 I'd like to thank everyone for joining us today at our briefing, Innovations in Weather Forecasting for a Changing Climate. I'm Dan Bercetto, I'm the president of the Environmental and Energy Study Institute. I'd like to start by thanking Representative Paul Tonko in his office and the Sustainable Energy and Environment Coalition for helping us with him today. ESI is celebrating 40 years of advancing climate solutions through congressional education. ESI was founded by a bipartisan group with members of Congress. And since 1984, we have worked to provide science-based information about environmental, energy, and climate change topics to policymakers and the public. What do educational resources look like? Well, they look a lot like this. We do lots and lots and lots of briefings. Our first briefing of the year was about the National Climate Assessment, the fifth National Climate Assessment. Two weeks ago, we were in this very room talking about DOE's Earth Shots Initiative with Deputy Energy Secretary David Turk. We have briefings coming up on the appropriations and budget process, ocean carbon dioxide removal, and even more. If you'd like to follow what we're up to when it comes to briefings, I encourage everyone to sign up for our bi-weekly newsletter, Climate Change Solutions. It's the best way to do that. You can RSVP for everything online at www.esi.org. But we do more than briefings. We do fact sheets. We do articles. We do podcasts. We do issue briefs. We do a lot of different stuff. And Climate Change Solutions, that bi-weekly newsletter, is really the best way to keep up with stuff. We try really hard to make sure that our educational resources are timely, relevant, accessible, and practical. We put a lot of thought into this because we know what it's like to be a congressional staff person and to have your boss show up late in the day. Everyone's trying to leave. Someone's got a plane to board. Hey, what's this climate change topic? Tell me about advanced weather forecasting. Tell me about carbon dioxide removal. Tell me about this or that. And I want you to not just Google search for what your boss is asking you about. I want you to Google search it and EESI and come to us and use one of our resources to answer those tough questions about climate change topics. We know that the Farm Bill at some point will be before us. And we have lots of really great resources, including a hearing tracker that goes back several years, tracking every climate-related hearing that the agriculture committees have held. We have side-by-side-by-side comparison charts ready to go to help you compare House and Senate marks when they come out. We try to really be there before you even know you need those resources so that you can rely on us. And briefings are part of what we do. And we're really happy to have all of you here today. But today we're here to talk about advanced weather forecasting and climate change. Climate change multiplies and intensifies acute weather events like hurricanes, as well as chronic stressors like droughts. Communities need to be informed not only of day-to-day weather events, but also of the ways our changing climate is affecting and altering weather patterns in their regions. As you hear from our panelists, there are really incredible, really incredible technological advancements being developed and deployed that are dramatically improving our ability to make actionable forecasts. But for a forecast to be truly actionable, it has to be communicated by the experts to the people who are likely to be affected. And then those people have to understand what they're being told. It's really hard to undersell the importance of climate communication and emergency outreach and how weather and disaster information can be delivered to the public in ways that support equitable, accessible, and effective community preparedness. My colleague is going to put up a slide in just a moment with a link to our survey. We read every response. We'll also have this link at the end. But in case anyone has to take off a little early, we really appreciate your feedback. We read every response. Let us know how you think we did today and we'll do our best to improve. We're going to get started with our panel. We may interrupt our panel a little later because we have a special guest joining us. But we are going to get off to a good start. Before we do, however, we will take questions at the end of our panel. If you're in person here in the room, we'll have a microphone moving around in the room. My friend Emily, one of our policy interns, will help you with that so you can ask the question to our panel. But I know we have several scores of people, because I just looked at the number, on our online audience. And if you are in our online audience, you can also ask us a question. The way to do that is to send us an email address to use as ask. That's ASK at EESI.org. We will also be having real time coverage on Blue Sky, our Instagram story, threads as well as X. And you can find more information, including presentation materials on the front table, as well as online. And our speaker bios will also be available online. Any further ado, the reason why we're all here is this great panel. And it's my privilege to introduce Michelle Minnelli, the deputy director of the National Weather Service. Michelle, thank you for being in our briefing today. I'll turn it over to you. All right, hello, everybody. And I'm really excited to be here. So happy birthday National Weather Service. So we just had a birthday on Friday, February 9th. We are 154 years young. We've been around a while, and the growth and the transformation that we've had over the years have been absolutely spectacular. I've been in the Weather Service. I call myself a Weather Service lifer. And they're over 33 years. And we are everywhere across this country. We are in 168 operational units across the country. We have over 4,400 employees. Our breadth is large. We have not only scientists. We have engineers. We have IT specialists. We have social, behavioral, economic economists. And we have the breadth of our administrative and project management that ties it all together and our modelers. We issue over 1 and 1 half million forecasts per year and around 50,000 warnings each year. And each year, they're going up. You look at the climate, and you look at where we're going with the extremes over and over. It's more and more each year. And we're a bargain. So if you look at what we get for our budget, for Weather Service, it costs about each person about $4 per year for what our budget is. So it's pretty spectacular about what we're able to provide. So taking a look of why Weather Service matters, look at the billion-dollar disasters that we've had across this country. If you look at 2023, we've had $28 billion disasters just in that year alone, and each year continues to increase. Our focus is with the communities of who we serve. We want to be eye-to-eye alongside the emergency managers, the decision-makers. And if you look at this map, we're everywhere across that country. And if places that we're not, that's where we want to get to, especially those vulnerable communities that we haven't had access yet to. You look at our impact-based decision support services. Our forecasts are not about a point-click, what's the temperature, or the wind's going to be at day five or day seven. What matters is whether it's a day out or seven days out. What are the impacts? What are the so-what of what is the information that's going to matter most to that community? What are their vulnerabilities? And how can we take the science and look at the impacts and translate that into the decisions that we're going to be doing? Do you want to continue? So if you look at, for example, of what we had in February, just this past month, last week, of the historic Southern California rains. And you look at the TikTok of that messaging and that impacts of what we're doing. So looking at two weeks out, we're communicating to the communities of what are the risks, what are the highest amount of impacts. And then we tailor that down. We look at our four to seven days prior, and we're getting more specific. What are the areas that you most need to look at? And then we're getting into the communities. We're in the emergency operation centers. We're talking with the decision makers. We're amplifying across the weather enterprise, making sure that message is going to continue and get amplified, whether it's on media, social media, broadcast media, across the board with our weather enterprise. And then down to the actual warnings that get issued. So what's our transformation? We are going through an incredible transformation in the National Weather Service. We start, our base is our data, our observations, our suite of observations. And we are looking at partnering with data buys in the vendors that are out there. What is the right level of satellite that we buy, that we fly ourselves, the data that we ingest, and that we build, and then also new technologies that are out there that can look to take measurements across the atmosphere. Our supercomputers are getting larger and larger with the amount of the ability to process more with our compute. But the model data is getting higher and higher, which means we have to move it around. Only we can't continue to move it around from location to location. So we're looking at cloud technologies of where we're gonna place it for better data access. So you take our observations and you take our model guidance. And we're not in the weather service only behind our computers, right? We are out there. We are going to be where we're needed the most alongside those decision makers. In some cases, the emergency managers, their office is their car and it's during an event. And so we will meet them where they need us to be. We will invite them into our offices. We will go through events, we will be stationed. We're having now positions that are located at state EOCs. And it's been an absolute, we're gonna be continuing to do that and expanding across the board. And where does that get us? That gets us to a more prepared, weather ready and responsive communities. And further connecting into those communities that need us the most. So this is our transformation roadmap. There's a QR code here. This is our 10-year plan. We have a new strategic plan. It's the shortest in government, three pages. I know it's hard to believe. We're focused on our people, our infrastructure and our future. And this roadmap really does set the stage of what we can accomplish in the next 10 years. And we're not doing it alone. We're getting that information on how to do it with our emergency managers, with our entire weather enterprise on how we're gonna get there. And we can't do it alone. And it's gonna take our whole communities to get us there. So I just wanted to say thank you. And if you need anything, there's my email. Thank you. All right, thank you. Thank you. Thank you so much, Michelle. Thank you very much, Michelle. That was a great presentation. Four bucks, that's not bad. Like those old commercials for the price of a cup of coffee. Less than a cup of coffee. Well, he's the McDonald's app. So mine was 99 cents this morning, but yes. Our special guest is here and it is my privilege to introduce Representative Eric