Episode 6: Craig Clements

[Music]

Vincent del Casino: Hi there, and welcome to The Accidental Geographer. My name is Vincent del Casino. I'm the provost and senior vice president here at San Jose State University and the host of this podcast. On today's podcast, we have Dr. Craig Clements, a professor of meteorology and climate science here at St. Jose State university. This is gonna be a really interesting conversation as we talk about the intersection between fire and weather and the emergence of fire weather research here at Saint Joseph State University, a place where we lead. Not just nationally, but globally, in the study of fire and its intersections with everyday life. So come on board. This is gonna be a great conversation. Craig, thanks for being here, I really appreciate it.

Craig Clements: Thanks for having me.

Vincent del Casino: So I always like to start with a sort of big picture question of why the questions, the sorts of things you're interested in, the intersection of fire, weather, how did you get there, what motivates you to do that work, what are the things along the way when you went, oh, that's the question I'm interested in or those are the thing I want to pursue.

Craig Clements: That's a tough question. I actually wanted to be a glaciologist. I didn't want to study fire. That was not my PhD. But either was glaciology. Anyway, so the way it happened was actually very accidental. So I had a field site as a PhD student in Houston, Texas. And I was in charge of building the site to do micrometeorology measurements, so like fluxes Heat and moisture and all that jazz and stuff. So we set this tower up, and then they did a prescribed fire around it. And it was just gonna be like a little chapter in my thesis, like, oh, that's pretty cool. And since my background was in mountain meteorology, I had a lot of friends in Cal Fire and a lot firefighter colleagues are like, why don't you do fire weather? And I was like, eh. Anyways, we did this experiment by chance. I presented at a conference and I had so much interest. I never had that much interest in anything I represented and. I just decided, I'm changing my PhD topic and did this experiment that became an international benchmark for fire model validation. And the reason I did that is I did a quick search and nobody had really collected those data before. And that's what started the whole thing. And I figured, gosh, if I could get back to California, fire weather's gotta be big. It just would be a good topic. And so I changed my whole topic.

Vincent del Casino: Yeah, that's interesting. Well, it's nice that you had the support. I mean, you have an advisor or something like that because that doesn't always happen.

Craig Clements: Yeah, it was a situation where I wasn't really assigned to a funded project. And I was funded to build the site. And I capitalized on the fact that I had infrastructure to measure, use measurements. And that was my background. And so that just continued. And that just keeps going. And so, that's what we're still doing.

Vincent del Casino: So tell me a little bit about what you found in that early work. What were the things when that became a benchmark? What were you looking at that other people weren't? What were thinking about and what did that mean for the kind of emergence of firesides?

Craig Clements: Yeah, no. So what we did is we set this, we had this tower, 150 foot tower, and in a grass prairie. And grass prairies in the United States, especially in the Southwest and in Texas and Oklahoma, they burn regularly. And so prescribed fire is really important for those ecosystems. And we had the tower and we were just like using it for measurements and, you know, not really anything that exciting. But we had it. And I was able to build it. And once we did that experiment, we were able to look at CO2 fluxes, because that was a big thing, looking at CO two flux from forests and different types of ecosystems. And even though that really wasn't my background or interest, I had the technical expertise or experience to do that. And we had never looked at that in a fire plume before. And so then... What happened was we decided, well, let's look at some of the flow interactions around the fire front because that would be very interesting. And again, literature of search, like nobody really has done this. And what we found at what we call this fire flux experiment, and we basically let a fire burn through the instrumentation as hot as possible. And nobody had really done that successfully. At least it was never published. And so these data show like these circulations around the fire front. So we call that micrometeorology. Typically, micrometric would be like flows around inside a forest, the moisture and heat fluxes out of a forest very small scale or flows around a building. This was looking at flows around of fire front and so we showed that there was down drafts behind the firefront, the updrafts and so able to just get these point measurements. For the first time. And what happened was a bunch of fire modelers were like, hey, can we use that data? Those data are really cool. And we'd need it for a model. And I said, sure. So people started using those data to validate their model. And what they could do for the time is the model generates updrafts and downdrafts from the heat flux in the model, but they could never measure that. So now we have these updraft circulations and dowdrafts that they can match their Model 2. And so that's how that became a benchmark. And then from there, we kept doing these experiments and eventually got into measuring those on wildfires.

Vincent del Casino: Got it, so you start with the controlled burn, and there's a, I mean, the US history relationship to fire and forest is fascinating. There's like this complex debate of forest cleaning, you know, what do you do, prescribed burn, versus natural ecosystem sorts of things, the ways in which human intervention, and even that, may have escalated fire over the centuries, right? I mean there's a rich history of fire science but what I hear you saying is literally nobody had ever really got into the middle of a fire in that way.

