Every weather forecast begins long before an icon appears on your phone. Behind each prediction lies a vast global system of satellites, sensors, data centers, supercomputers, and decades of scientific research working together to observe our living planet.

In this episode of the ALLATRA Podcast in collaboration with ALLATRA Global Research Center, Dr. Stephen M. Volz shares insights from decades of leadership in NASA and NOAA. As former Associate Director of NASA's Earth Science Flight Program and former Assistant Administrator of NOAA's National Environmental Satellite, Data, and Information Service (NESDIS), he helped build and oversee the systems that monitor our planet from space.

Drawing on decades of experience in NASA and NOAA, Dr. Volz explains how humanity has transformed its ability to observe Earth, from studying individual atoms and the origins of the universe to managing constellations of satellites that continuously monitor our planet. He shares why Earth should be understood not as separate oceans, atmosphere, and land, but as one interconnected living system.

The conversation explores how satellite observations become weather forecasts, why climate prediction is far more complex than it seems, how artificial intelligence is reshaping Earth observation, and why human expertise remains irreplaceable. Dr. Volz also discusses the critical role of oceans in Earth's climate system and why understanding our planet is one of humanity's greatest scientific challenges.

In this episode:

• How satellites observe Earth from space

• How weather forecasts are created

• Weather forecasting vs. climate prediction

• The role of oceans in Earth's climate

• AI in Earth observation: opportunities and limits

• Why human expertise still matters

• The future of Earth observation

The podcast is hosted on the ALLATRA platform.

This material is for informational and discussion purposes only. The statements, assessments, interpretations, conclusions, and hypotheses contained herein represent the personal views of their authors as of the date of publication and do not constitute an official position of the ALLATRA or any other public association, organization, project, partner, affiliated entity, or individual unless explicitly stated otherwise.

Publication of this material is intended to encourage open scientific dialogue and does not in itself imply institutional endorsement, confirmation, or scientific verification of the views expressed. The material should not be interpreted as an official statement, expert opinion, or professional advice.

Full Podcast Transcript

Dr. Stephen Volz

When I got to NASA, we started building satellites to observing the whole universe. My first mission was actually studying the Big Bang.

Valeriya

Wow.

Dr. Stephen Volz

The Earth has been shouting at us for a while. We're not listening. We don't understand the ocean as well as we do stars a thousand light years away. We are primary actors in world change today. Humans. Not hurricanes, not solar storms. It's us. Eight of the hottest years on record were in the last ten years. There's definitely a signal there. It's not random.

A one degree temperature rise in the ocean is equivalent to in terms of energy to 100 degree temperature rise in the atmosphere. You're building this juggernaut of power, of energy that's going to take a thousand years to dissipate. There is no away anymore. As you know, the plastics in the ocean, the biosphere, the CO2 in the atmosphere and the ocean, the pollution everywhere. It's here and it's not going away. Understanding the planet is critical to surviving the life in a changing world.

Voiceover

Dr. Stephen Voltz, a physicist and distinguished leader in earth observation, satellite systems and environmental prediction. Over the course of his career at NASA and NOAA, Dr. Volz has helped guide the systems that allow humanity to observe our planet from space, transforming satellite measurements into the data behind weather forecasts, climate monitoring, ocean observation, disaster warnings and space weather prediction.

From satellites in orbit to the forecasts that help communities prepare for hurricanes, floods, wildfires, extreme heat and solar storms, Dr. Volz has been deeply involved in the infrastructure that helps society see climate and environmental threats before they become disasters.

Valeriya

Doctor Volz, it's our great pleasure to have you today with us.

Dr. Stephen Volz

It's great, it's a delight to be here as well. And thank you for asking me.

Valeriya

Your journey is extraordinary. From condensed matter physics to building and managing NASA Earth Science Flight Program, to leading NOAA satellite and information service and serving as acting Assistant Secretary for environmental observation and predictions. You spent decades building and overseeing the systems that show us our planet's pulse. Now, from the vantage point of a long and extraordinary career, what has Earth become to you personally, and how has this journey changed the way you understand the planet itself?

Dr. Stephen Volz

Well thank you. It's a heck of a question to cover all of that at once. What I would think is when I look at it from looking backwards, it's sort of starting with the small and going to the large. When I was a researcher in physics in graduate school, I was studying quantum mechanics and microscopic, atom-by-atom physics of helium at low temperature. And as I moved on in my… so it can't get much smaller than that. I guess you could get a little bit, but that's as small as you can get. But as I started studying other events, when I got to NASA, we started building satellites to observe the whole universe, so we went… My first mission was actually studying the Big Bang, the origins of the universe, and the cosmic microwave background. So you go from individual atoms to everything, the Big Bang.
I think in my career, I've been sort of riding the wave of technology, of space technology as it's gone from individual Apollo Day, kind of individual astronauts and spaceships, to whole constellations of observing systems that are monitoring everything. And what I've seen is I've gone through that as I've learned and watched individual atoms to the universe, to different elements of Earth, I really started to learn how they're all connected, how watching, looking at the Еarth through different lenses, through, you know, an imager of just pictures or through something that looks at the temperature of the planet, or you can see the radiation coming out from it with a different kind of camera.

It's really remarkable as how when I look at it, I don't see all these different things. I don't see the oceans and atmosphere and land. I see an interconnected biosphere of everything working together. And by studying all these things separately, as the manager of the constellation of earth observing satellites, I see how all these different perspectives fit together. How looking at a hurricane from the ground, from space, from an airplane, you see it in different angles and different perspectives. You see the whole thing much better. So you see, it's really the connectedness that I see as much as I see the complexity of the planet, but how all the different elements of it are really co-dependent and  co-mingled so that you can't understand it without looking at all of it. So it's really just a big picture, but it's with a thousand different pixels that are all necessary to understand how they  work together to create the world we live in.

Valeriya

How do you see the planet for yourself? Like, how is it like in your mind? How do you…

Dr. Stephen Volz

I see it as a living, breathing organism of all the different elements working together, sometimes not aware of each other, but all necessary for life together. I see the evolution… from the seasons to the planet over thousands of years. How it changes, how humans are impacting the planet. I see that in most of the data we have, how we are probably the primary actors in world change today. Humans, not hurricanes, not solar storms, not other fires. It's us who are affecting. And I really see our agency and the action and the future of the planet. So I really do see and feel my own ownership and responsibility for what we as humans  what we're doing to the planet and to life on the planet and a responsibility to see that we do, we treat it well and we treat it carefully.