Sorenson. Representative Sorenson serves the people of the 17th Congressional District of Illinois. And before being elected to Congress, he was part of the fabric of communities across Central and Northwestern Illinois as a local TV meteorologist for over two decades. Communicating life-saving information that impacted those people's jobs, schools, farms, and safety. Today, Representative Sorenson has a seat on the House Agriculture Committee and the Science, Space, and Technology Committee. And on the Science, Space, and Technology Committee, Mr. Sorenson serves as Ranking Member on the Space and Aeronautics Subcommittee. So really, there's no one better in Congress to join us today for this briefing. Mr. Sorenson, welcome you to the lectern to say a few remarks. It's so nice to see you. Thank you. I will say, much like my previous job as a meteorologist, in Congress, you can wear whatever shoes you want. And so, that's the great part. Well, thank you, Dan, for the introduction. I really appreciate it for EESI, for hosting this briefing today. Truth be told, the only job that I ever wanted when I was a kid was to be the meteorologist on Channel 13 in Rockford, Illinois. Why? Because when I was a kid, the meteorologist on TV was the one that was helping me not be afraid. His name was also Eric, so it was like he was talking to little Eric. So, I got to live my childhood dream for 22 years for most of that time in my hometown of Rockford, Illinois, but more importantly, as a trusted source of weather information. I worked in Texas for five years, and then the rest of my career was spent in the congressional district that I now serve. My job was to help neighbors make good decisions, when to stay off the roads for winter weather, when to stay safe in heat waves, and also when to go in the basement because the tornado was confirmed. It was also taking the risk, long about 2008, of communicating climate. And ever since then, it was going and talking to other meteorologists all across the country to say that there's a way to do this without alienating your audience. I talked about climate regularly on television. I stood up there and I said to the camera, if you think that this is a political issue, then it's probably challenging your own politics, but I don't believe that this is a political issue. I believe you need to know about it, and you need to know about it from a trusted information source. Little did I know, I was actually talking more to the farmers of Northern Illinois because they were the ones that were already having to deal with it. One of the big concerns that I have are the family farmers today that even with the technology that we have, it's still harder to be a farmer today. Why? Because we have more extreme droughts. And then in between those extreme droughts, we have times of extreme flooding. And so it's harder. Even with the technology, we have more farmers that, for you and I, case in point, I changed careers in the middle of my life. But if you're a family farmer and you're great, great grandparents who are on the same land, we have farmers that don't quit their jobs, they can't. They quit their life. And so that is a challenge for us as meteorologists, as communicators of climate to figure out how we're going to make sense, have these people make sense of it with us. So when my congresswoman decided to retire, I knew that I had to at least try to become the first meteorologist in Congress in nearly 50 years. And getting here had to make sense to the people of my district. In the end, they just want the communication to be rooted in good science. And I think of how important it is to welcome the next meteorologist to Congress. Maybe the next broadcast meteorologist to Congress. As challenges arise, we need more trusted communicators in our government. Here in Congress, I'm proud to be a part of the space, science and technology committee, where I helped draft the Weather Act. For the first time, a meteorologist was included in the reauthorization language. How we support reanalysis and reforecasting, which helps NOAA and the National Weather Service improve our forecasting capacity by learning from the analogs of past storms. And I don't need to tell the folks in this room what some meteorologists think that the GFS stands for. Well, you know what? Since there's a meteorologist in Congress, we're going to put more funding into the global forecast system to make it more accurate than it's ever been. We have the attention now to source more accurate conclusions that come through additional data points in the beginning of our equations. So what does that mean? Well, story time real quick. I remember one of the biggest storms literally hit my home. I was without power for a week. I lost a tree in the front yard. It was the 2020 Eastern Iowa Durecho. I was on the air for this for about three and a half hours. It was a storm so strong, I didn't believe the radar data when it was coming in in real time. The velocity data from the Quad Cities National Weather Service office was showing a hundred mile an hour inbound wind. But then we started getting the ground truth from the radar that was coming in from the spotters that were in the field with animometers in their hands. And the fact is that after that storm moved through Cedar Rapids, Iowa, something like 70% of the mature foliage, the mature trees were gone. Cedar Rapids has a population of more than 100,000 people. So the electricity grid was even impacted. Durechos are produced by immense uplift of air within a storm that are pushed for many hundreds of miles. This storm remains one of the worst storms that I have forecasted and that I covered in more than two decades as a meteorologist. So the process for understanding thunderstorm development, it's well understood partly because we use model data but we also need the know-how of the meteorologists that can recognize which situations are similar to what we've had in the past. But that also will get more challenging as we deal with situations and storms that have never occurred before, all thanks to our changing climate. Storms we didn't know could now exist. Rapidly intensifying hurricanes that we've seen in Houston, dumping years worth of rain in just a matter of a few days. Extreme droughts that happen in record wet seasons and I'm not going to get into the more complex issues that we have at the higher latitudes that are happening in the Arctic and the Antarctic. That's why it's so important that our weather forecasting agencies had the tools that they need to keep people safe. When a tornado touches down or a hurricane makes a landfall, we have to have the people in place to keep others safe. And with an increasing accuracy rate of forecasts, people we know are making better decisions than ever before. I was proud to vote in support of this Weather Act and Committee. I look forward to it coming to the full house and eventually to the President's desk. And finally, I hope that sharing our weather stories helps illustrate to you the importance of why we need to be investing in the tools to help meteorologists inform the public. When I think of the abilities that we have today to forecast the weather versus when I was even a kid, it makes me ponder what's in front of us. Will we be able to evacuate a town before the tornado hits? I'm a believer that that's 10 to 20 years away. That means that the kids who are in first grade today will have that discovery to make. So today, we can sleep well knowing that we're investing in the technology so that they will have the ability to make it. So again, a big thank you to EESI for having me here today and for everyone who's attending the briefing. We can maybe take a couple of questions. Absolutely. We would love that. Thank you so much for your remarks. Quick round of applause please. Oh, thank you. If anyone has a question in the audience, please make yourself known and my friend Emily will bring the microphone to you. There's even a one minute left card and I didn't even get that one. Not for you, especially. That would have given me flashbacks to television when you got to wrap it up quick, we're going to commercials. No, it's not like the Oscars. It's because of your strong sock game. You mentioned your shoe. Sock game deserves a shout out tip. Any questions before Representative Sorenson leaves? Well, I have a question. And in your two decades as a meteorologist, were there things that came up, tools that came available to you over that two decades that made your job as a weather communicator and a climate communicator easier or better able to be relevant to the lives of the farmers and other people in Northern Illinois? Right, I will say my undergraduate study at Northern Illinois University, I'll date myself, 1999, we really didn't have much climate science in the curriculum. So that was something that I had to learn on my own. One of the first things that came up was the polar bear. It was really hard for me being a meteorologist in my hometown to get people to care about a polar bear that they've never seen. So what I needed to do was localize the impact, localize the, equalize the impact because there weren't the resources that were there. There was a thing very early on, it was called Climate Matters in the Newsroom, which morphed into Climate Central, climatecentral.org for everybody out there. It allowed us to localize the impact of the changing climate to my specific television market to be able to have the graphics that I could just put up on the screen and say, here's the trend line or to say that here are the forest fires and the floods that happen. Or one of the best results I ever got, the responses from the viewing audience was here's how climate change is going to affect the process of making beer. That got people to perk up. Wait a minute, climate change has an effect on beer. You can find that on Climate Central's website today. So very simply, that allowed me to have the tools to explain it. But now, if you look at the map of different television markets where meteorologists are using that critical data to communicate it, it's almost in every TV market in the country. Well, thank you so much. And Representative Sorenson, it means a lot for us to stop by, for you to stop by the briefing today. Thanks for sharing your story with our audience. And yeah, good luck with the Weather Act. And if anyone has any questions at all, sorenson.house.gov. And if there's anybody out there that thinks that they can't do this, I am a firm believer and I will support whoever wants to be the second meteorologist in Congress because we need, certainly, imagine what happens to climate science and communication of science in general if all of a sudden congressional districts start electing their science communicators. Thank you. Thank you so much. Thank you so much. As Representative Sorenson walks out, he mentioned that we have, that there's the brewery issue. We actually have a podcast episode coming up, not too long, about climate and the winemaking industry because that's more my cup of tea. But my colleague Allison is busy putting that together. Excuse me one moment because I forgot my screen. I wanna make sure that I get Thomas' title exactly right because I can't see it from the other side of the card. Our next speaker today is Thomas Cavett. Thomas is Vice President of Government