Latest / Atmosphere with Emily Gracey / NSSL: Inside America's Severe Weather Research Lab

Transcript

NSSL: Inside America's Severe Weather Research Lab · Atmosphere with Emily Gracey

• Emily Gracey: There's a lab in Norman, Oklahoma, where scientists chase tornadoes with LIDAR, deploy radar fleets to wildfires, and send drones into the heart of the storm. There's so much that can be learned from such up close camera videos, you know, of what's going on within the storm. It's a place where curiosity drives the science and scrappy creativity fills the gaps that billion-dollar technology can't quite reach. Decades of chasing storms has sharpened their understanding of the atmosphere. Now they're turning that same curiosity toward the people caught inside it. When the meteorologists go out to the storm, the social scientists also go out to the storm. What are people thinking when a warning goes out? What do they do and why? Understanding that can help, I think, with the the fine-tuning of how Things are being communicated in different regions and being able to integrate the human element into the weather that we all experience. Dr. Pam Heinzelman is the Deputy Director of Science at the National Severe Storms Laboratory. Today she's taking us inside the institution that's been at the center of severe weather research for over 60 years. I'm Meteorologist Emily Gracie, and this is Atmosphere, a podcast where science becomes conversation. Each week I sit down with the experts behind the forecasts, the research, and the headlines to explore not just the science, but the human stories behind it. Through natural curiosity, we're making science more understandable, more relatable, and hopefully bringing a little more comfort in how we experience the world around us. The first time I ever heard of NSSL was in 1996. It was mentioned a couple of times while I was watching my favorite movie, Twister, on repeat. I had no idea what the acronym meant, but I assumed anyone mentioned in that movie must be the real deal. There are several things the movie got wrong about weather, but the vibe of this lab, that was spot on. The National Severe Storms Laboratory has been at the center of severe weather research for over 60 years. It's part of a remarkable cluster of institutions, all under one roof. NOAA Researchers, University of Oklahoma Meteorologists, the Local National Weather Service Forecast Office, and the Storm Prediction Center. The people who study storms and the people who warn you about them are quite literally down the hall from each other. What they do is cool, like nerd cool. Back in the 90s, when Twister came out, the script was actually inspired by real work being done at NSSL. Remember Dorothy, that contraption that Helen Hunt and Bill Paxton were trying to launch into a tornado? That was inspired by Toto, a real life and slightly less sexy version that NSSL was testing at the time. Fast forward to two summers ago, it happened again when Twisters was released, shining a giant movie spotlight on the super cool work of NSSL scientists. One of the scientists behind that work is now running the show. Dr. Pam Heinzelman started as a research scientist. She earned her PhD in 2004 and worked her way up through phased array radar, storm modeling, field programs, all of it, before becoming the Deputy Director of Science. Today, Pam is giving me the rundown of the latest in severe weather research. A new fleet of mobile radars, a close-range tornado project called Lyft, what a high-speed camera pointed at falling hail can actually teach us, and why, for the first time, social scientists are riding along to understand how people think and act when a storm is bearing down. Dr. Pam Heinzelman from NSSL. ⁓ we haven't met, so this is a great honor to meet you. You are the deputy director of science for NSSL. Is that correct? Yes, that's correct. It's a pleasure to meet you as well. What do you do as the deputy director of science? I get to work ⁓ with all the scientists across the entire laboratory. So on a a lot of different topics related to severe weather and you know help to set the science priorities and support the scientists and you know make make all the exciting things that we do here work and move forward from doing work in the field to ⁓ producing new forecast tools and then getting them into operations, you know, to to help forecasters on the public. So it's it's really an exciting ⁓ place to be. It seems like it seems like the fun stop over there in the National Weather Center because you're kind of in this epicenter of not just research, but like a lot of stuff is going on right there as well, right? There is. It's it's such a unique opportunity because we're co-located with the University of Oklahoma School of Meteorology. So ⁓ we see the students, they see us, ⁓ we work with them. We have students who work with our scientists I advise them through their graduate programs, for example. And ⁓ right outside the door from me, we have the National Weather