Valeriya

That's true. We really carry a huge, huge responsibility for our planet for our environment. And that would be great that we will understand more about this responsibility, right?

For most of us, a weather forecast is just an icon on a phone screen. But behind that little icon is an invisible universe of satellites, data centers, and supercomputer models. Can you walk us through that hidden chain? What has to happen, like step by step, before the phone can say to us that it's going to be rain tomorrow?

Dr. Stephen Volz

Sure. Yeah. You're right,  when you think of a weather forecast, it's two dimensional. You look at it on your camera or on your phone, you might look at it on TV, but the generation of that involves thousands and thousands of people working in many different disciplines. It also involves decades of work and research, which gets us to the understanding that so we can take an observation and turn it into an information that you can use. So we use the term numerical weather forecasting, numerical weather models, which is basically a mathematical model of the atmosphere of the planet. Think about it. Every atom, every piece of air in the planet is moving at any time. And you want to understand how. A weather forecast, it is just how the air from here is going to move there, is it going to be hotter or colder or is it going to be rainy or not? You're seeing how these things move in time. To understand how they move in time, you have to know where they are today. And you have to know what's driving, what's forcing them to move. Things move because they're driven by energy, by forces. You see high pressures and low pressures. So a high pressure pushes air out. A low pressure pulls air in.

So you see air moving from one to the other. By measuring the air density, the air temperature, pressure and humidity all around the world at once with dozens of satellites —by the US, by Europeans, by the Japanese, multiple nations— by these satellites, download the data. They take an observation from 900km in orbit, or from a geostationary from 36,000km, they take a picture. The picture has a certain resolution in. It sort of takes a cube of the atmosphere, resolution of the air in that particular part, so you fit all these pieces together that gives you what we call an initial state, which is what the Earth is today. And then you apply those years of research: how does air, how does energy move through the atmosphere? How does wind move over clouds or over mountains, over the ocean? What happens if water is pulled out for rain? It changes the temperature of the air. All of these things are little processes that occur, and you model those with physics and with chemistry and with modeling. And you put all this together into a global numerical model, which requires a supercomputer because it's a lot of data in real time. And then you say, what's going to happen to this little cube of air over the next week? So this one kilometer cube of air is going to move this way.

It's going to move up, it's going to move to the left. And that is really forecasting. You're seeing how that air is moving, and that's the dynamic of wind and atmospheric movement. And it really is. You need that global because what's happening today in the Gulf of Alaska is driving what's going to be weather here in the Chesapeake Bay five days from now. If it's a high pressure dome there, it's pushing air down there. If it's low pressure, it might be pulling and it would change the wind direction. So you need that global picture all at once. And then the models and the researchers who spent generations and PhD programs and graduate students and others over the years to pull all this together. So it's an immense amount of community that pulls all this, makes all this happen. And it's your weather forecaster on the evening news who is just taking these data feeds and communicating with the National Weather Service or with researchers. It's a really incredibly dynamic dance that all these players do together, to provide weather forecasts on a daily and hourly basis.

Valeriya

Wow, that's really a huge work. And really, like you said, decades of research and work. So it's, well…

Dr. Stephen Volz

We’re constantly trying to improve too. And we always look, “Did I get it right? Did I get it wrong if I predicted it and it didn't happen, why not?” So we go back and say, “So what was different?” And we correct our models as we do that or we missed an observation wrong. Or the calibration of an instrument was off and we had to correct that. So constantly we‘re re-correcting or revisiting to improve it. And that's why forecasts have gotten incredibly much better. 20 years ago, a ten day forecast, a three day forecast is as accurate as our seven day forecast is today. So we've more than doubled the length of time for which we can predict. So you're actually now… you'll say it's going to rain at . It's not that it's going to rain tomorrow. It's going to rain at . And it'll be gone by . And you'll find it's pretty good. So you can actually, “I'm going to be walking from 6 to . I don't need an umbrella, but I might need one if I go if I'm delayed.” And we are that accurate now. Whereas 20 years ago you'd carry an umbrella if it might rain and it would be, you know, you'd look at it differently. So it's, we're constantly improving. But there's obviously more that we can learn, but it's an iterative process. And there's a lot of lessons learned that we take in every day from what we observe and what we do.

Valeriya

Now a lot of people will finally understand what kind of work stays behind when we all open the phone just and see that, “Okay, I can plan my day according to the weather,” right?  So that's ...

Dr. Stephen Volz

It takes an effort of many, many people and long time frames and lots of investment by governments and by private companies and the like to make this happen. It's not something that's easy to build, and it's something we should be treating very carefully because it's a very important asset for all of us.

Valeriya

Of course. Definitely. And it's just the only one simple thing that we say about the planning the day, but that's a lot of things when it can save lives for people, the heat and everything.

Now, if we talk about the satellites, when a satellite looks at Earth, what is [it] really seeing: light, heat, chemistry, motion? What is it really capturing up there? And once those raw signals reach the ground, what happens inside the NOAA system to turn them into meaning?

Dr. Stephen Volz

So what a satellite.. we call it remote sensing. We're basically it's remote, your distance from the object that you're looking at, and it's sensing something. So remote sensing satellites are basically looking for photons which come up and they measure the effect of photons. There are some that look at gravitational forces, but mostly it's photons coming from the planet. A photon is just a unit of energy radiated from the planet. Now it can be radiated by a bunch of different things. So basically every instrument is some kind of photon detector. It's like your camera is a photon detector. It measures incoming photons in different spectrums, which will be different colors. And you generate a color picture based on its red, a red photon or a yellow or blue photon.

Similarly, in space observations, you're measuring radiation coming from the planet, coming from a spot on the planet which has a certain spectral characteristic, and you then take those… And so we have something called geospatial reference. So, it has a location. If you're looking at imagery, the planet, it's, you know, geography, it's the two dimensional planet. If you're looking for the atmosphere, it has three dimensions. It has, you know, X and Y, which is the land, but also altitude. So you have a three dimensional space that you're trying to monitor and look at. For the atmosphere you obviously have to know not just the bottom of the atmosphere but the whole column. So you measure these emissions that come from the planet. Mostly they're passive, which means the planet is always radiating its 270 degrees Kelvin or Celsius.