Affairs and Strategy at tomorrow.io. Thomas, thanks for being here. Welcome to the briefing and I'll turn it over to you. I'm following on the heels of a member of Congress so bear with me here. See if I can make this work. There we go. All right, good afternoon. Thanks you all for being here and for those of you who are joining us online. I really appreciate your time today. So Thomas Cavett, I lead all of our Government Affairs policy work here on the Hill as well as internationally. So I'm gonna tell you a little bit about who we are tomorrow at I.O. And then I'll talk a little bit about more sort of structural changes within the industry around innovation and how I think the industry is really in a new position to really support and augment what the US government is doing from a weather and mostly weather but somewhat climate forecasting standpoint as well. So tomorrow I.O. was founded in 2016. Our founders were in the military, received weather forecasts that they found didn't really enable them to make really good operational decisions. They lost colleagues to weather incidents and things like that. And so they really wanted to change that. And so first and foremost, we're a weather intelligence software company and our software platform it's called the Weather Intelligence Platform. Really it's about contextualizing and translating that weather information into the decision that you need to make. So we serve a really broad array of customers on the commercial side. It's big companies you've heard of like JetBlue and Uber and Ford. They're using us in their day to day operations. We span really every industry, right? Even restaurants, shipping companies, electric utility providers because all of them are impacted obviously by weather but in very different ways. We've been around since 2016, as I said, we're, although we're building this software platform, we're also a space company and I'll get into sort of how we got there. We also work with a number of government agencies, the US Air Force, NOAA, NASA and many others. We're about 250 employees and we've raised about $300 million. So really pretty high growth in that period of time. We also have a nonprofit that we created a few years ago that works mostly in the developing world but they're really leveraging this idea of weather intelligence to reach smallholder farmers and other really disadvantaged communities because the impact of weather on what they're doing and their livelihoods is so dramatic. It's something as simple as a text message with instructions of what do I do about this weather? There previously they had nothing but maybe sort of folklore and tradition to deal with. We're seeing really dramatic changes in their productivity and the output of their crops and things like that. So what is weather intelligence, right? We've all gotten a weather forecast our entire lives in really kind of a consistent format, right? It's for the most part, here's what the weather's gonna be like today in your region, in your city, in your zip code but not necessarily contextualized to the specific thing that you care about and it's very unique to the individual. So here you might see something like, oh, let me move this mouse out of the way. You know, suspend train operations on this specific section of track because of crosswinds or something along those lines. It's really unique to your business because if a train derails, that's tens of, if not hundreds of millions of dollars in damage, not to mention lives potentially lost or injuries. So what we do is we make out this logic-based system where you can set thresholds of across any weather variable that you care about but what that allows you to do is automate this at scale. So of course we still have meteorologists in the loop because this is not a perfect science but what we can do is make it very unique to your use case. And so we build these templates across industries and then we allow 100,000 person airline who previously was relying on maybe a dozen meteorologists. Every employee at that airline can't call that meteorologist all day, every day, every time they need to make a decision. So what we can do, automate that, disseminate that information, text messages, push notifications, whatever it might be and we really find a lot of operational value in disseminating information in that way. So our company is really innovating across the entire value chain, not just the end product of how you consume weather. We also run our own proprietary models. Of course we leverage government models as well but we're able to run our own models which allows us to be more accurate in many ways and that also allows us to ingest more observations and so we're building a constellation of satellites because we realize that one of the biggest challenges is observations. If you run a higher resolution model you need more inputs, otherwise it's just garbage in and garbage out. So this quality weather information is not really universal, right? I mean in the United States we're fortunate enough to have a pretty high quality, reliable forecast but it doesn't exist in many parts of the world, not to mention over the oceans. One of the sort of backbones of our weather infrastructure is ground-based radar systems. So this is what the map of ground-based radar systems looks like around the world. This is how we see precipitation, really quantify precipitation. Other sensors are useful but for the most part when we're leveraging those other sensors they're passive which means we just sort of infer what's happening inside of clouds. So next rad, our ground-based radar system as you can see is quite dense. Western Europe has something similar, really the developed world but over the oceans where many weather systems develop there are no observations almost. Maybe a few observations from a ship or from an aircraft nose cone and not to mention the developing world where again this is sort of disproportionate impact to those populations. Their infrastructure is disproportionately impacted by extreme weather events and so we set out to really solve that. Well there should have been satellites on here. There are three I promise. Two of them are on orbit today so apologize for that but we're launching 30 in total and what those consist of are radars and microwave sounders. Again taking that radar concept from next rad which is on the ground putting that in space. Previously there was only one radar in space built by NASA. We launched the first two commercially built weather radars in history and we've actually found that the data quality is even better than what's coming out of that satellite which is called GPM from NASA and JAXA. The really important takeaway here is with this constellation we're gonna reduce the current average state-of-the-art revisit rate which is about two to three hours for the sort of total constellation that we leveraged today to less than 30 minutes for the entire globe. So a really dramatic step change in observations and of course maintaining really high quality data along the way and we expect to have that up by the end of 25 or at the latest sometime in 2026. The really important thing here is that this presents a cost savings opportunity so the commercial industry has started to invest in this because there's a dual use here. Commercial sector is realizing they need better weather information and they're willing to pay for tools like this. And so what we can do is building on the history and the scientific advancements of NASA with GPM we're able to reduce the size, reduce the cost and maintain the same quality orders of magnitude which is really significant from a public and private partnership standpoint. So a study that came out many years ago but basically said that if anybody's able to do this you're gonna revolutionize weather and climate science so that's what we're set out to do. When you compare what our satellite data quality is against that GPM mission so the correlation to MRMS which is a ground-based NEXRAD derivative product our correlation is actually even higher and that's just with two small spacecraft that are about the size of a mini fridge versus a billion dollar school bus. So pretty exciting to see that data quality coming out of that system with just the first two. So together this constellation is gonna provide across the full spectrum of weather forecasting and of course over time climate data as well in ways that we've really never had before. There's never been a space-based global radar dataset like this so we're pretty excited about the applications across really the entire world. Now one thing I should mention is that and this is primarily sort of a civil conversation but there's a defense and national security element to this as well, right? I was in the Army and I can tell you from firsthand experience that weather forecasts in those other parts of the world where we operate are usually pretty worthless if I'm being totally honest. The Air Force and the Navy do a great job of creating a weather forecast but there's no infrastructure and so they're relying on really sparse data to create these systems. Not to mention the fact that China has also launched a copy of GPM when ours is in extended operations and could fail at any time. So we're in some ways falling behind as a nation so really important to what I'm about to talk about NEX. So what I would call weather 1.0 is really the traditional weather enterprise where the private sector was mostly focused on the last mile delivery of information. They were building software, leveraging government model outputs and really disseminating that information for you. There wasn't a lot of innovation happening up the chain. Most of the work and investment was done by the government agencies and it also meant that there's this sort of disproportionate expenditure of investment into these observational networks and this infrastructure which gets to kind of the global problem that I mentioned. And of course that matters because those weather systems that hit us form in other parts of the world sometimes. Today, things are changing, right? The industry is really innovating across that entire value chain from data and observations to the models themselves and ultimately to that end product and there's really an opportunity for us to collaborate more to leverage what the private sector is doing. I got the one minute mark so I'll speed up a little bit. The main thing here is we're taking on the risk, right? We're leveraging private capital to do these investments. We're building these systems sort of whether the government is a customer or not because the commercial sector sees the value in what we're doing. We also know that over time, industry is likely gonna surpass what the government can do just because we can take bigger risks and we can move faster. There are a number of mechanisms that exist today for the government to leverage what the commercial sector is doing. NOAA has their commercial weather data pilot. NASA has the commercial small-sat data acquisition program. The Air Force has another similarly named commercial weather data pilot, but these programs haven't grown as quickly as the industry has. Since they were created, this industry has suddenly sort of blossomed into something new, but there's more opportunity to expand these. Lastly, I'll just say, so