Water whisperer
By Cat Hofacker|September 2023
Nadya Vinogradova Shiffer
★ Click at right to open this story’s Knowledge Guide entries.
★ Click star at right to open Knowledge Guide.
As a tempest of wildfires, floods and record-breaking temperatures swept across Earth in June and July, Nadya Vinogradova Shiffer and her colleagues at NASA and the French space agency CNES were preparing to begin science operations with a satellite that could help climate researchers predict which regions are at the greatest risk of those and other long-term impacts of our warming planet. Both agencies have satellites that survey the oceans, but SWOT, launched in December, will be the first to measure the cycling of water across Earth’s entire surface, including its oceans, lakes and rivers — hence the satellite’s full name, Surface Water Ocean Topography. Success of the three-year mission hinges on turning the interference patterns of reflected microwave radar signals into measurements of surface height, a technique never before applied in Earth observations. I reached Vinogradova Shiffer, the top scientist for SWOT, at her summer home in New Hampshire. We discussed the satellite’s brush with failure and the revelations SWOT could bring now that that problem is behind it. Here is our video conversation, compressed and lightly edited.
Q: Between the wildfires, floods and record temperatures, it feels like we’re getting a glimpse of the severe weather ahead if the planet continues to warm. What role will SWOT play in answering questions about what is happening on Earth and what can be done about it?
A: All water originates over the ocean, then travels through the atmosphere and sinks on land, where it’s being used or stored in lakes and rivers, or used for drinking water, agriculture, energy, you name it. Then, it comes back to the ocean through rivers. Any deviations from this perpetual cycle result in what we call hydrological extremes. If the supply of the water from the ocean is too large, you get storms and floods — which is what we have now in New England, where I am. Or if the ocean supply becomes interrupted and there’s a deficit, you get droughts, you get heatwaves; sometimes droughts could become big enough that they produce fires on land. SWOT is the first mission that allows you to look at this complete distribution, storage and cycling of all water in one mission, so you don’t have to stitch together different missions, like one over the ocean, one over land. That means consistent measurements; it means also that predictability is better. Think of the ocean as a warehouse of moisture. If you go to a grocery store and there are no strawberries, that’s too late to prepare, right? But if somebody had previously told you back at the warehouse that there is the shortage of strawberries, that’s lead time for you to prepare. So when we predict floods or droughts on land, it helps to see what happens with the sources of moisture before it travels or it doesn’t travel on land. So this complete look between the supply and demand chain, it’s very critical for prediction of those weather extremes.
Q: You alluded to the stitching together of measurements collected by different satellites. Why can’t today’s government and commercial remote sensing satellites do what SWOT is doing?
A: With SWOT, we kind of redesigned the way we look at water. For 30 years, NASA and international partners have been using altimeter technology, but the resolution of these measurements were not always useful for decision makers to address rapid changes we see at the coast. So at some point 20 years ago, there was a push to increase the resolution of water measurements. The idea was instead of just using a nadir altimeter that constantly emitted a single pulse, like a dot, as a satellite travels around the globe, what if we create a two-dimensional picture of the water with a wide swath in one scan? It’s like instead of painting with a little brush, you use a roller. The final product was the interferometry concept that SWOT is using. It has a large antenna with a boom, and then two antennas sending a microwave signal. The signals are a little bit out of sync because they travel different distances to come back to two different sides of the antenna. Because the signals are out of sync, there’s differences in phases and basic geometry that let you compute the height of the water column as a 2D swath. And because of the long boom, you have this big, 120-kilometer roller that you are making the swath with one pass. It does require a lot of stability. This antenna has to be stable to about 1 micron — thinner than a piece of my hair. So it’s not really a viable commercial endeavor as I would call it, because commercial satellites are useful, but there’s a difference between a bunch of bicycles versus a Rolls-Royce.
Q: Many of the previous altimeter satellites focused specifically on ocean measurements. Why were lakes and rivers left out?