So, and space is cold so it is radiating heat to the world, to the universe. We're measuring that radiated heat, and we're looking at very small variations like one Celsius or half a Celsius variation in a piece of atmosphere from another one. By measuring that, absolutely, you can see temperature changes, temperature variations. We take all of those in, and on spacecraft we convert those into digital measurements: zeros and ones. So that you have a datum from an observation: would be a spatial location, time that it was taken, an intensity of the zero and one. And depending on which measurement you're making, it would be a type of observation you have, whether it's the colors or the light that you have. Those are all integrated onto a spacecraft and then beamed down in data packets to the ground, usually from a geostationary satellite, which is orbiting at the same rate the Earth orbits. It's constantly in contact with the same  ground, but a low earth orbit satellite is rotating as the earth spins under it, so it beams down data when it goes over a ground station. Those data then are taken from these antennas spaced around the planet — and there are hundreds and hundreds of weather satellites that do this — and then brought together into data centers where you take these zeros and ones, and you convert them into information. So a zero and one doesn't tell you if it's temperature or humidity or radiation or something. You convert it into what we call engineering units. You convert it into a temperature, a pressure, humidity, and a geospatial location.

So then you have a map, a grid of information which has physical content of what's going on in the atmosphere or the land or the water. Those then are mapped or put into, as I mentioned before, those models which are pre-made to take in data from.. with geospatial reference points and temporal stamps. Because if I make a measurement of air over Kansas at , and air over Washington at ,  it's not the same picture. Since the air is moving, it would be blurred. So you have to make sure the timestamps are the same too. If you want to have an accurate picture… like if you take a picture of you moving, the picture from one half of the lens would be different from the other half because you moved. So you get a blurring, so you don't want that to happen. You take all this together. You put all the data from dozens and dozens of satellites, integrate these into what we call interoperable, so they're intermingled now into an information content which is much richer. That goes into models then, which then goes into the different forecasts. And the forecast could be “is it going to rain?” It could be “what's the probability of a flood?” It could be “what the height of the tide is going to be” because we measure ocean altimetry, the height of the oceans, all these different parameters can go into different models and, and hundreds of different information outputs that go to users around the world on a regular basis. And all this happens with weather 24/7, as we say, 24/7, 365. forever, every day, every hour.

So that the information flow is gigabytes of data happening from dozens of satellites. You never see this, but it's in the background. It's in the systems that are going through all the time. When I worked at NOAA, we had a reliability requirement to not be down for more than 30 minutes ever. So our satellites have to be working, the ground stations working, the data systems working without interruption, because an interruption would be catastrophic to somebody who's relying on storm information or your airplane flying around a hurricane or things like that. So that's how we go from radiation measurements to actual information content that people use. And the data are not done once the weather is done either, they go into storage, they go into an archive, which are then used by people who want to do climate modeling or climate studies, or what's it like for months at a time. They're trying to do more integrated studies of the evolution of crop formation and, you know, crop productivity and humidity in the ground and the soil. Do we have droughts coming or are they relieved? Things like that. There is a lot more information than just is it going to rain or is it going to be sunny. Because all of this feeds into that living, breathing system that I mentioned earlier, where all of the what.. how you feel today depends on what happened yesterday, and it's going to influence what's going to happen tomorrow. And we have to have all that tied together too.

Valeriya

Of course, and for climate models, it's very, very important because it's like the archived information that you can basically predict the future. And what are the observations we most cannot afford to lose, the ones that, if interrupted, would immediately make society less safe.

Dr. Stephen Volz

Well, the one that jumps out most immediately to that question would be I mentioned two types of satellites we have. One is a low earth orbit, low earth orbit, we call it, or Leo, which basically orbits the earth about 14 times a day and gives us that global picture of the planet. The geostationary ones are different, and we call them GOES: the Geostationary Operational Environmental satellites. They actually sit over a spot over the planet. And we have we, the world, have about a dozen of those spaced around the planet, around the equator. So as the Еarth rotates, if this is the satellite and here's the Еarth, as the Еarth rotates, the satellite rotates at the same speed. So the satellite orbits at 36,000 km, but it's always over the same spot. The benefit of that it is looking at what's happening right now. So if there's a tornado that's in your neighborhood or is coming your way, I don't rely on a low earth orbit satellite, which might pass over in a couple of hours to spot it. I'm watching it with that geostationary satellite, I'm able to do extreme event updates in real time observations of crises as they occur, a hurricane as it makes landfall, a fire that is changing and growing, a dust storm or haboob or a derecho, which is moving through at 50 km an hour.

You can track it in real time. And we have examples where we talk, we talk with emergency managers, with emergency services to say it's you have ten minutes before the eyewall, the hurricane gets to you, move now. So that kind of if we lose that geostationary visibility, we lose that real time observational capability. And we would be blind basically for significant events when they occur.
A second effect would be in low earth orbit. If we were to lose some of our atmospheric pressure and temperature measurements, it would degrade the models, the forecast models. You wouldn't notice that right away if I took out a third of all the satellites in orbit, I'd still have a forecast and just I wouldn't be able to tell you if it would rain at 6 o’clock. I'd say sometime between 4 and 7. The accuracy would be reduced. But you still have a forecast. So that's why the geostationary is so critical because there's no backup to those if one goes out. I mean, not immediately. We can bring another one on board quickly. But they are the sole source staring at a real event, which can happen in minutes or seconds. And you want to have that real time.

Valeriya

Interesting. I don't think a lot of people really understand, I mean, the role of that. We often hear about long term climate models. Can you please explain for the general audience, what is the difference between a weather forecast, a seasonal forecast and a long-term climate prediction?

Dr. Stephen Volz

The time is an obvious one, right? But the time.. the measurements are the same, the climate forecast models are built on the observations we take to do weather forecasts. The accuracy required to measure temperature and pressure that goes into a numerical weather model is the same kind of precision that you would use to do a climate forecast. When a climate model does it generates that, it is not looking to see whether it's going to rain here on Thursday in 2029. It's going to look at… it takes these sorts of these narrow observations and generalizes them to will the pattern of rain change over time? So it's taken the same models, the same information, but it's looking for more longer term effects that come into how you would interpret then the… So I would call it the human factors that build into the changing of the environment. So what drives the weather today? It's radiation from the sun. It's the amount of radiation we keep in the atmosphere. It's the radiation balance basically, it drives the long-term climate forecasts. And things can disrupt that radiation, that heat balance of the planet. In historical days, there were outbreaks of massive outbreaks of volcanoes, which put a bunch of dust in the atmosphere, which changed the albedo, the reflectivity of the earth, which changed the energy balance, which brought on ice ages, for example.