tomorrow I was also a member of something called the Commercial Weather Alliance where 11 companies that are operating all different types of weather systems, observational systems, balloons, buoys, drones, across the whole spectrum. And all of us are really excited about this new era where we can really support and augment what the government is doing. And yeah, I'll wrap up there. Thanks very much. Happy to take questions. I suspect we'll get some of these opportunities when we get to the Q&A. And sorry about the conversion via your slide, but one of my colleagues is gonna try to pull up a different version later so that you can show the audience the satellites because they are pretty cool. They're shiny and they go to space and we wouldn't wanna miss that. Our third panelist today is Pierre Jean-Tin. Pierre is the Maurice Ewing and J. Lamar Wurzel professor at Columbia University. Also the director, Learning the Earth with AI and Physics, National Science Foundation, Science and Technology Center. Pierre, welcome. You traveled in this morning for the briefing. We really appreciate that. Take it away. Thank you. Great. And thanks for having me today. It's a pleasure to be here. Okay, so I wanted to give you a little bit of a sense as to how AI is really transforming the weather and the climate landscape. And it's not just my work, but trying to give a little bit of a sense what we can do in 10 minutes to the best of my ability. And trying to give you a little bit of a taste as to how that field is really moving really, really quickly now. And as was mentioned, so I'm directing a center at Columbia, which is a partnership actually with NYU and also UC Irvine and University of Minnesota. And what we are trying to do is actually using AI to improve what we call climate models. So things that are used for the projections such as used in the IPCC report. And I also launched co-founded a company a year ago, which is called Telus, which we are actually looking at sub-signal to seasonal forecast. So it goes without saying that climate adaptation is needed. So every year now we are actually witnessing more and more droughts, flooding events, heat waves even very early in the spring now. And we're also witnessing very peculiar events such as the wildfire smoke that we've been witnessing for instance in New York and on the east coast back last July. And really it goes without saying now that climate change is happening, but the question is, are we prepared for it? And I would argue that at this stage we are still really lagging behind in many ways. And kind of the bulk of the funding at this stage really goes to what we call climate mitigation, which is basically trying to reduce emissions or even trying to remove carbon dioxide from the atmosphere. But a very tiny fraction of that only goes to climate adaptation. And you could say it sounds dire to say that we need to adapt to climate change but it's just that things are already changing, right? So we need to have a plan and at this stage in many, many ways we don't have one. And so why is that needed? That's actually needed and some issues with the slides as well for me. It's because the cost associated with climate events and especially extreme events is actually on the rise and that's been dramatically affecting everyone. So Michelle has been talking about that a few minutes ago. And so we are really seeing a rise due to for instance, flooding events, droughts and wildfires, a lot of that being witnessed in the Western US. And that cost really is impacting the private and public sector alike and we need to get better prepared for that. In tandem with that and we could get into the weeds here is actually the fact that we also overlaying population and there's actually no movement of population and population are really sitting at places that are affecting the most such as in the Asian regions. So how can actually AI help for that? So it can help on different fronts. The first one would be to first help with the resilience of society and public and private stakeholders. So just to give you a quick sense as to how AI is being used these days. This is an example here from Google where they've been developing this flood forecasting infrastructure and basically what they have is two AI based algorithm. The first one is kind of trying to predict rivers and how the river flow will be actually responding to an inundation and basically to a flooding event. And then they overlay that with another AI based model which is actually showing you the inundation. And the advantage of that is it can be running very, very quickly as opposed to a physically based model and that pace of response can be very useful to actually provide that to a broad range of stakeholders. So you could actually get an information on your cell phone saying you're basically at risk now of inundation. People could actually move and be prepared for early warning systems. So that's really tremendous. This has a huge impact in terms of being prepared as a society. There are other examples. Another example from a company. Actually they are based in Brooklyn. A good colleague. She was at Columbia at the time. And they are looking at flood monitoring in almost near real time using constellation of satellites of meter, submeter scale satellite data to actually map especially inundation in urban context. And they can actually basically predict or monitor that in the area of time so that people can also be better prepared and making sure that people could also have like early warning systems. That's also being used for insurance and the re-insurance sector so that people could have better assessments of inundations in particular in the urban context. AI is also being used for different some other types of extreme events. So wildfires in particular, hurricanes as well. This is an example from the Department of Energy. This is called the radar rapid analytics for disaster response. Where they are also using satellite data to monitor in near real time basically disasters and especially wildfires but also to try to predict their trajectory. So for wildfire and hurricanes so that people could get a sense as to where things are gonna be hit next and that's a great way to be prepared. So that's a great early warning system so that we could actually move populations and also we could allocate resources at the right time and at the right location. So again, those things are really transformative so that we can better use our resources and mitigate the impact of climate. There's been what we call a second revolution so that's been happening very, very quickly over the last two years. So we talked about weather forecasting just a few minutes ago. This is actually now really AI super powered. So there's been two years ago a lot of push actually started a lot of that by a former collaborating student, Stefan Roth, that really started the groundwork and the laywork to actually start looking at how can we actually improve weather forecasting using AI? And basically in that sense, the idea is that we're looking at the historical data so what we call re-analysis products that are actually developed by the weather agencies and we are trying to replicate that with some sort of sophisticated AI model. And what people have shown recently over the last year and there's been a lot of battle basically moving in that direction is that those AI based model can do a better job than weather forecast or physically based models that we've been using for the past decades. And that's really exciting and people call that the second revolution for weather forecasting where now we can actually get better forecast and also we can get that at a much faster pace again. And in fact, you can even run that on the desktop or a laptop now. So it's so easy to run. And so basically it opens up lots of different possibilities that we didn't have before. You don't need a supercomputer to run that. It does take a lot of time to train those models. They are really greedy and very hungry in terms of the data. But once they are trained, they are actually very easy to run. So there are many people in that landscape. So DeepMind is an example and Vidya, Huawei and the European Center for Medium Weather Forecast is actually pushing very much in that direction and they are now actually using the AI based model as part of their weather forecasting. So that's operational right now and that's been happening very, very quickly. And I would mention that Europe is really leading the way in that effort, especially ECMWF. So we really need to push in the US. I want to point out that they are limits to those models. So they are great, but just over a couple of weeks. So typically they do a great job from one or two days up to two weeks. So that's just a schematic here showing you basically the error as you go up. Basically it means that you have more and more errors building up as a function of the forecast lead time. And those are some of those different models. So for instance, the PENJU weather forecast model, the forecast model from Nvidia and the graphcast model from DeepMind. The main take home here is that they do a great job for a couple of days up to two weeks, but then they do actually a very poor job. They are not, they are actually worse than just the mean seasonal cycle. So just looking at historical data. So they do a terrible job after two weeks. The reason for that is that they, how many droughts have you seen in the past? How many El Nino have they seen? Very, very few. So you've seen many days of weather, many weather patterns, but you've seen very, very few extreme events. So there are limits to those models that would argue especially for extreme events. So we really need to push that forward. And so that's why actually with a former student we created this company called Tails. But there's a lot of work to be done in that space. And just to wrap up, I would like to mention, so that's one thing we're actually doing in our center at Columbia and where you, you see your van in University of Minnesota. We are looking at what we call climate projections. So long range forecast, so 10 year plus, which is really used to think about climate adaptation in the long run. Like really it's thinking about investment and all that. And basically what we've been doing for decades and what is actually basically the substance of the IPCC report and the Paris Agreement is we take a bunch of physically based models, so what we call climate models and we typically take some sort of democratic vote across those models to give you the best forecast of the future, right? So democratic average behavior. And the issue is that we know that those models are very, very divergent. There's a lot of inaccuracy in terms of how we can actually forecast the future. And that's an issue, right? Because you'll be, and actually the representative just a few minutes ago was saying you need to think at the regional scale and you need to be very precise, right? So you can actually use AI there as well to actually use past data and combine that and merge that with climate models so that you can do a better job, especially at the regional scale. So you can do a better job at predicting temperature or precipitation with much more accuracy than what we have in those physically based models. So this is just an example here. We're on the