A: Mainly because of the footprints. If it’s a more than 100-kilometer lake — major lakes in the U.S. or if the lake is big enough — we do have some inventory, and we have some lake scientists who keep track of the water storages. We also have, of course, other missions like Landsat that are helpful for land hydrology. But in terms of floods or water storage, Landsat gives you just the extent. It’s a map that looks like a photograph, which shows you “Oh, it flooded this direction,” but it doesn’t give you the height. So imagine a flood: It could extend this way, but it could be ankle deep, or it could be neck deep. SWOT would add this 2D view for water storage, for water management for dams or anything related to water, power, energy production, agriculture, storage. And with other altimeters, the limitation is mostly the resolution. Right now, there is a database of thousands of lakes. Now that we have 10x resolution with SWOT, we are hoping to have 5, 6 million lakes being observed. So it’s just now we can really focus on the smaller water bodies or water bodies that feed larger water bodies, etc., and the same with rivers. One hundred meters is what we’ve promised to the world, but based on our first results, I think we’re going to do better — 15, maybe 30 meters. In our first images, we saw a small channel in New York Harbor that’s 30 meters.
Q: It makes me think about the unknown unknowns — all of the things you might discover that you haven’t yet considered.
A: When we were launching SWOT, I thought about it in a similar way as our colleagues from the Planetary Science division that launched the James Webb Space Telescope. They’d been using Hubble since the ’90s, and it was amazing. Then we launched Webb and boom, “Is that a new exoplanet? Oh, what is that?” They also had some expected mission scientific goals, but then when you start doing it, you are open to new things that you did not plan. With SWOT’s improved resolution, it’s a similar feeling. One of the interesting questions that we want to discover is what happens with the Earth energy as a planet. Ninety percent of global warming is being absorbed by the ocean. We’ve never observed it before, but our models suggest that a big portion of the energy exchange and energy budget happens at smaller scales. And so if this is true, SWOT gives us an opportunity to kind of close this climate puzzle. Our models suggest that sometimes one-third of the work is done by those energetic eddies, the small features that constantly move in the ocean. We think of the ocean as big inert machinery, but actually it’s very, very turbulent. So what is the tipping point when we saturate the ocean with heat and it starts releasing it and kicking it back? It’s like boiling a teapot. It takes forever to boil, but then at some point it boils and the heat is released.
Q: So much of SWOT’s mission hinges on its star instrument, the Ka-band Radar Interferometer. Because the KaRIn instrument had never flown on a satellite before, how did that influence mission planning?
A: We thought, “What kind of a campaign do we need in order to maximize this 2D imaging of the ocean of the satellite versus 2D imaging of the ocean on the ground?” In other words, how do we make sure that what we are seeing from space is actually reality? So we were just trying to design a robust calibration/validation — cal/val — experiment. When we launched the observatory, it had this dedicated special orbit that’s a little bit lower than we were targeting for operations.
SWOT was temporarily placed in an 860-kilometer orbit so the satellite would pass the same areas each day at the same time during the six-month commissioning period. The multiple measurements collected in this “fast-sampling orbit” allowed the scientists to more quickly validate SWOT’s measurements against those taken by ground probes and other devices. — CH
And then for about six months, we’ve been flying in that lower orbit and collecting measurements every day over a certain location over the globe. It was what I called a “daily water show.” So every day we would have the view of the water, whether it is over the ocean or it’s a lake or a river or a coast. That allowed us to really fine-tune our measurement algorithms and our calibration system, so we would be more comfortable and ready to boost the satellite into its science orbit. The world community has really mobilized across the whole globe for our commissioning period by sending ships and airplanes and water instrumentation. So in addition to the NASA and CNES calibration/validation campaign, there were more than 50 field campaigns, self-mobilized in a coordinated manner. They were helping us by collecting ground data that helped us validate these novel measurements with SWOT that nobody has collected before.
Q: You had a scare during commissioning when KaRIn shut down for about two months. What was the cause?
A: The component that shut down is essentially a very sensitive circuit breaker. There are just so many electronics on the KaRIn instrument that the circuit breaker just blew, so we turned it off to protect the whole instrument. Then we did a series of tests to understand, “All right, is it really something going on? Or just by design, the system shuts down?” We tried to restart it. We tried to analyze a bunch of voltage experiments jointly with NASA and CNES engineers. What we finally ended up doing is switching to a similar unit on the spacecraft. NASA, in its wisdom, when we launch something new, we build two of them. So we had two KaRIns on board, and we’ve always treated side A and side B completely identical. So we decided to switch to side B and restart, and it paid off.