Greenhouse gases right now are causing some changes in our energy balance in the earth now, so that if you model what the greenhouse gases will be based on measurements, but based on projections of human action, how much carbon, CO2, we're going to generate, how much fuel we burn, how we change our energy usage will lead to different future states of atmospheric carbon dioxide density, which leads to different climate outcomes. So the models basically are the same. It's the physical processes are the same, the observations, the land and we're not going to change the land or the ocean largely. What changes is the feedback mechanisms that are driven by the things that we do in the atmospheric content. We're projecting, the sun's not going to change dramatically. It hasn't for thousands and thousands of years. But what is changing is the atmosphere, content. And we can then use our projections of what is likely to be the case and how that will affect the energy balance which will affect our climate. Second to that is land use change that comes into the climate models. The amount of land that's taken for farming changes the energy balance and how much moisture is held in the soil. So that changes the water cycle. These are things that have to be modeled to give you different projections. So when you see a climate forecast which has this distribution of possible outcomes, it's based on different assumptions about how people will change our behavior and how the atmosphere will respond to those changes.

So all of these are projected, and we could change a lot by going away from carbon fuel right away, it would be a big change in the atmosphere, CO2 content. Or not mitigating methane production would have another big change, which would change the atmospheric… the radiation balance. So those are the why you see in climate models, much more variation over a 20, 30, 50 year timeframe than we see in weather models over the 3 to 5 day or ten day forecast. Because the underlying parameters of the balance of the planet are highly affected by what we do and what we don't do in our life on earth, our humans, how we behave. But it's the same physics parameters. It's the same interactions between different elements, the same observations. And from a climate perspective, we look longer in history to see how we might have seen climate changes in the past so that informs some of these feedback mechanisms I mentioned for long term projections into the future. But the observations are the same. We're not doing a climate measurement or a weather measurement. We're doing an earth measurement, and it goes into different models in different ways.

Valeriya

And if we talk about these basically the temperature records. So 2024, 2025, and now 2026, the temperature records, especially in the oceans, have been striking. Some have said that the data looks like it's deviating from what models predicted. You know this observing system better than almost anyone. When you see those numbers, do you feel that the earth is telling us something that we're still not hearing clearly? And if so, what do you think it's trying to say to us?

Dr. Stephen Volz

The earth has been shouting at us for a while, and I think we are not listening, or we don't want to hear it sometimes. Or we don't want to admit it. We don't want to take action on what we're being told. When you think, go back to the question. I observed that it's a question of energy balance, which leads to changes in our climate warmer or colder.
So what you're seeing in the last recent years, like ten of the eight of the hottest years on record were in the last ten years. I mean that as a mathematician, as a scientist, you see that as a trend. There's definitely a signal there. It's not random. It's not variation, random variation. There's a trend that's occurring.

And when I mentioned before about climate models there's a variation of outputs of and I'm mimicking here the temperature variation in the future. It's going to be very hot or not quite so hot. And that spread, that uncertainty in the forecast depends on input assumptions.

And as you mentioned, invariably the rate of heat warmth increases greater than we had than our mid-range projections had said. The models are not wrong, but the range of models are, the models that predict that it would not be as bad as it is have been invalidated because the systems, the effects are stronger.

We didn't maybe understand all the feedback mechanisms. The ocean for example, the ocean is… The amount of heat in the ocean compared to the amount of heat in the atmosphere is like three orders of magnitude greater. The ocean can absorb an incredible amount of heat but it absorbs it slowly. So one degree temperature rise in the ocean is equivalent to in terms of energy to 100 degree temperature rise in the atmosphere.
Atmosphere heats up really fast and cools down fast. The ocean doesn't heat up fast. It doesn't cool down fast either. So when you look at the temperature in the ocean, it's going.. you're building this juggernaut of power, of energy that's going to take a thousand years to dissipate. If we can dissipate. The atmosphere will cool off really relatively quickly. But the ocean's been storing this energy and it's possible… And what you see is one of those feedback mechanisms, say, what happens when you heat up water? If you have a pot of water on the stove, what do you see? You see vapor coming off. So as we heat up the ocean more, it's going to generate more humidity. It's going to generate more water vapor. It's going to put more water vapor into the atmosphere.

It's going to be more rain, but you're going to get these feedback loops. And that's what you're seeing is that the change in one parameter is changing the balance that it would have had in the past for other parameters. We're definitely seeing an increasing trend in heat. The ocean is generating is now hotter, that's why we're seeing temperature sea level rise is mostly expansion. It's not ice melting, it's the water expanding as it gets warmer. So two thirds of every inch of sea level rise is because the ocean is warmer now. And another point is we don't understand the ocean as well as we do the atmosphere. We don't understand the ocean as well as we do stars a thousand light years away. It's very hard to get down into the bottom of the ocean, understand what's going on. Observations are difficult, so the more and better we understand what's going on in the ocean, the better we'll be able to understand the future trajectory of our planet and our climate. But it's shouting at us. The planet is. And the signals are not going away, and they're just going to be getting more and more glaring. And the impacts are going to be felt in everything we do.

Valeriya

I truly agree with you about that point. And that's really the earth shouting for a while. And that would be great that we will stop ignoring that. Right? And finally will do more, more and what we can.

Dr. Stephen Volz

There are  Things we can do. There are ways to, you know, mitigation is certainly within our power. It's just, it's not easy, but it takes thought and it takes a recognition that it is necessary. And I think that's the part that we're struggling with right now. We as collectively, individually, some have definitely agreed, but not everybody and not enough to make concerted action possible.

Valeriya

Doctor Voltz, you watched weather prediction evolve across three decades, from physics-based models running early supercomputers to today's AI systems that can generate a forecast in seconds without our solving an equation of motion. You've seen both worlds from the inside. In your opinion, what can AI genuinely transform in Earth observation and prediction? And what, if anything, must remain human or physical, or simply beyond the reach of a machine trained on the past data?