left-hand side, this is some work we just submitted where you actually have a correction based on machine learning of some of the IPCC basically models, so the SEMIP couple modeling comparison project. And you can see that some of the regional structure is actually changing, right? We can see for instance, what we call the Arctic Amplification. We know that basically the normal regions, normal regions is actually warming more and we know that models tend to be deficient. That's something we can correct when we expose those models to data and do that fusion with AI. So we believe that can be quite transformative and we can narrow down, so this is those estimates that are used here for the Paris Agreement. We can narrow down the uncertainties in those estimates so we can really do this mapping between how much carbon dioxide do you have in the atmosphere and what is the temperature of the planet that corresponds to that. We can actually narrow down those uncertainties. So just to wrap up, I would say that we're actually in a much better place now, thanks to AI than we were just a few weeks ago, a few years ago. And that's on many different fronts. That's actually on the monitoring side of things. And also on the prediction side on the short term, so weather forecasting again is witnessing a second revolution and we are really seeing some of that going also all the way to longer forecasting time scales. There are still gaps. I would argue that sub-seasonal to seasonal forecast, which is really critical for predicting droughts and impact is still very difficult to do and we don't really have great tools yet, so we need to work on that as a community. And training data is very important. We talked about data before, making that data also available to the borders community so we can work together and basically improve those models is really, really critical. So data availability is part of that. And with that, I'd be happy to take any questions later and I'd like to mention that, as was mentioned before, that public-private partnership is becoming really critical and pretty seamless as well. There's really great connection these days. Thanks so much. Thank you, Pierre. That was an excellent presentation. As a reminder, everything that our panelists are presenting today and other materials as well, they're posted on our website, on the briefing page. You can visit us online, www.esi.org and we'll have time for questions at the end of our panel in person and online. If you're in our online audience, you can send us an email and the email address to use is askaske.esi.org. Pierre, you talked a little bit of artificial intelligence. Now we're gonna return to the natural intelligence portion of the panel and it's my privilege to introduce someone who's part of my morning routine, Dan Stilman. Dan is the co-founder and meteorologist at the Washington Post's Capital Weather Gang. Dan, welcome to the briefing today. I'll turn it over to you. Just not artificial. I'll leave it there. All right, well, pleasure to be here and to have an opportunity to speak to this audience. So I wanna go over a few recent weather forecasting challenges. I wanna preface this by saying we have made dramatic gains in weather forecasting over the decades, over many years, but I've cherry picked here some examples of challenging forecasts because, especially we're speaking to Congress here, we're speaking to funding. We don't want you to think that forecasts are perfect and we don't need any money because that is not the case despite the many gains that we've made. So I wanna talk about where weather forecasting is today in terms of accuracy and then what are the key initiatives going forward to make forecasts more accurate, more useful, and drive better decisions. So, and I'm gonna skip through some of these slides, but if you've seen something that you wanna ask me about later, feel free, but in the interest of time, I'm gonna focus just on a few specific topics. So first forecasting challenge was just a few weeks ago, late January here in DC. We had, if you remember, temperature hit 80 degrees. That was the first time we got up to 80 in January on record. It was the earliest 80 degree day in DC in the year on record by several weeks. Beat the previous one was February 21st. So this was an extremely warm day occurred during a worldwide warm spell that touched nearly every continent. And what I wanna show you here is I don't want you to try to understand everything you're looking at, please don't, but I wanna focus in on the forecast for that day and on the left side, you can see what circled in black were the forecast from a variety of models five days before that day. And you can see that forecasted temperatures were all around mid sixties, about 65 degrees, a couple of outliers cooler and warmer than that, but about 65. So that's 15 degrees lower than the actual forecast, which in the winter, not a big deal, people appreciate an 80 degree day in January, but imagine you push that 20 degrees higher in the summer and you're thinking you're gonna get 85 and you get 100, there's ramifications of that. Even one day ahead of time on the right side, there are forecasts were still six to eight degrees too low. So these are some of the sides I'm gonna speed through a little quicker, but just to say that extreme warmth was the combination of both that long-term climate change, warming from climate change and natural fluctuations in the atmosphere, natural cycles that tended to line up. That included the polar vortex and ongoing El Nino, as well as record warm oceans. I wanna take up just a second on this slide. You see 2023, the orange line was the warmest year for the world's oceans on record. And 2024, the top black line were already off to an even warmer start. And so that heat in the oceans translates into the atmosphere, it does influence and increases the odds of extreme warm temperatures like we saw on January 26th. So another forecasting challenge was Hurricane Otis. So Hurricane Otis was October. That was a Mexico landfall, so not in the US, but still a storm tracked and predicted by the National Hurricane Center. Same models that would be used for any storm here in the United States. And the forecast was not a good one, unfortunately. You can see on the charts on the left side there, the top left, the top right, and the bottom left. Those were all model forecasts from 24 hours before a landfall. So just a day before a landfall. You can see all the three models there predicting a relatively weak storm, a tropical storm. Otis ended up a category five, the bottom right was the actual image of Otis at landfall. So it intensified from a tropical storm to a category five in 12 hours. So we call that rapid intensification. And what we're seeing is that rapid intensification, more hurricanes are rapidly intensifying like that. Those kinds of hurricanes are harder to predict, they're harder to prepare for, and there are real consequences when the forecast is off of those. And then the last example for now of a forecasting challenge is just this week. A few days ago, the North Easter, we had this huge shift. If you're familiar that further up in the Northeast, we had a significant snowstorm. We had a huge shift in the snowfall forecast just the day before the storm. The models shifted the storm south just from late Sunday night or early Monday morning to Monday evening, and that span forecasts changed from they dropped by six inches in some locations, they went up by six inches in some locations. You had Boston's forecast initially expecting about a foot. That forecast went down to four to eight inches. They ended up with 0.1 inches at Logan Airport, so almost no snow. Similar situation at Albany, New York, so this was a tricky storm for the models and for forecasters. So again, despite these examples, we've made tremendous gains in forecasting over many years, and you can see on the left side, general forecast, three days, five days, seven days, 10 days out, those rates, those curves signify an improving forecast over many years, but notice towards the right side of that graph, those lines start to flatten out. So we've seen a bit of a flattening of that improvement, not improving the forecasts at the same rate we used to be. Similar for precipitation forecasts which were shown on the right side there. So how do we get things moving back in the right direction? Well, observations are a crucial part of that. They are considered backbone of weather forecasts. Models use observations as a starting point. They need to ingest as much detailed data as possible about the current state of the atmosphere. Excuse me, that's why NOAA, as we heard, and other government agencies, they fly weather satellites, they send up weather balloons, they've got all these sensors. As part of this observation network, it's an international collaboration which is extremely important, right? Any local forecast, any town and city in the US, that depends on data from all over the world. So NOAA works with its partners internationally. The forecast models we have, though, are limited by incomplete observations. We don't have as many as we need of the surface of the earth and the lower part of the atmosphere. Satellites can only probe down so far into the atmosphere and we can only deploy so many sensors on the surface of the earth. So that's why you have agencies like NOAA that are continuing to maintain their own observations networks and plan for the future but also work with the private sector, work with partners like tomorrow.io and others to explore commercial satellites, commercial radars, augment the government data with affordable data from the private sector. I wanna just follow briefly to Pierre on the area of AI. Because this is an area I've been covering for the Washington Post, as Pierre alluded to, there's huge developments in the area of AI, weather modeling, in just the past couple of years. And AI is harnessing this historical, vast archives of historical data, finding patterns in that data to predict the weather. In a different way than our traditional models do where they use these mathematical equations that model the physics of the atmosphere to predict the weather. And this approach with AI really has some benefits, right? You can run the models faster, you can run them cheaper. And in the past couple of years, these AI models have now started to catch up in their ability to predict the weather as accurate sometimes, in some cases, a little more accurate even than some of our traditional weather models. So there is a lot of excitement about what AI can do for weather forecasting and also an acknowledgement that AI is not here to replace our traditional weather models like the GFS that was mentioned earlier and from the European Center and others. It's a symbiotic relationship AI models use data that come from the traditional models. They depend on, they're trained on data from the traditional models. So it's a hybrid situation where we need to keep supporting both. And then last but not least, every bit as important as an accurate forecast is commuting that forecast effectively, right? So social science is a key part of that, key part of turning forecasts into well-informed decisions. NOAA has a number of initiatives in social science, including in its Weather Program Office. You're gonna see