds, there is a thick haze over most of the country, especially in the northern part, caused by condensation on dust particles (I guess it would be called a "wet haze"?). The haze means that if you look out the plane window, according to the team members who fly, you can't even see the ground. Fortunately, our laser DOES go through the haze, as does the hyperspectral camera. We found this out on the first flight day and we were very relieved.
question about laser altimetry: how do you determine the position of the plane with that much exactitude?
Good question! The short answer is that we combine GPS data with data from an inertial navigation system, which allows us to refine the trajectory from the GPS. The long answer: The plane has a GPS antenna, obviously. We also set up a GPS base station at the airport so we can use differential GPS. (It's my job to haul the GPS receivers out to the antenna every day and hook them up.) This is more useful in Gabon than it was in Chile - the Antarctic flights took us so far from the base station that dGPS didn't really help. Standard GPS can resolve a position to 3m or so. Differential GPS and some other tricks can bring that down to <15cm. But that's not good enough for our purposes. Plus, we also need to know the orientation of the laser when it fires, and we need to know it a lot more precisely than the 1 degree the plane's sensors measure. The inertial measurement unit (IMU) is a device consisting of 3 gyros and 3 accelerometers that measure changes in velocity and orientation. We have 3 - ours are black boxes about 5 inches on a side mounted on the laser assembly. For each time step, the IMU measures the change in each of the 3 gyro angles and the acceleration in the x, y, and z directions. In a perfect world, you could integrate the acceleration to get the velocity and the velocity to get the position of the IMU (not the same as the position of the plane or the laser - I'll get to that in a second), and you could integrate the changes in the gyro angles to get the orientation. It's a fancy computerized way of doing dead reckoning, basically. Unfortunately, we don't live in a perfect world. The IMU has a finite sampling interval and because it's digital, not analog, it also measures in discrete steps. It also inevitably has some internal biases just because it's not manufactured perfectly. The good thing about IMUs is that they can sample faster than the GPS, they measure orientation, and they still work even if you lose your GPS signal. The bad thing is that an IMU by itself drifts off course in minutes. The way to fix this is to combine IMU and GPS data. The GPS position, because it doesn't drift with time, can be used to correct the INS drift, and the INS can provide more precise position and orientation measurements and also fill in any gaps in the GPS data due to GPS signal loss. The mathematical algorithm that does this combination is called a Kalman filter. This is a pretty decent explanation of what a Kalman filter does. When combining the GPS and IMU data, though, you have to also keep in mind that the GPS antenna, the IMU, and the laser are all in different places and facing different directions. We can roughly measure the lever arms (distances between the 3 objects) and the mounting angles, but to get the kind of precision we need, we have to calculate them from the elevation data. That's what the test flights where they do roll and pitch calibration maneuvers over the water allow us to do - knowing that the surface we're looking at should be flat(-ish), if we see large "waves" when the plane rolls and pitches, it means we haven't gotten the biases right. There's an absolute monstrosity of a program (it's almost as old as I am and written in Fortran) that solves for the lever arms and mounting angles and several other variables that are required to make the flat water FLAT.
thank you for the detailed answer! i love this sort of thing, the clever solutions people come up with for these kinds of problems.
ok here is a total speculation science fictiony thing i've been pondering on lately. in an episode of 'agents of shield', there was some kind of satellite imaging that let them look 'miles' into the earth in search of buried structures. they were vague about what kind of imaging. now obviously that's not real, mcu is science fiction and they have all kinds of impossible tech (helicarriers! woo!). but what's on my mind is... if you were going to invent such a thing, what would you use? sonar and seismographs give a really smeary image, and i don't think sonar can go deep enough, though if you do it right seismography can image all the way to the mantle. but in a super blurry way, like you can detect a magma pocket a few miles across, not access tunnels for a buried city. obviously visible light's a no-go, but i wonder if some kind of clever fuckery could be done with longer wavelengths. neutrinos pass through the earth with no problems, but they also pass through your detector.