Dr. Stephen Volz

So AI is an incredibly powerful tool. It is not a replacement for thought and for fundamental process understanding. So the reason.. so we are seeing great leaps forward in the ability, as you mentioned, the speed with which AI… And I'm not a software engineer, but what I would call it is a correlation engine. It can, it sees the correlation of different observed phenomena and then projects how the next, the next probability of an occurrence is going to be. So it's not saying why high pressure forces air to move from one place to another. It's saying, when I get high pressure here and I get low pressure here, the air moves this way. It doesn’t understand why, it just says it does. And if you have enough data that has been demonstrating that it creates these probabilistic correlation models, which say this is what's going to happen.

And they can be very powerful for short-term forecasts, and they can generate forecasts much faster than running a supercomputer on a physics-based process numerical model, which is very labor-intensive, very time and computationally intensive. So AI will be very helpful in updating minute by minute, hour by hour models and forecasts. It will not replace physics-based analyses and understanding of what's going on. And what it won't do It will not mean… it will not predict something that we don't understand or that we haven't observed. So where the human comes in and says and sees, you know, their own experience, and we are in a sense, we are correlation machines as well.

We remember what happened before and we remember how things seem to go together, if we're observers of weather and our environment.  So what an AI will not do, will not take that leap and say, aha, this is why that's happening. And because I understand why it's happening I can understand another thing that I wouldn't have expected.
An AI might, will never see that, take that leap of knowledge and say, “because I've seen this correlation and now can infer something completely different.”
So that's why I think that what we do with.. what we have found in observation science and in observing the earth is when you bring in a new measurement, a new look at something from a different angle, it looks very… you understand it in a very different way. So when you start observing for example, gravit… We have a satellite which measures the gravitational constant of the planet. It measures gravitational distortions. And it's not radiation. It's looking at actual mass distribution on the planet. And that was a completely different way to understand the environment. And we found we were able to measure underwater aquifers by looking at the mass change. When you pump water out of the ground and you use it and it distributes the mass of the water changes and the gravitational constant changes. That's a whole different way of looking at water movement. And it allows us to start seeing how mass movements of water are actually affecting the way that the ocean currents move and, and the way that the land is subsiding.

These are things which were not predicted, well, maybe predicted in some way, but not anticipated. And it would not have happened without that new observation.

So that's where I think the human ability to ask the “what if" question, which AI can't do, how about or what about if I try something new will always be there and which actually will lead to the breakthroughs of conceptual understanding, not just predictions of the near-term changes that will occur. So I think it's a great tool. It's necessary to take advantage of it, but it certainly won't replace observations ever. And it won't replace humans in the loop who are there to interpret what you see.  Because an AI will say it's going to do something, and a human could say, that's a stupid forecast. It obviously won't happen, but AI won't know that.

You see lots of examples in other cases where AI has produced something which how they're trained can distort what they predict. And the other thing I would say is every AI is only as good as the data it's trained on, and the data are the ones that make it possible. So the observations of history and of ongoing are necessary for AI to be even as capable as they are. So there's no replacement for what we do. These are great augmentations though.

Valeriya

Definitely. I definitely agree. And, it's a great tool, like you said, we really need to use it. But at the same time, the biggest, the brightest mind, the brain, is in people. And that's what needs to, of course keep and, you know.

Dr. Stephen Volz

There's, I'm sorry to interrupt. I'll give you an example, an analogy, not an analogy, but an example. I had a conversation with somebody who was at the Department of Commerce, the weather service. He said, “With a good model and a good dashboard of information, I can handle what you now have going on with weather forecasters all over the Midwest. I don't need to be sitting in a weather tower in Lincoln, Nebraska, to be able to see if there's going to be a storm and to tell people in Lincoln, Nebraska, there's going to be a storm. I can do that from New York. I can do that from wherever. As long as I have the data and a dashboard and the information.” That is an ignorant statement, because it's not just that, yes, I could do the same model. The model is true here or there, but unless you know the history of Lincoln and the people who lived there and how they respond to storms, and that it's two in the afternoon, which means there's traffic in this part of the town and not that part of town, your forecast is not going to be used... It's not going to be as useful if you say it's going to be a storm, go to shelter right away. If you're in the middle of a school transition, some kids are in their school buses.

That's not the answer you give them. You say shelter in school, don't go home.

You have to know the environment you're working in and the people and the communities that you're working with for the information you provide to be useful. And AI can't do that single, you know, Uber dashboards with great information are only… they provide information, but not, it doesn't make it usable unless you understand the context with which people are using it.

So you've got to have people in the field. You've got to understand the community you're working with and you're working for, and you have to have that. You have to listen in order to understand what you need to say. And that's the part that's missed. I think when a lot of these pushes towards efficiency and innovation, I can replace that with a machine. The machine will not know the people. The people will not trust the machine. And if you want to change people's lives and have a good outcome, you need to be able to talk to them in ways that they will react to. They will act in the right way to protect themselves. And just the best data is not the best, will not produce the best outcome unless you understand the people you're trying to talk to.

Valeriya

Yeah. And there are definitely different conditions in different areas. Otherwise, like you can't just each area just count as just equal, let's say same conditions or something like that. Thank you so much.  And that was I think it's a good explanation to the people, so people also understand that it's not always like everything…. Because, you know, a lot of people are using like ChatGPT and they go and they start doing this themselves and trying. So it's a really good explanation. So they also understand that, you know, because there are different areas, some areas are more affected for fire, some more affected for floods. So you can't treat it the same, right.

Dr. Stephen Volz

Some people have cars and some people have mass transit and some people don't have buses. So the same forecast which says seek shelter will be different block by block almost. So you have to have that local understanding to be able to service, to provide a service that people can react to.

Valeriya

Doctor Volz, you have argued that weather data is public good. Why does public access matter so much when lives are at stake?

Dr. Stephen Volz

I start with the assumption that we're all part of a community and that what affects one of us affects all of us. So when significant weather and environmental events take place, whether it's a flood, a fire or a hurricane or a heat wave or something, or a cold snap or the like, everybody in the community is affected.