later this year, come hurricane season, the National Weather Service is gonna be testing out a new cone forecast, cone map. You can see the existing one on the left and the experimental one on the right that they'll be sort of piloting in parallel that tries to highlight hazards beyond just the cone in the center track but also other hazards inland, such as flooding and what have you. So the ability to communicate the forecast and understand how do people interpret probabilities? When they hear 40% chance of this or 30% chance of the answer, what does that mean to them? What kind of actions do they take? That's a whole field of social science that is an important component of this forecast process. And then lastly, some messages to Congress. More accurate forecasts, earlier warnings, better characterizing extreme scenarios and communication. All of that is what we need. That's gonna save more lives, economic benefit. And then I think it's really critical to understand from a funding perspective, when you properly fund weather and climate agencies, you're investing in the entire climate and US weather and climate enterprise because it's collaboration, because the public sector works with the private sector and academia, you are lifting up the entire enterprise, not just the government portion of that. And on the flip side, when you underfund any one component of all those ingredients we talked about, because they're all so critical to the end forecast, you undercut that end forecast and the potential for personal or organizational or economic benefit from better weather forecasts. Thank you for that one. I guess I could keep this as a souvenir, but I'll give it back, just in case. This is my natural. Yeah, exactly. That was great, thank you so much. And our final panelist today is another person who's part of many people's daily routines. This is Caitlin McGrath. Caitlin is a meteorologist at WUSA 9, our hometown CBS affiliate. Caitlin, welcome to the briefing today. I'll turn it over to you. All right, thank you so much, Dan. Meteorologists are oftentimes referred to as the station scientist in a newsroom because in addition to our daily weather forecasting, we report on anything from earthquakes to astronomical events. But the role of station scientists has truly never been more important than when it comes to communicating climate change. So we're gonna spend the next few minutes talking about what it means to be on the front lines of climate communication and how we can effectively communicate the impacts of climate change on extreme weather events and global disasters. Because the fact of the matter is, people trust meteorologists. Research shows that broadcast meteorologists are a highly trusted source of information about climate change falling only behind scientists, which meteorologists are, and family members. WUSA 9 works with Magid Consulting, a firm that specializes in media and entertainment, and they recently conducted a nationwide survey of adults ages 25 to 64 who reported watching at least one day of local television per week. Of those adults surveyed, 64% reported coverage of climate's impact on local weather as a strong reason to watch TV. 62% report coverage of climate's impact on major weather events and disasters as a strong reason to watch TV. It is information that people want, and it is in our part to deliver that information. Because meteorologists have a big platform, and now it stretches beyond our local broadcast market. Thanks to social media and digital media, our reach is greater than ever before. But meteorologists have to be willing to step outside of our traditional role. People no longer need us to tell us whether or not they need to take an umbrella in the morning or what the higher low temperature will be on any given day. They need climate information. And it's also our responsibility to help educate our viewers on what we can do to mitigate the impacts of climate change. Because, well, yes, there are changes that need to be made at a level well above most of our heads. I truly believe and research shows that there are actions we can take in our day-to-day lives to help curb greenhouse gas emissions and reduce our personal carbon footprint. And while tremendous strides have been made by closing the gap, I do still think there's work to be done in helping the public understand the climate science that will ultimately lead to greater understanding, acceptance, and hopefully action. So how can we do this? First and foremost, we have to meet our audience where they are. We know people are not watching TV like they used to. So all of the information you hear us talk about on broadcast is information that you'll find on our WUSA9 app and website. We post these updates on YouTube, Instagram, X, and Facebook. And I've found that the climate message is always best communicated when you can make it personal and when you can make the big picture a little bit smaller. I do a lot of school talks with students ranging from kindergarten through high school and I talk at a lot of community events with people of all ages. And when I ask the question, have you felt the impacts of climate change? The answer is almost always yes. It might not always be obvious, but when I phrase it like, do you remember the poor air quality last summer? Because of wildfires burning in Canada? The answer is yes. Or have you been down to the tidal basin and noticed the new sidewalks we have to use because the tidal basin floods twice a day every day in part because of rising sea levels? The answer is always yes. A couple other local examples that always seem to hit close to home here in the DC area includes talking about cherry blossoms, one of the most iconic things about DC. Now those cherry blossoms, they traditionally bloom in early April, but because of warming global temperatures, they've been blooming in late March. And I'd be willing to guess everyone sitting in this room and maybe even watching at home on your computer has experienced a wild DC summer storm where it just pours. Linking that excessive rainfall to algae blooms and how they pollute our local waterways every summer makes it feel a little bit more personal. We also talk a lot about the impacts of extreme heat and how it affects our health. But we cannot do this alone. We partner with a number of organizations. The media and broadcast meteorologists historically have a great relationship with the National Weather Service, both at the national level and with our local forecast office. We work with NASA, NOAA, the National Oceanic and Atmospheric Administration and local universities. We are so lucky here in the DC area to have phenomenal resources right in our backyard. We work carefully with George Mason University and their Center for Climate Change Communication, the University of Maryland and their Department of Atmospheric and Oceanic Sciences and Virginia Tech. So many wonderful local experts always willing to help us tell the climate story. We get a lot of data from covering Climate Now, the Yale program on climate change communication and Climate Central. Representative Sorenson touched a little bit on this, but Climate Central is truly changing the way that broadcast meteorologists can tell the climate story. They release free weekly graphics to more than 245 media markets in both English and Spanish that help us quantify the impacts of climate change on our forecast. We utilize those resources and report on what I refer to as solutions journalism. Solutions journalism to me is making the connection between what the climate issue is and what can be done about it or what is being done about it. An example of how we utilize these resources is here on your screen. On the left hand side, you see a graphic from Climate Central detailing the chill time required for fruit trees to effectively produce fruit. Not always what you think of when you think of climate change, right? So I took that information and I tried to localize it. I found a group of local scientists who are working on a solution to this very problem. A group of scientists from Virginia Tech are working on two solutions. The first being genetic modification, altering these trees before they are even planted in the ground. And the second is a topical solution can be applied to existing trees. Both of these solutions are helping the problem of late spring frosts. They're delaying the bloom of these fruit trees. Pardon me while I get a quick sip of water. I usually talk for three minutes, Max, not 10 minutes. So this is an example of how I took the climate issue, our warming winters, and I addressed what's being done about it. And most importantly, how it's impacting our viewers and our consumers. Because at the end of the day, we're talking about the fruit that you buy at your local grocery store or your local farmer's market and the cost of that fruit because of the impact on climate change. Another example, the power of trees and how they can help our environment and our climate. And this is data that's specific to Washington, D.C. And I wanna reiterate that whenever Climate Central releases this data, they release all of their sources, all of the research behind it. We are not just mindlessly regurgitating these numbers. But this graphic shows you that our local trees help avoid 276 million gallons of stormwater runoff annually. They help absorb 1 million pounds of air pollution and they help remove 40,000 tons of carbon pollution. And that doesn't even get into the benefit of trees and how it helps with the urban heat island effect. So I paired this information, I paired these eye-popping numbers with information about a local nonprofit, KC Trees, who has an incredible mission of enhancing the tree canopy here in Washington, D.C. to 40% by 2032. They're already up to a 37% canopy. How do they do that? They plant over 4,000 trees annually. How does that get accomplished? They work with thousands of volunteers. So the goal here is that someone watching at home or maybe scrolling on their phone might see this report and think, hey, that's something I can do. That's something I can do to help our environment and to help our climate. But not all climate data needs to be turned into a full report or a story. Sometimes we just incorporate it into our daily weather forecast. An example of that includes this. This details the number of days above 90 degrees that Washington, D.C. experiences and how it's changed over the years. And you can see that since 1970, we now experience 10 more days of temperatures above 90 degrees than we have in years past. That's pretty incredible. And it quantifies how we're seeing climate change impact our local weather. It is hard to do. It's a hard message to get across. But products like this are helping. Products like this are making it even easier. This is the Climate Shift Index. It's a relatively new product by Climate Central. It was released back in 2022. And it compares how often a given temperature will occur in our current climate with the frequency of that same temperature in a climate without global warming. They've created a scale ranging from zero to five, both positive and negative, that details how likely a temperature is in today's altered climate. Temperatures that are more likely receive a positive value, negative values indicate conditions that are less likely. And