Service office in Norman, Oklahoma. So those are the ones on the front lines, you know, deciding when to issue warnings for flash floods and tornadoes, ⁓ severe storms. And then we also have the storm prediction center, which is right down the hall. And you know, they're the ones looking at the bigger picture across the whole nation. deciding when watches will be issued, for example, ⁓ for tornadoes, which is one of the things, you know, that we're really looking at at this time of the year. Yeah, I know there's a lot of like official collaboration among these different groups. Is there like unofficial collaboration that goes on too? Like do they ever pop in and say like, hey, I'm looking at this storm, like what do you guys over here doing research think of this? Is there just like chatter in the hallways? There's definitely chatter in the hallways, especially on big days. You know, when there's there's a lot that could be happening. But one of the places where we really work together the most is a space that's located right between the Storm Prediction Center and the Norman Forecast Office. And it's called the Hazardous Weather Test Bed. So that's a place where, you know, we might be running ⁓ a mobile radar or a phase ray radar, and we are Piping the data to the forecast office so they can see it. And then if they want to come in and talk about something they're seeing on radar, you know, they can do that. We bring forecasters there to test new tools to improve warnings, ⁓ for example, or you know, providing information between watch and warning to, you know, keep people informed along that timeline. And so it's ⁓ we even bring in people, ⁓ not only forecasters, which you know, Storm Prediction Center, Weather Service Offices. But also academics and private industry as well. And that's that's starting early May to look at the latest forecast models. And so, you know, there we're looking at ⁓ short-term forecast models out to a day, as well as some of the new AI type tools. And then it's really a ⁓ you know, place for testing what is the future going to look like in terms of models that are going to be used operationally in the National Weather Service. You, I want to hear a little bit about your background. ⁓ you know, I've interviewed a lot of men from NSSL. You're the first woman. When you were a kid, was this your goal to be doing severe weather research? Yes, absolutely. ⁓ so I I've wanted to be in severe weather really since I was a child. So I've always had an eye to the sky. You know, enjoyed looking at all the different clouds and the wind and you know, I even built a little hydrometer when I was a kid. Like for a science fair, one second prize. You know, I always wanted to understand better what's going on. You know, a lot of curiosity, you know, why are tornadoes forming? Why is hail forming? Why is it so large in some storms and not so large in others? And so that is really taking me ⁓ from being a a research scientist, ⁓ which I earned my PhD in 2004 and started working with phased array radar and sampling all different kinds of storms and looking at How ⁓ adaptive scanning can help with ⁓ forecasts, short-term forecasts and warnings of ⁓ tornadoes and other severe weather, up to then working with Warnon forecast, which is a modeling endeavor. And ⁓ got to really bring that into our test bed environment and work with the weather service forecasters, which I always love working with people and testing these tools. And then moving on, you know, to more of a ⁓ you know, entire lab re research ⁓ leadership role. And so kind of worked from the the ground floor up, I think you could say here. And it's ⁓ it's been really fun to ⁓ do that across my career. Awesome. ⁓ okay, well, I want to ask you about some of these individual projects going on at NSSL and and dig into a little bit of the research that you're doing right now, starting with you mentioned the radars, but you have a new fleet of mobile radars, correct? Yes, we do. We have a fleet of two X-band radars and one C band. And ⁓ so yeah, they're bright and shiny and our first. Our first goal is to test those. So taking them out in the field, you know, as storms come our way and test, you know, what what do we want our mobile operations to look like, seeing how, you know, the quality of the data, you know, prepping to be able to take them into longer field programs, ⁓ as well as testing their capability of just kind of sit and stay and observe for a while. And ⁓ you know, be able to run those remotely so that our our scientists don't always have to come out, you know, in the middle of the night and and can run those from home. Interesting. Cool. ⁓ I've been surprised over the years in some of the ways that mobile radars have been used, like unexpected ways. Can you touch upon some of those? Because I I mean we've seen twisters and like the way they're used there, but what are some of the unexpected uses for them? You know, what are the great things about having mobile radars and