And if the information that people use to make life-saving decisions, do I, you know, “Do I have to evacuate? How do I evacuate? Do I put a kerosene heater in my kitchen or not?” These decisions have life or death consequences on an individual basis.

And they should be informed by, as part of our culture, our society informed by the best information that we as a people can provide to each other. And I think we've decided as a country and many, most countries have, that it is in the best interest of the people to provide that information in a way that they can make decisions. They can help themselves take action in the case of emergencies.

If you put different levels of information behind paywalls, where you have to pay to get that information, you end up with circumstances where some people have the right information, have good information, some people have poor information.
It's almost like when NOAA created the NOAA Weather Radio and we everybody didn't have radios. So we distributed radios. So people had them so they would get the forecast. If you don't have the information available and you don't make it publicly available, people will not know how to behave or how to react.

And yes, those individuals will be damaged as real or will die. There'll be deaths and we all suffer from that. There's no benefit to the rest of us if some people weren't smart enough or capable enough to get the information. It is our responsibility to see to it that they have sort of a life-saving public information content.

So when I extract that, to extrapolate that to what we do with the weather products and weather information, if the information we collect is privately owned, then it is not available to, it is restricted in its use, it's restricted in its application. And yes, some people would make profit by selling it to you, but if you can't afford it, you won't use it. You'll go to someplace else where you get free data, which may not be as good, or you'll get no data and you'll be suffering with ignorance.

And I think there's a certain level of information content that we've decided should be available to the public. And that's where it is available from a public source, from a public provider, and not through a contract or a vendor who is selling or doing it under contract to the government.

So it starts with that fundamental principle that we all benefit if all of us are taken care of and we all suffer if any of us suffer. And that's an important part of public service and public benefits that we provide.

Valeriya

That's where we come into the question of how much important  the information itself, the value of the information.

Dr. Stephen Volz

Information is as essential as Internet access these days. It used be water and sanitation, even in some areas, has become sort of regionalized or localized or privatized. But yes, information is the lifeblood of survival in today's world. And we've 100 years ago, that wouldn't have been the case, but it is now.

And recognizing that is not as we would call in the government requirements creep. We're not encroaching on the private sector's ability to do something. It's a real realization that people need a certain level of information awareness to be, to survive and to, to live in the world today.

And we are all part of the country and the citizens. We need to provide that and not just to the country, but to the world. The world. Not everybody has access to satellite data. And I think it's our responsibility, as those who do, to see to it that the world is elevated, and is given the same information that we have, so they can make decisions as well.

Valeriya

You've also spoken about the connection between Earth observation and human health. For many people, satellites still feel distant and very abstract, something far above us. Can you explain how satellite measurements can eventually become a health decision on the ground? For example, protecting people from extreme heat, wildfire smoke, poor air quality, flooding, changing ecosystems, etc..

Dr. Stephen Volz

It already is. And most people, as you said, are not aware of the infrastructure that leads to the information they deal with on a regular basis. And take for example, a heat wave. What when we will get warnings like last week, we had a whole, whole East and this last couple of weeks, the whole East Coast, most of the United States is under heat, heat warnings because of predictions, projections of temperatures over 100°F for sustained periods of time and not being. And warm and hot at night as well.

And that has led to major efforts to mitigate the potential impacts of that which creating cooling centers, making sure people don't plan to be outside. You know, people are adapting their lives based on the intimate knowledge that's been blasted to them by TVs and everything else, appropriately, that there's a major heat wave coming.

That didn't come because AI Claude said there's going to be a heat wave or ChatGPT said. It became because we have weather models and observations and we have forecast models which say this kind of heat dome is being created. It's going to persist for weeks and it's going to be over this period of time.

So that's a clear example of satellite data and satellite-based weather models affecting your daily activities for weeks at a time, weeks ahead of and on an hour to hour, day to day basis.

And that's just the obvious one on weather. Another example might be air quality. A couple of years or two ago, the fires that were occurring in Canada, because of dry air, dry conditions and fires were creating terrible air quality all along the eastern seaboard of the United States from Maine all the way down to the Carolinas.
And that was an example where you could actually image those geostationary satellites I mentioned earlier, whereas imaging fires occurring, you could see the smoke coming up, the wind and models showing the wind trajectory would say, it's going to be down here along the East Coast on Thursday. It's going to be here on Friday, here on Saturday here, be ready for that.

That's another example where the clear air forecast and the air quality warnings to people who had asthma or said, don't go outside, wear a mask, etc., was based on intimate access and ready access to satellite data and the models that they feed to show the trajectory of this information, of these environmental impacts that are coming along. We see that.

I'll give another example: water quality, something called algae. Algae grows when warm water with lots of nutrients. You've probably seen the news of the reflecting pool on the mall when it gets green from algae.

Well, that's basically just warm water nitrogen runoff and hot air and algae just blooms and grows. That happens on whole seacoasts at times and in fresh water supplies. You can see that from space. You can see the precursors of it and you can project its growth. And then you can anticipate how you have to change your water intake plants that don't draw water from that resource because it's going to have algae in it, take it from that resource, or start rationing water. And cities and municipalities do that on a regular basis based on water forecasts, based on satellite data.

And one last example I'll give sort of climate based as we're seeing the winters get warmer and the late frost and the early thaws occur earlier in the year,  later in the year for frost and earlier in the year for thaws. Insect. Insects are surviving more. So you're seeing the migration north of what had been tropical issues like Zika and West Nile virus that would not survive because the mosquitoes couldn't make it and couldn't survive through the winters and the like. You're seeing them migrate north. So now we're able to forecast that, say, in the next five years, there will be an increase in Zika and West Nile virus in West Virginia, where it never existed 20 years ago.

That's something that you don't just flip a switch and say, okay, I can deal with it. You have to get entire health systems ready for. What does Zika look like as opposed to the flu? It's a different way to deal. I mean, one is a disease, a vector-borne disease, which can be catastrophic to some people. The other one is just a cold. The  symptoms may look similar but slightly different, training health communities to be able to know the difference now so that they can anticipate that.

And knowing it five years before it's going to happen is an important part of advance warning that climate models and earth observation satellite data are all feeding to support, along with health and human services activities. These are all examples of how understanding the planet is critical to surviving life in a changing world, and environments that are changing in ways that your local history won't help you because it never occurred in your domain before.