this is data that we get every single day for both high and low temperatures. And it looks a little bit something like this. I chose an example from this past Monday, where our low temperature was 13 degrees above average here in D.C. That landed it as a two on the Climate Shift Index, meaning that climate change made these conditions at least twice as common. CSI, Climate Shift Index levels of a two and higher indicate a dominant climate influence. This is a way to really help us quantify the impacts of climate change on our local weather. It helps us answer the question, is this climate change or is it just weather? And we need more products like this. Thank you so much for your time. My contact information is up there on the screen. I look forward to continuing the conversation. And you can also find all of the resources that I referenced during this presentation on ESI's website. Thank you so much. Amazing presentation, Caitlin, so much. While we're transitioning to our Q&A, my colleague is gonna come up and bring up that slide, Thomas, with the satellites. But that won't get us, keep us from getting started. Thanks for mentioning KC trees. We've featured them in Congressional Education in the past. They're amazing. And also the Yale Center. We did a briefing with them. Tony came down last February, maybe January, and did a part of our Congressional Climate Camp. So two other great ESI educational resources that if you wanna learn more about them, I encourage that. So we're gonna start taking questions. My colleague, Emily, has a roving mic. Dano, you're welcome to be as invasive as possible, or not as possible as necessary. Yeah, be careful what I ask for. So we're gonna go ahead and get started. I'm gonna kick us off while we're waiting for people to work up their courage to ask a question. And that is something that Representative Sorenson mentioned. He mentioned a derecho in Iowa. And that made me remember the first time I heard the word derecho was living here. And I was like, I don't know what a derecho is. Michelle, maybe we'll come back and start with you and then we'll go down the line. But how can we effectively communicate novel or previously unknown weather events like extreme heat or derachos to the public? And specifically to people who have no idea what that weather phenomenon is even like or what it even means. Great question, Dan. And I mean, yes, there's the science and all the terminology, but we need to, as scientists, we need to remember who are we serving and who needs this information. And communicating it, Caitlin, I liked how you mentioned it of taking a big picture and making it localized and making it smaller of what is the so what for that community? So you're gonna get a line of wins and they're gonna be strong and they're gonna come on fast. And what does that mean, right? So how does that translate? What are the local communities where there are power lines above ground in that case that are going to most likely gonna have power outages? What's the likelihood of the wind ranges and the impacts? So really looking at the risks and taking the impacts and applying those to the communities and using visuals were possible. You can mention, okay, we're gonna get three inches of rain. Well, how does that translate to a vulnerable community with very little ability for drainage? That water is gonna pile up fast and it's gonna impact it. So having that visual of what does that mean? We look at that with storm surge, with tropical cyclones and hurricanes making landfall. Oh, it's just six to seven feet, just six to seven feet. That means that water is gonna be halfway up your door, most likely coming into your house and flooding it. Looking at the heat risk mapping on some of the products that we're doing, we're rolling out and coordinated with the CDC on a new heat risk map. We've been testing it in western region, but communicating the impacts of that worst case scenario, we're gonna be coming out in April with a graphical numerical based index, zero, no risk, four, meaning high risk and it's color coded. So it matches to what the community can see on translating that out seven days. So you know from practitioners from the health side of where you need to engage in those communities, where cooling centers are gonna need to be staged and having that advanced notification, more and more. All right, I'll stop because I can talk probably a lot on this. That's great. Thomas, and just so you know, your satellites are behind you. Great. I'd love to hear what you think and if you have anything you wanna say about your satellites. Let me first respond to the question and then I'll talk about those guys. So I totally agree, Michelle. I mean, I think there's an element of defining the impact and making it relatable and being specificity is also key, right? We have to understand what does it mean to me? What's the so what, right, to your point before? And what we see especially on the commercial side with businesses understanding this information is they're so used to kind of broad and vague and maybe too scientific of an explanation that it almost just passes over them. They make very binary decisions. Like today I'll just wholesale stop my operations or I won't send these aircraft off the ground, but in fact, we can give them even more specific information. What's the window in which I can operate or what are the activities I could get done today because those are safe while this other set of activities isn't safe. So it's getting away from using too scientific of language and making it relatable and putting it in the format of the decision that's being made. Thank you. Oh, sorry. Let me just talk about the satellites really quickly. So on the, your left, my right side is our microwave sounder. So those are, we're commercializing a technology developed by NASA and MIT Lincoln Labs called Tropics. So we sort of commercialized that system, made it a bit more robust, but there's quite a lot of data already that shows that that system is really impactful and high quality. The middle is actually what we have on orbit today. So we have two of those are called our Pathfinder radars. They're ESPA class, 75 kilogram, roughly the size of a college mini fridge. Those are our radars and they're really validating the scanning modalities of the radar system in really powerful ways. And then the far left side is our operational radar. So this is a much bigger spacecraft. It's about 350 kilograms. It has a panel array scanning system which allows us to scan a large 400 kilometer swath. So when we think about that GPM satellite that I mentioned, you know, it's swath is only 130 or so kilometers. I'm slightly off on that number, but it's around there. So we're increasing that scanning ability quite dramatically and then of course putting more of them on orbit. So again, the sort of key takeaway is reducing that revisit rate but maintaining that high data quality. So that's it. Thanks Thomas. Those are cool. I'm glad we came back to that. Thanks for bringing them back up. Pierre, thoughts about how we communicate and sort of how we evolve to explain this stuff to people who haven't experienced it before? Yeah, I would say. And that was explained very well by Caitlin and Dan, which is basically what we call the science of attribution, right? People want to know more and more. One particular event can be attributed in parts to climate change or how much more likely it would have been with and without climate change. So there's a lot of work being done to be quite quantitative about it, to say, you know, how many, I mean, and Caitlin explained that very well. So people are very interested and that's really a way to communicate how weather is being impacted by long-term climate. So there's a lot of work in that area. Thank you. Dan S? Yeah, so, I mean, first of all, I really like what Michelle said about showing people what it's gonna look like. I think that's really important and I don't think we generally do a great job of that right now. And how do you put those visuals in where people are getting their forecast? So it's easy for when they happen to be watching the local news at 11 o'clock at night or noon or whatever. But now so many people are getting their forecast from apps. How do you get those visuals into an app and what's that flooding gonna look like or that hurricane damage gonna look like? So I think that's something like sort of as a community we could work more on getting those visuals out there at the right time. And then my other comment would be with these extreme events, communicating sort of, better communicating to people, the confidence in the forecast and when you've reached that threshold and confidence, so you always want people prepared for the worst at any time. But with any specific weather event, whether it's a heat wave, flooding, a hurricane, there's gonna be a specific time and it's gonna vary from event to event where you reach a threshold of confidence that like yeah, this extreme scenario, it looks like it's gonna happen. And so how do we better communicate to people when they sort of even signaling ahead of time, here's when we think we're gonna have that knowledge. So keep checking back and okay, now we feel at this level, now's the time to take action and prepare and being able to communicate that confidence and uncertainty for extremes that we haven't necessarily experienced that often before. Thanks, and Caitlin? Yeah, I always like to think of the so what. So if I'm watching at home and I hear someone say, there's gonna be a derecho, there's gonna be 80 mile per hour winds, so what, what does that mean to me? How do I need to prepare? Well, we tell people to treat events like derichos like a tornado, it can bring down trees on houses, cars, you need to treat it as if it's a tornado. And the visualization I really do think is key as well, what is that going to look like in my neighborhood? So I think utilizing not only resources that we have at the station, we work with a couple of third party vendors as well to help create those really powerful images of what something like a derecho or a tornado or a hurricane would look like really helps bring it home. But I think answering the so what, what does this mean for me is critical when it comes to relaying that information. Great, thanks. And I know we have a question in the audience, so gentlemen in the front row, please feel free to ask away. Thank you, Dan. My name is Jared Bloma, I happen to serve as chair of EESI. And I want to just thank you very, very much for being all of you here. What you're doing is so important. I mean, the Yale University study recently on public attitudes about climate change revealed that, oh, a large majority, 73% believe the climate change is really important. A very small minority believe it affects them personally. So what you do needs to be done and please keep doing it more and more. Pierre, you blew my mind with AI. I mean, the concept of what we're able to do and where, and Dan has all brought that in as well. My question is this. It's great that you're able to tell people for a day or two or three in advance and people may get prepared. My understanding is from, in the issue of resilience and adaptation, the built environment is really key. And the model building codes and energy codes over the last number of years have