especially, you know, something like an X-band radar can really look at the wind field within storms. And so you can get a lot closer to a supercell that might produce a tornado, for example, and focus on the low level wind field, which is really critical to understanding which supercells might become tornadic and which ones may not. So I think that's a really critical aspect of that and having multiple radars and you can bring multiple to the storm at one time because each radar can only look, you know, you kind of get radial views of the storm that are like perpendicular to the radar. You bring in two radars, you can get a much better perspective on the wet wind flow. ⁓ but other things that we've taken the radars to and we're excited about doing with this fleet are even taking them to like a debris flow situations. Yeah. So, you know, when you have fires, ⁓ we can bring these to fires as well, you know, to sample ⁓ the characteristics of how fast are they spreading, how quickly are they moving, ⁓ what's the intensity of the plumes associated with that, and you know, convection that can result from that selectional storms and even rotation within fire. And then after the fire, you know, the the landscape is just so different. So you get a heavy rainstorm, you know, in some situations and you can have flash flooding events. ⁓ so, you know, we can sample close to these fires to understand what they're doing, you know, the damage, the directions they're moving. And then also look at after the storm, you know, what what are the rainfall rates and drain impacts after a fire and the flash flooding that could occur and yeah, really be ⁓ dangerous, you know, to life and property. Are there holes in radar coverage now, like just places that are not as populated that could use, really utilize mobile radar units like this? Yeah, so there definitely are. ⁓ the you know, the radar network that's around the country is driven mostly by the higher population centers, you know, where they're located. And so there are some in private industry ⁓ who are bringing radars to some of those gaps. But we can also bring our radars to those gaps to understand, you know, like the network of radars. They're all S band radars. And like that's what the Weather Service is used to using. And so we can look at, okay, if we sample storms, you know, in this gap with an X band or C band, you know, what additional data is that providing? Right, which is really important to fill those gaps and also looking at how that additional data might help models perform better. But also, what do those signals look like? And how does that inform what a radar mosaic across the country will look like? And you know, the usability of the data filling those gaps. ⁓ I know the Dow Doppler on Wheels folks. they strive to get to like hurricanes too as they're making landfall. Is that a goal of NSSL to kind of get these into hurricanes too? That is another goal of ours. We have a long list of ⁓ different phenomena that we would like to sample. We we've gone to hurricanes ⁓ before but I think with with these radars, ⁓ what we will be focused on for those kinds of events, it's really the smaller scale things that are happening there. So Yeah, like where are the tornadoes spinning up, you know, within the hurricane structure and you know, sampling on those smaller scale features that you know really are within our line of expertise and that these radars, you know, should be able to sample well. Gotta be hard to kind of get to the coastline too. Norman Well, you you need a plan ahead of time, right? So you need a good forecast and you know, be ready, be there, yeah, to go on the storm. Yeah, yeah, yeah. Yeah. ⁓ okay, let's talk about the Lyft project. I love NSSL's ⁓ always great with all of the acronyms. ⁓ Lyft. What does Lyft stand for? And ⁓ I saw close proximity, close range studying of tornadoes. That sounds a little intense. Yes. So ⁓ lift is low level inflow into tornadoes. So that's what it stands for. ⁓ Do you guys come up with the acronyms first and then fill in the letters later? Like Vortex, that was like too good, right? We always try to be creative and have names, you know, that catch on but really capture the essence of what we're doing scientifically for sure. I think that's part of the fun of being a scientist is getting to come up with the names for the field projects. ⁓ so yeah, there ⁓ we ⁓ have ⁓ what's called a LIDAR that we bring. very close to the storm. And that's able to sample the winds even better than some other instruments can. And so we sampled our first tornado last year, like the complete cycle of a tornado tornado with LIDAR last spring. And so we're analyzing that data and you know we think that a lot of the process that's important is in the details. that haven't necessarily been sampled, you know, much before. And so that's why we're bringing this instrumentation to the storm to be able to understand, you know, the wind flows that are producing the tornadoes. And of course, you