Q: That was a lucky break. So there are two of this specific component, or duplicates of every single component on KaRIn?
A: It’s not every single component of the observatory, but for the key subsystems, we have a unit A and unit B — two redundant units.
Q: You gotta love that redundancy. I know the extra weight makes it challenging when you’re designing —
A: Like imagine I ask you, “Hey, build me a nice Porsche, and you have to build two of them because you’re not sure!”
Q: Right. So you switched to unit B, and it seems like everything’s working as expected?
A: Yes, it was perfect. We’re still looking at the failure, but it is still a mystery to us. It could be just that we did not program it in the right way. We don’t know; we never came to the bottom of it. Also, because it happened during commissioning, we were restricted on what we could and couldn’t communicate about the mission. It probably looked worse than it actually was because NASA just goes quiet. We give no updates. We’d already taken our “first light” measurements on Jan. 25, but then KaRIn went dark. So do we release the first light image? It was a little bit of a delicate dance around what’s happening with SWOT, so we were sitting on the first light image for six weeks and not sleeping [laugh]. It’s like you deliver a baby, and then the baby was in the ICU [intensive care unit]. You waited for it, but you cannot take it home.
Q: Switching gears: I was really interested in the decision to release SWOT data much earlier to the public, instead of giving the science team first dibs on publication. Did the progressively strange weather we’ve experienced the last few years influence that decision at all?
A: This is a good question. SWOT was the catalyst for NASA to move to the cloud maybe about 10 years ago. At some point we realized that NASA Earth Science data streams were going to triple with SWOT on orbit, because it was going to be such a large volume of data — from 5 petabytes to like 15 petabytes. So there needed to be a paradigm shift, because people couldn’t just download the data on their laptops. So we’ve been going through these birthing pains — what is a better word?
Q: I like birthing pains since we talked about SWOT being in the ICU.
A: [laughs] Right, so how do you transition to the cloud? How do we give people access to this data? And then, step by step, we expanded toward this concept of open science. NASA has had open data since I can remember, meaning all the data is available, but what we decided to do is also make available the tools. We are creating new tools to work with the cloud anyway, so let’s just give people that. So we started thinking of giving the tools and the codes and how to handle the data, how to analyze, how to make a plot to publish it and so on. And since we are doing all that, let’s just also have free knowledge: If you publish a paper, make it open access. As program scientist, I made it a requirement. When people submit a proposal to us, they have to agree that they share their codes, they share their results. And of course I said, “Let’s create a framework where you are also rewarded for your authorship.” It’s not like people steal your work; they take it, and they reference your work. They acknowledge your work. When I was a scientist, somebody would ask me, “Oh, you did such a good job, can you just email me your script?” I would email the script, I would get no credit for it, and they would use my tool to write their paper with no attribution. But now, because I published my tool, now there is a reference. It’s a win-win situation. So in the spirit of open science, we thought, “All right, let’s deliver the data not in this science-ready format but in this pre-validated state.”
She means that scientists would process the SWOT data with the aforementioned tools to make it ready for publishing in papers in Nature and other scientific journals. — CH
We’re kind of expanding the number of cooks in the kitchen. Of course, not everybody wants to do that. But we are opening this process to those who are interested, as long as there’s understanding what is this data good for — it’s not yet ready for Nature papers, but it is ready to help us to improve the quality of the data products.
Q: As a scientist, are you at all conflicted about releasing information before it’s been extensively vetted?
A: I mean, it is a culture change, and it’s an experiment that we are going through. The goal is to be more open in the future. With SWOT, we are doing pre-data. I think with the next-generation missions that would launch in the next 10 years, NASA wants to share everything. But it is a process. How do you work out the kinks? How do you create a framework where you are both opening the process, but you also do not jeopardize credibility or the quality of your data product? There can’t be too many cooks in the kitchen; there has to be a robust process where the quality is vetted, and this is what we’re trying to figure out. What will work in practice? How do you stop people from taking precooked data and submitting it to Nature? You learn as you go, and we are doing it with SWOT. We are happy to provide lessons learned and figure out the kinks.