tried to understand. I use that term advisedly because it's very hard to understand what does it mean, the once in 500 year event, once in 100 year events that are now obviously more quickly happening. What does it mean that it's not coastal events, it's riverine for building codes, et cetera. Are you finding the use of your material more and more for civil engineers, architects, municipalities taking it and saying, okay, what does that mean for us on how a 30 or 40 year plan for people in the city? Yeah, I would say the US Army Cops of Engineers, I mean, they are really updating a lot of their tools to actually account for climate change. Because for a long time you could just use past statistics and that was working, like as you said, like a one in a 500 year flooding event. But now we know that this is changing, this is going to be more frequent. So they are updating a lot of those things and that's been happening over the last couple of years. So that's definitely been worked out. One thing I'd like to mention, and that's kind of the plot I had, which is the fact that it's climate change, but it's also overlaid with basically people staying in regions that are being affected. And FEMA is a great example of that. We are actually incentivizing people to stay in places that are getting flooded, right? So there's also like, we need to think about incentives to actually get people out, right? Because at some point, and of course, there are psychological issues related to that. But that's an issue as well, right? More and more flooded and always at the same place. Great, thanks. Any other panelists wanna make a comment in response? One thing I'll say is that we're also seeing a lot of promise in these AI models and we're using some of them ourselves. One of the challenges I think is that as these AI models become more and more of the norm, the question is what observational infrastructure do we need to support that new form of forecasting? Because what we're building today is for these physics-based models and the very pretty well-defined needs that we know that they have. But with AI, we might need a fraction of the observations that we have today or different types and it's a question I think no one really knows the answer to yet, but it's something interesting to think about. And just following, just building on that, Thomas, of the foundation of our observation suite, like that is the foundation of making sure everything else in the value chain works of getting the improved forecast of communicating in that end result. And it is changing and figuring out what is the biggest bang for the buck, right? Of what you're gonna get the most out of, of the new technologies. It's gonna be an all-in across the entire weather enterprise. Yeah, sorry, can I say one last thing? Which is that there's another risk which is a guy in his basement or girl or anyone in their basement with a laptop who can pull an AI model and run it and spit out a forecast on Twitter with no atmospheric science background or meteorological background and no real accountability but they've got 10 million followers on Twitter or X or whatever we're calling it now. So there are a lot of challenges we have to think through as a weather community to say how do we control that or not and how do we verify the validity, you know, all of these things, so yeah. I'll pull down on that one. Sounds like he found my X feed, my blue sky feed, yeah. So we are running low on time but we started a couple minutes late so I'm gonna invite you all maybe into a lightning round. Thomas, I think it would make sense to start with you for this question and then we can go down the line and give Michelle the last word on the panel today. But what's one thing that we should be on the lookout for that from your perspective would just change the game in terms of your ability to do what you do in kind of the ecosystem of weather forecasting and communication. Yeah, well I think we've hit on a lot of trends here and a lot of really exciting changes that are coming. One thing that I always harp on is that in a lot of ways we're sort of fundamentally re-educating people about how to think about weather information, right? Like I mentioned, and when I was at the podium we've all been receiving this same more or less same format forecast for 100 plus years, right? Since forecasts were first created. Here's the probability of temperature being this and relative humidity being that. And if you're not an atmospheric scientist and you're not a weather nerd, that's really hard to know what to do with. Like I'm making simple decisions in my day-to-day life like should I cut my grass or should I walk the dog or should I put covers over my faucet so they don't freeze but when you aggregate that across like really broad areas and really complex decisions and people have very different needs it's really hard to get people to understand what to do, right? And I think the most important thing is like we've all kind of hit on is contextualizing it. What decision are you making? What does this matter to you? And I think tools like AI are gonna help us to do that. They're gonna help us see what are the things that this population cares about versus that population rather than a sort of generic broad-based information dissemination, if that makes sense. Pierre, one big thing we should be on the lookout for. I would say for me I would add to that I very much agree with the contextualization and I'm sure AI will actually help and especially large language models. But I think what's kind of missing at this stage is really like a centralized data infrastructure especially cloud type infrastructure. You can see NASA for instance has different satellites on different platforms so they are starting to work on that but that's a major roadblock for the community like trying to create this unified platform and especially now that we have to deal with petabytes of data so that's a major roadblock for the community. Dan, what's one thing that would help you do what you do a little bit better? So there are so many things but if I had to pick one I'm gonna piggyback on the AI again because I think one thing that we're gonna see sooner rather than later is something and I'll keep this very brief, very brief. It's a fancy word. But when we run models today, we run a model, you can run it one time or you can run it many times, 50 times say where you tweak the data that goes into the model of the current state of the atmosphere, you just tweak it a little bit each run to reflect the uncertainty in that information and the incompleteness of that information and it's called an ensemble. And so if you do that 50 times and 45 out of 50 forecasts say you're getting a blizzard you can be kind of confident you're getting a blizzard. If 25 say the foot of snow and 25 say a dusting well you're not as confident, right? AI has the potential because it can run faster, it can run less expensive, it's very expensive to do that kind of modeling to be able to run hundreds or thousands of those ensembles instead of today we can only run about 25 or 50 of them. And so it gives you the potential to give you a much stronger statistically robust probability of what could happen and also a better idea of what some of the extreme scenarios are and how likely they could be. And I think that's coming sooner rather than later and it's something, especially if you're involved in weather legislation that you wanna be aware of that's gonna need support. Great, thanks. Caitlin? I would love to see more products like the Climate Shift Index that really help us communicate climate attribution. Climate attribution is really hard, really complex. There's so much science, research and historical data that goes into it. But products like the Climate Shift Index make that an easier message to deliver. They make it a little bit more digestible for viewers no matter how you're consuming it whether it's local TV on your phone how cool would it be to open your weather app and see the Climate Shift Index impact on any given day. So more products like that I really think will help with climate attribution. Thank you. Michelle this gives you the last word today. Wow, last word. So yes to everything that everybody said and truly we need to embrace the technology and not run away from it. And it is only going to enhance and it's going to reshape on the forecasting and the preparedness in communities as we move forward cause it's gonna shift where our focus is. We're focused so much more kind of behind the scenes and making sure that forecast is perfect. At the end of the day, a forecast doesn't have to be perfect. It needs to be communicated in a manner that is going to resonate with that community of who we're serving and what is the most vulnerabilities of that community? What's that threshold of confidence? And going back to what Dan you mentioned about the last snowstorm in New England. I mean it was we communicating that you know what there's low confidence here. There's a shift but here's the range. Here's the most likely, here's the worst case and communities based upon what their vulnerabilities are are able to prepare for that very well. I was reading about that this morning in the Washington Post. Is it gonna be a boomer bust on Friday night and Saturday morning? So we'll stay tuned for that. Tremendous panel. I think they all deserve a round of applause. Thank you so much. We had to postpone this briefing last October and it was just killing me and I'm so glad that we did this panel today. I've been looking forward to this panel for like more than a half of a year. So thank you so much for joining us today. I'd also like to say thanks to Representative Tonko and the Sustainable Energy and Environment Coalition for helping us with the room today. Huge thanks to Representative Sorenson for joining us and sharing his remarks and his thoughts. I'd like to thank my colleagues Dan Oh, Omri, Allison, Aaron, Anna, Molly and Nicole for all their help pulling today together and assembling this tremendous panel. And our interns, Emily and Kylie. Kylie's first briefing today and thanks to Emily and Megan. I know you're not here today, but I hope you feel better soon. Also, big thanks to Scott who's helping us with our videography today. And big thanks to Lloyd Ritter for helping us, for serving as part of the inspiration for bringing the panel together today too. So thank you, Lloyd, for being here as well. This is a link to our survey, if anyone in our audience. And I know we have many people on the online audience as well. If you'd like to take the survey, we've read every response. If you have any feedback about how today went, we'd love to hear it. We will be back. I don't know if anyone has leap day plans, but we do, we'll be back with a briefing about the budget and appropriations process. You won't want to miss that. We have three tremendous panelists. And then after that, we'll be back in mid-March for the next installment of our IIJA IRA progress report briefing series updates from Rural America. We'll be looking at USDA and DOE programs that are making investments in rural areas. We will go ahead and wrap it up. Thanks to everyone for sticking with us a little bit past 4.30. Thanks to our panelists again. And we'll go ahead and wrap it up and see everybody back on leap day. Thank you.