need to do that with a lot of different storms because you don't get the answer from you know just looking at one. What's the difference between how works and how LIDAR works? A LIDAR is like a radar, except that it's more sensitive to ⁓ sampling particles, you know, like the air particles ⁓ versus particles from precipitation. I have questions about drones and tornadoes because my first thought is like, how is this little thing surviving that? Because these are, you know, phenomenon that like do so much damage. How is this little drone surviving a tornado? So, you know, you have to have a ⁓ a probably a very ⁓ well-formed drone that can handle high velocities. I think one of the the opportunities with that video that we saw is because it was a low precipitation supercell. So you can see the tornado without a lot of rain around it, ⁓ which you know can make it a little bit easier to do that. ⁓ but they need to be tough and be able to fly in high winds. And I think drones that we have for sampling ⁓ the boundary layer, you know, against those low level winds, you know, that really matter. They can fly up to fifty-five mile per hour winds. ⁓ but you know, really making them ⁓ able to withstand, you know, the ⁓ severe, you know, the extreme environments in which tornadoes form is is really key there. But There's so much that can be learned from, you know, such up close camera ⁓ videos, you know, of what's going on within the storm. Can we talk about hail for a minute too? Cause it seems like there's been a lot of research on hail in recent years and it's maybe not quite as fatal as tornadoes, but certainly costly. and at the AMS conference this year, I saw the cool hail. Vehicle that Sean Wah brought out. ⁓ like his like MacGyver approach to all of this. It seems very scrappy. Can you tell me where we are with hail cam and and what's been learned so far with it? Yeah, so hail cam, I think it's a unique instrument. ⁓ so you know, we have this truck and we've got a cage over it so we can bring the truck to the hail and not have the truck be totally damaged, you know, keep safe inside there. And then there's a really high speed. camera with ⁓ a lot of bright light that can capture the hailstones as they're falling to the ground. And yeah, those the videos from that thus far are pretty impressive seeing, you know, what the hail actually looks like because, you know, when it's on the ground, it's, you know, we don't know, you know, to what degree has it fallen and broken up, you know, broken up on the ground. ⁓ what are you know the different shapes, you know, it gets to the ground, it's already starting to melt. So being able to capture the hail and hail fall helps us to better understand what, you know, hail is it's falling to the ground and impacting things, you know, whether it's a roof or your car or, you know, hopefully not someone out in the storm getting hit by hail. ⁓ you know, what is what is what does that actually look like? How does that compare to signatures we see on weather radar? For example, to help verify, you know, because we have signatures that are associated with hail size, but we don't always have really good data to verify how those signatures relate to actual hail on the ground. And so that's one of the focus areas for us is to better understand those relationships. Interesting. Okay. So like a more precise forecast and knowing. This is very likely this is what we're seeing on the radar. This is what you're going experience. Cool. Yeah. There's so many things about hail that are interesting and unique and different from other types of hazards. Okay. Well, speaking of warnings, ⁓ let's talk about warnings. I know you have like a history in this and kind of dabbling in the social science side of things as well. I guess the goal with all of these really cool projects is. warnings and getting that information out and saving lives, right? So that final piece I guess is like how to translate this to what people understand and make the appropriate action. So where does NSSL work when it comes to that part of the puzzle? Yes. So I think we're all highly motivated ⁓ by that, you know, being able to make a difference through the the science that we do. And so much of what we learn, we develop into tools to help forecasters. And ⁓ so sometimes there is it's a new conceptual model that comes from the field, you know, that forecasters can use in their forecasting. ⁓ but then it's also, you know, how do we take the data that's out there, bring it into a model, and then be able to forecast, you know, what's going on on the short term. And that's where warn on forecast comes into play. And it is focused on producing forecasts every half hour, out to three hours at the bottom of the hour, six hours at the top of the hour, ⁓ over a 900 by 900 kilometer domain, which is about the size of the four quarter states. And so it provides ⁓ probabilistic information about. You know, like what's the probability of a tornado occurring, a severe storm occurring, ⁓ you know, hail, different rainfall rates, ⁓ things like that. And it's really focused on the time period between when the storm prediction center is issuing a watch and then the weather service putting out a warning. And so, you know, currently, ⁓ You know, prior to Born on Forecast, there was not model information that was coming out on a regular basis that could provide information for them to ⁓ put out products in that kind of like tweener zone, you know, to ⁓ to keep people informed. ⁓ and especially people who are out in the field, like emergency managers. ⁓ for example, they've used products ⁓ from the weather service during this, you know, the between timescale. ⁓ to decide to ask their community to say seek shelter, for example. So, you know, they received ⁓ information, I think it was it was one or two hours before this was a storm that was in the Missouri area last year. And so they asked their community to seek shelter. They did seek shelter. There was a significant tornado that ⁓ happened in their area. And so they're able to know provide this direction to people in the community, whether it's a school or you know, perhaps people in mobile homes, you know, who need a place to go and they need time to get there. And so those are the kinds of situations where providing ⁓ this more like specific information about where are the storms going to be, how are they going to move, how intense and impactful are they going to be of, you know, between the time the storms form and then a warning is required. Hmm. I've never really thought about that tweener time before and how big it is. Do you think that one day there'll be like an additional alert during that time or will warnings get like stretched out longer and fill that gap? Well, it's interesting because I think both approaches are being explored. ⁓ so you know I think there's You know, the the opportunity and already the weather service is starting to put out graphics that will show like this is the area of high probability, you know, of tornadoes. Keep an eye out, you're like in the next two hours. ⁓ and so we've seen a lot of different you know, graphics like that, or even for flash flooding, ⁓ you know, hail, all all different kinds of things. And then at the same time, we have this ⁓ project that's called probabilistic hazard information or fear. fee and so that's more observation focused whereas more on forecast model ⁓ the is more observation based machine learning type tool and so the goal there is to provide probabilities on specific storms out to an hour that they could you know what is the probability they could produce a tornado, hail, strong winds, those kinds of things. And so these ⁓ plumes kind of look like storms or like object plumes will go out ahead of the storm with this information associated with them that forecasters are now starting to look at. And so ⁓ they're able to have perhaps more confidence looking at the trends from those probabilities, which can enhance confidence and getting a warning out. ⁓ but then they also have tools now that. Have been developed that provide the capability for warnings to expand like as the storm moves. So like the warning could move and then, you know, expand with time and its size, depending on you know so what's going on. And so that's a new tool that is being tested and ⁓ with the National Weather Service. You know, ideally that will become just another, you know, thing that can help. people when they're seeing these warnings on the screen, you know, whether it's a broadcast meteorologist or on an app, you know, you can see where am I and how is it moving? What is its intensity related to where I am, right? Interesting. Yeah. So it's it's exciting. Yeah. Yeah. That's very cool. And it in your career, because you've been doing this a minute, like what has been the evolution of just like the precision, the warning time? What is like you've really noticed has been like a huge transition to something so much better than it was, say like 20 years ago? Forecastability is, you know, one of the biggest things. ⁓ like being being able to really hone in on that period, you know, an hour out to maybe six hours out ahead of the storm. And I think, you know, that's because ⁓ You know, we've learned better how to use the data. We can use it more quickly. So, you know, we data comes in rapidly, but you need to be able to take advantage of that. And so a lot of the tools that have been developed really bring all the data together in a rapid way so that it can be ⁓ you know analyzed or you know put into a model and used to provide you know information quickly. Like with convective models, like I was talking about with Warren on Forecast, that's all about bringing the data fast, being able to run the model fast, and getting all the products out fast, right? You can't have it take an hour to to get out information when storms are evolving right here in front of you. So I I really think being able to make good use of data models in a very