There are multiple other examples of a similar nature. As the world changes, our local history does not reflect it is no longer applicable. Past performance is not indicative of future behavior, and we have to be. And understanding the Earth as a system allows us to anticipate that and to start preparing for it decades ahead of schedule.

Valeriya

Thank you so much. I think that is honestly even more not obvious for a lot of people. How much is that affecting the health system and how much is that really gives the health departments, those who work on that to understand, and predict it and be ready within like some years in advance.

Dr. Stephen Volz

Yeah, right.

Valeriya

Wow. So how important is international data exchange for science itself? And what would be lost in forecast accuracy in model quality if that flow of data were interrupted?

Dr. Stephen Volz

The  weather community is probably one of the most globally integrated communities in humanity. It's something that the World Meteorological Organization, the WMO, has been around for decades, many, many years, and the weather community recognized exactly what you just said,—that weather ignores boundaries and national boundaries for 150 years. So we have been working together in the weather community as an integrated community, understanding that we're all in this together and we have to share the data and we have to share observations for a long time.
With the advent of remote sensing and satellite data, in particular, the last 60 years, the global nature has become even more obvious. And the US and other nations through the WMO have established and maintained a whole network of data sharing expectations, requirements, capabilities, which allow all of us to share data seamlessly, rapidly, with no time delay and with no cost to each other so that we all benefit from that common data and information source.

And just as an example from the US side, as a NOAA lead on the satellites, intimately aware of our dependency on European and Asian and Japanese and Korean data into our models, for every observation we put into that global consortium we get two back. So in terms of co-development and co-investment, we're getting a three fold three times whatever we put into it. We're getting that back on our own use.

And so if we tried to do this without the partners, sure, we could fly our own satellites and do everything ourselves. It would cost 3 or 4 times as much, and we would not benefit from the expertise, the intellectual capacity, the capabilities of all of our international partners, just as they benefit from us.

So we all recognize that we have what we call a friendly competition. We all try to improve on each other to do better models, to do better satellites. But we all recognize that we're learning. What worked in European models here, hey, we're going to try that and we'll make it work on ours, but we're going to tweak it a little better, make it better here. And they're going to learn from us. So that friendly and competitive nature, we all are improving the system and working well together to make it happen.

Again, another argument for keeping this in the public domain: many countries don't even have a private sector, and many countries don't share their public data with other nations. But in weather they do, because they recognize, they see the value of it, and they see the return on investment to their own people, their own nations from doing it. And actually, weather has been a good example for environmental data writ large, in the larger context, which leads to climate modeling and land use changes.

So the earth observation data, not just weather, has generally become an open data source for the world. And we've all benefited from that by rapid access to a global collection of data, which makes us all smarter, not just in what's happening over there, but in how to interpret what's happening here as well. So it really is, I think, the best example of  humanity working together to take best advantage of what we all need to know what's going on a daily basis.

Valeriya

I was about to say that   this is something that really can unite us all and unite the efforts, understanding that basically, do we want that or no, but we need each other. And it's I really like that example that you said that, basically it's a friendly competition and that we can learn and it's something that remains open, you know, like for everybody, it's very, very interesting.

Dr. Stephen Volz

But it's not inevitable. It's not necessary. It doesn't, it's not guaranteed. There are nation states that have arguments with each other. The US doesn't use Chinese data in our models. I don't know. I don't know how the Chinese use our data. But when there are competitions between nations, you have restrictions and desires to pull back information sharing. And in some areas for national security purposes, you can see where I'm not going to be tracking some of these airplanes and wars and things like that, but we should resist that for flowing over into weather data.

I know weather is strategic in terms of some conflicts, but it's still so important. If you start chipping away at this collective nature, this collective agreement, we have to share our observational data. It could fall apart if we're not careful, if we're not aware of the value of it, and we fight for the public and the common sharing of the data.

So it's not something that should be assumed in the given. We have to highlight how it's important and make sure that we don't go away from it because somebody has a different idea.

Valeriya

Now if we talk about space weather. So space weather begins 150,000,000 km away on the surface of the sun. A solar storm begins with the flare of the sun, hurls plasma across space, and days later hits Earth's magnetic field, scrambling GPS, threatening power grids, endangering satellites.

You spend years overseeing the systems that watch for this. My question is simple: if a massive solar storm were coming, are we actually ready for that?

Dr. Stephen Volz

Massive. Depends on what you consider massive.

Valeriya

Fair.

Dr. Stephen Volz

We have a good system to observe it right now. We do not have a system that forecasts a solar storm.

So we can look at forecasting a hurricane. We know how to do that. We can see a little bit of a storm cell over the Western Africa and we know with models and everything else, it's going to grow into a hurricane. And we can do forecasts of that.

We don't know what causes those solar flares. We know they're correlated with sunspots and magnetic… We have magnetic models of the sun. So we don't forecast a solar storm as much as we see it when it happens.

And we get a few minutes, a few hours to a few days warning time because the light gets here in 9 minutes, the solar, the flare, the CME, the coronal mass ejection may take days or hours to get to us. So we can project that far in advance.

So are we ready for a big one?  We can see it coming. We can project what it's going to do to the magnetic field of the planet and what it will do to surface currents on the earth, which is what causes power problems. Our systems are slowly adapting to be ready for that.

And so we're aware of the risk. It is not easy to insulate us from the most massive storms.

And there have been some big ones. You probably heard of the term Carrington storm in 18, I think it was 59, which caused aurora down at the equator and had telegraph lines burning out all over the planet because of the induced currents it created.

That kind of solar flare would probably cost trillions of dollars worth of damage, in our infrastructure, because you don't build every transformer on every city block to deal with an event like that. But they're very rare.

What we are more concerned about is as we become more technology dependent our satellites, our GPS satellites, our communication satellites, our remote sensing satellites, can they handle a major storm?

Most public satellites, like the ones that I would have built at NOAA, are made to be hardened against massive storms. So. And when we have a flare warning, we will go into what we call a safe mode. We'll turn the satellite off so that the system's not running where it's more vulnerable to a storm, but it'll be in a safe hold mode so it can recover slowly.

Not every commercial satellite has that, so they may not be protected against it.

It costs more money to do that. It costs more mass. So large systems of the commercial sector will not survive storms, whereas a public or a federal system might.

So we're partially protected. The most important systems are built to be radiation-hardened.