Nah, that just makes it so you don't need as many magnets or as big magnets because you can just use the existing magnetic field and manipulate it a bit instead of having to generate the field from scratch. *nods sagely*
huh, then in a science fiction context at least, you maybe COULD do it with satellites -- you don't have to build a donut, you just have to have satellites ping the magnetic field. you wouldn't get realtime of the whole earth at once without a donut, but you could get slow features like glaciers and stuff no problem.
Science! (speciality: maths and science, but all branches of science are absolutely wonderful and I wish I found this thread earlier)
Ooooh, science, my specialty! Several random biology related facts that may or may not have come up before: 1. Don't these orchids look like bees? They sure do. They also smell like bees. How do we know? Because they've evolved to be pollinated by horny male bees on the search for a lady. They see/smell these flowers, go "AHA! I FOUND ONE!" and then desperately attempt to have sex with it for a few minutes before flying off in frustration, covered in pollen. Nature is a beautiful thing. 2. Australia has tree leeches. Yes you read that right. So, my intro Bio professor used to do a study-abroad tour that went to Australia. At one point, he was out doing field work with an undergrad when, unbeknownst to him, a tree leach dropped off of its strand of mucus onto his head. Totally unaware of its presence, due to the fact their saliva contains a paralytic agent, he and the student got back into the car, with the student driving. A little bit into the ride, the leech decided it was done and let go. So, head wounds already bleed a lot. Now imagine one that just had an anticoagulant put on it. Basically, this is what went down for the poor student: There he is, in the car with his prof, when suddenly there's a veritable FOUNTAIN of blood and he looks over and there's this ominous black THING inching toward him. Frantically, the student pulls over to the side of the road and runs shrieking off into the Bush. (Everything turned out ok in the end, so the mental image is really funny.)
A cool science thing that combines my love for human geography and human genetics is the evolution of lactose tolerance, or lactase persistence, in European vs East African and Middle Eastern populations! In the 2 geographic areas, the mutation that allows lactase persistence is different. It's really interesting to me, because you have these 2 completely separate populations just beginning to domesticate milk-producing animals and realizing that hey, this stuff makes baby animals fat, which is something that I would really like to happen to me and my babies; and the mutation that brings about the ability to take advantage of this handy source of fat and protein develops completely differently in the 2 populations of humans but with very similar selective pressure, aka Real Fucken High selective pressure, because there's a renewable source of nutrition that takes way less effort and resources to get than raising and slaughtering animals and the ability to use it really helps out survival chances. So the 2 mutations both had the same effect and were both selectively favored, giving 33% of the population the ability to digest lactose. And it's development is pretty much confined to the pastoral populations that relied on their livestock for their livelihood, aka most of Europe, East Africa, and the Middle East, whereas in areas such as most of China where milk wasn't a Big Deal, the mutation was neutral, which is why a lot of Chinese people are lactose intolerant!
Ladies and gentlemen and those not falling into either category, let me tell you about LAGEOS! LAGEOS is a satellite that was launched 40 years ago last week. It is a sphere 60 cm in diameter covered in 426 reflectors. It has no power or moving parts and it's been in the same stable orbit since its launch. It is basically a space disco ball. But what we can do with the space disco ball is bounce lasers off it and thus calculate the distance to it with very high precision. The precise distances from ground sites around the world to LAGEOS can be used to make measurements of Earth's gravitational field. It's been used to map the movements of tectonic plates and to measure frame dragging, one of the predictions of general relativity. LAGEOS, its clone LAGEOS-2, and a constellation of similar satellites are expected to be in orbit for literal millions of years. The continents will shift significantly before these things de-orbit. Human civilization could fall and the new intelligent species descended from raccoons could measure tectonic plate movement with the same satellites. (Assuming they could find them. GSFC is one of the places that generates orbit predictions for LAGEOS so we don't lose it.) More info
It's a medium earth orbit, lower than geosynchronous at ~5900km. I'll see if I can dig up some more orbit details tonight.
ALL HAIL THE DISCO ORB no but for real, thats cool on a lot of levels! thank you, local space nerd :3