timely, efficient way, ⁓ that is really representing the structure of stores, that that is a huge leap over the the ⁓ length of my career so far. Cool. Where do you think the most need is when it comes to future progress? Like it seems like forecasting has gotten so good and so precise. Like, is it social science? Do we need to understand decision making better and like communication better? Or is there still work to be done kind of in the forecasting side of things too? I think it's both. ⁓ so ⁓ here at the lab, we also have social scientists ⁓ who are working ⁓ to understand, you know, they help us in the testing of tools because how you design experiments to test them is like really important. And also ⁓ learning, you know, when when when storms are approaching, how are people thinking about their risk? Right. So there's there's been a lot of focus perhaps. After the storm, you know what happened? People go out and do damage surveys or maybe, you know, interviews with ⁓ people afterward. And of course, you know, we always wait a while because we want to be very respectful, you know, of people who have ⁓ experienced these ⁓ types of events. But one of our our our things we're focusing on now is like when the meteorologists go out to the storm, the social scientists also go out to the storm. So while they're Setting up to figure out, you know, where are we going to collect data? You know, they can talk to people in the local area and learn about what are you thinking about? What are your concerns? What have you heard about the weather today? And then ideally be able to understand kind of like that decision process from before the storm to all the way after the storm. And so, you know, understanding that can help, I think, with the the fine-tuning of how. Things are being communicated in different regions. And we're really excited about that work and you know being able to integrate the human element into the weather that we all experience. Not always coming naturally to a bunch of scientists, but it's a definitely a good thing to have in the back of our minds. Yes. Yeah. Well, I mean, yeah, the concept of having social scientists in the field, you know, talk about another change in my career. That's another big one, really the the integration of social science science into what we do and you know, understanding the importance of that, really being involved at the beginning, you know, of different types of projects that we are seeking to tackle. So important. It was not even remotely a part of my weather education. So ⁓ it's it's great to be learning about it now though. Pam, is there anything else you want to share about research going on at NSSL? I would share, like right now we're active in the field. So this is our really busy season. ⁓ that's when we are testing a lot of our new models. You know, like I was talking about, we're on forecast, ⁓ probabilistic hazard information. It's really getting a lot of eyes on both of those. ⁓ from forecast centers and ⁓ other places as well. Like a lot of those tools are available on the web so people can look at them. ⁓ spring is our our very busy season and that's when we learn so much. We gather so much data of that then we can learn from the rest of the year. Awesome. We can't wait to see all the information that comes out of this. Yep, I'm excited too. Thank you so much for talking to me today. This is very educational and ⁓ fun too. Fun for me as well. Thank you, Emily. Appreciate the opportunity. Thank you so much to Dr. Pam Heinzelman for joining me on this episode of Atmosphere. I also want to extend a huge thanks to the folks at NSSL and the NOAA public affairs team for helping set this interview up. If you have questions about today's show or weather in general, you can always find me on social media. On Facebook, I am Meteorologist Emily Gracie. And on Instagram, you can find the show at atmosphere underscore podcast. And guess what, guys? Atmosphere is now on Substack. If you want to take a deeper dive into the science and the stories we cover here every single week, please head over and subscribe. It's free. If you enjoyed today's episode, please follow Atmosphere on YouTube, Spotify, or Apple Podcasts so you never miss a new conversation. Leaving a rating or review takes just a second, but it truly helps support the show and can help other curious minds find us as well. And of course, if you loved it, please share this episode with a friend. Atmosphere is a production of Mike Burst. It's produced, edited, and hosted by me, Emily Gracie. If you're looking for even more weather content to close out your week, make sure you check out the weekly weather brief, a quick rundown of the week's biggest weather and science stories hosted by me and my fellow meteorologist Karen Jerriman. New episodes drop every Friday on YouTube, Spotify, and Apple Podcasts to catch you caught up before the weekend. I'm meteorologist Emily Gracie. Thanks for taking the time to slow down, stay curious, and explore the atmosphere.