The most vulnerable, and infrastructure, like the power grids, are being designed and slowly being redesigned to be de-integratable. One of the problems that can occur under a massive solar storm is that you overload one section of the power grid, and it pulls down other sections. When you have a storm occur or when it's coming, you can have them decouple so that I may lose one power grid, but it won't pull down the next one because they'll be running autonomously on separate sections rather than all integrated like they do normally.

That's one way to mitigate the impact.

So yes, we have ways to protect ourselves. A massive storm of Carrington or a larger event which occurred like 150 years ago, would be hard. It would definitely have a big impact on us.

But I think we've done good diligence in understanding where we're vulnerable. We're still in the process of trying to protect ourselves against those vulnerabilities being realized under a big storm.

Valeriya

You spent a lifetime observing this planet, its pulse, its rhythm, its changes. Looking ahead, what areas of Earth observation or prediction do you believe are most critical to improve right now? And where is the next breakthrough most urgently needed? The one that could genuinely help us better understand and respond to what's happening to the planet.

Dr. Stephen Volz

So I think that we're in the middle of an information revolution right now. I'm not talking about AI. I'm talking about massive processing and computer systems and the ability to bring data together and work with it.

I think the biggest near-term change and without changing observations or information is the integration of information.

You've heard of the term the Internet of things.  That's just an example of data everywhere, but not data integrated into a common data system where you can actually use it together.

So we have something called the FAIR Principles. You probably... it’s a practice where you try and make data interoperable. The FAIR Principles are called findable, accessible, interoperable, and reusable.

But the idea is that data collected for one purpose, like a weather forecast, is made available for another purpose, like a crop forecast or an economic forecast of future water requirements.

So I think the biggest innovation that we are on the verge of realizing, if we do take the time and effort, is to bring the data together so that we can do these, collect these conceptions, these consistent integrated models of the Earth, and not just Earth observations, socioeconomic data, energy use data, people, humanity data, health data.

So if I want to know if I know a heat wave is coming, I can say these are the neighborhoods that are vulnerable because they have people who can't move or they don't have air conditioners.

That kind of information is available, but it's not integrated. You have to go through multiple steps to bring that together.

I think the integration of all this will make it possible to be much more proactive and responsive and protective of the communities and of the people.

Now, the greatest change is coming forward. I think there's one area I mentioned, I alluded to earlier where we underobserved, and that is in the oceans, the impact of what we're doing above ground and in the air is being felt in the oceans.

And the ability of the oceans to absorb our sins, so to speak, our pollution, our carbon dioxide, our heat is waning, is being used up, but also is being unobserved. So we don't know what's going on in the deep oceans and in the oceans overall.

And the amount of energy, power and biology in the oceans is immense and undersampled and underobserved.

So I think if and when, as we get more sounder probes that go down into ocean profiling, more, deep ocean observations like the ones that are along the Gulf Stream current - these things and then integrating these into our global climate system models will be, I think, transformative in the way we understand the future trajectory of our planet.

And I think it'll be pretty alarming because of our… We have not realized how much we've relied on the ocean to be that infinite sink, as I called it, for our sins, but for our our refuse, our throwaway, our throwaway heat, our throwaway pollution, our throwaway CO2, all of that stuff that we generated and not worried about - it's been absorbed by the oceans.
There is no away anymore. You can't… you throw something - it's still here and it's coming back. The plastics, as you know, the plastics in the ocean, the biosphere, the CO2 in the atmosphere and the ocean, the pollution everywhere. It's here and it's not going away.

We've been counting on the oceans. I think the oceans are the place where we need to observe better and understand better.  And that will change the way that we see our future life on this planet because of the way that we have used that resource ignorantly and blindly for most of the last couple hundred years, as we've been living on the surface.

Valeriya

That's for sure. Thank you so much for sharing. And that was a great pleasure talking to you today. Thank you really for your time, for a lot of explanation. And I'm sure we're gonna see a lot of comments with gratitude for explaining how those systems are really working. And, maybe you have something very important that you wanted to share with the people that you probably didn't say today?

Dr. Stephen Volz

I would just say one comment I would make is that you keep saying, since I know this better than many, and I've done this for so many years. I am just one of thousands who have been doing this for so many years. I'm reflecting the information collected and generated and integrated by so many dedicated and intelligent and thoughtful and passionate engineers, scientists, human resource, sociologists who see information and try to figure out how to use it to improve life on our planet. So, and those are the people, many in public service, but many in the private sector as well, in the academic world, who really are just trying to understand the way things work and, and how we can live better on the planet.

And I really want to give a shout out to all of them. As you mentioned, people don't recognize how much they depend on the efforts of so many other people who are not doing it just because they're trying to be helpful. They're doing it because they like to do it and they love it. But we all benefit from the passion and the investment of time and effort by people like yourself in what you're doing here. Your interest in asking these questions is something you like to do, but it also helps us all get smarter and better.

So it is really that network of invested people working towards common objectives, which I'm just thrilled to be a lead on, a piece of. And I've spent my life being able to participate in it.

Valeriya

And thank you so much for that. And of course, we truly appreciate every single person who is  really doing all the best and also acting out of their curiosity.

We talk a lot with different scientists, researchers, and it's definitely… You understand and the society understand more the role of the science and how much we really need that, and how much is it important, what the role it plays in our life and in our future and the future of this planet.

So we truly appreciate your work over three decades, you know, and we're sure you will continue your work because I know scientists they're the ones who’re never retired.

Dr. Stephen Volz

So I'm retired, but I'm not finished.

Valeriya

Exactly. That's what is really  true. So…

Dr. Stephen Volz

Retired from federal service, but I'm still very active in the Earth observations community.

Valeriya

Yeah. That's what I was trying to say is that if you're retired from the… I don't know, governmental work or any other work, it doesn't mean that you're retired generally.

Dr. Stephen Volz

That's right. Yeah.

Valeriya

So we wish you all the best. And we truly, truly appreciate all your actions and all those years and the years that are waiting ahead, for sure. And we truly believe and hope for those amazing and important scientific breakthroughs on our paths and, I mean, as the scientific community, in our shared humanity and our shared planet.

Dr. Stephen Volz

Yeah. Thank you. It's been, it's been a great conversation. Thank you, Valerya.

Valeriya

Thank you.