Craig Clements: In the field scale most field scale experiments were like let's time the fire front We know the fuel conditions. We know, the weather conditions, but and we want to know the spread rate Versus the structure of the plume the structure, of the flows the circulations around that fire and Yeah, it's again by luck that this thing just kind of landed

Vincent del Casino: So is a geo, PhD in geophysics, right?

Craig Clements: Yeah, atmospheric science, but.

Vincent del Casino: So you kind of became, for lack of a better word, passionate about instrumentation as well, or you always had that, that kind of like...

Craig Clements: I always had that and actually when I moved to the University of Houston, I followed a woman to be her first PhD student and I wanted to be a modeler. She's an expert in modeling the mountain atmosphere and I was like, that's what I want to do. But I had a lot of background in instrumentation just as an undergrad, my undergraduate thesis was in Yosemite. I started. Pilot balloons or weather balloons and getting observations. And so I was always an observationalist and I didn't really know what that meant when I was an undergrad or experimentalist because I wasn't from a physics background. I actually studied geography as my bachelor's. I know, I know. So I am an accidental geographer if anything.

Vincent del Casino: Yeah, yeah. Well, it is interesting. There's a within the within the field of climate science, there is the sort there's the modeling community. And then there's this the on the ground applications group, the instrumentation group. And obviously, those work together in certain ways, a lot of ways. But how did the work you were doing on the ground you you talked about this, maybe flesh this out for me a little more specifically, how did that work inform some of the modeling work that was going?

Craig Clements: Yeah. So that's kind of where my path intersected with a lot of modelers, is that we collected these data. And based on that early experiment, bigger programs were built. There was a thing, a project called the ARX cadre, the Prescribed Fire Dynamics Project or whatever. And so I was brought on there. So I started interacting with a bunch of different people, Theracologists. Geographers, engineers, and collecting data at the field scale became more popular and because modelers realized that we could collect data that they needed to develop their models. So there is this interaction, this intersection between the modelers and observationalists and we're still doing this today. We still have, we have an NSF workshop coming up. Where the modelers are gonna tell us what they need and we're gonna tell them what we can measure and it usually doesn't always correlate very well. But there's usually an overlap but not that well. So, because you want as many data points as possible and it's very hard to measure. The atmosphere itself is hard to to measure because it's expensive because you're trying to measure the air above the ground. You need aircraft, you need towers, you needs lasers, you need radars, you all that. But the modelers, they want that point on every grid point of their model, and we can't get that. It's impossible to get that many observations. So what happened was, to get back to your question, we started working with modelers saying, hey, modelers need really good infrared imagery of the fire front, we need really great rates of spread, and we need to know the wind flow around the fire perimeter and within the perimeter. I think I'm the one that kinda kept pushing and we need. Measurements inside the fire right and there's a limitation that because fires melt instruments and I've melted lots of instruments

Vincent del Casino: I imagine there's an expense rate.

Craig Clements: Yeah, and the funny thing was, when we did the Firefox experiment, I changed my PhD halfway through, and I dropped everything I did and changed it. And, I mean, it was a quick turnaround.

Vincent del Casino: It's unusual.

Craig Clements: Yeah, I ended up here a year after that experiment. And that is pretty crazy. And I just, I don't tell a lot of people that because we're like, what? You didn't postdoc for five years? And anyway, so what happened was, when I was standing there in this fire, they lit this fire and I remember the burn boss and this is my first time really, my second real, second or third fire I've been on. And I had all this instrumentation set up in this field and these guys were like, are you sure you want us to run the head fire through there? I'm like, yes. It can't be that bad. And as soon as they lit it, the flames were actually, the flame front of a grass fire is, people think grass fires, that's not a big deal, but they're like 20 feet tall.

Vincent del Casino: Yes, yeah.

Craig Clements: And there's a lot of radiation, there's lot of convective heat, not as much convective as you get on a forest fire, shrub fire, but the flames are big and they're intimidating and they move very fast and I thought to myself, ooh, I'm probably gonna be in trouble here. I might damage a lot equipment, this may not go the way I want it. And I thought, well, I have weighted tables for 10 years, I got that CV as well. I mean, you know, this is a chance. It was very experimental for me because I didn't have enough experience in wildland fire to know that this was kind of trivial. Now that I've done probably 50 of these to 100 of these prescribed fire experiments in my career, after 20 years now, which is kind of a scary. It's like, wow, OK, this stuff's kind of, I don't even care about grass fires anymore. But they are still really fascinating, because they're actually quite dangerous. And most grass fires end up starting larger fires, because they are easier to ignite. Anyways, I got off topic. Modelers found out about these data. I had them. We improved the experimental methods. It still wasn't good enough, and now we've scaled up to larger experiments.

Vincent del Casino: So I was actually living in Canberra in 2002 when it lit and it was grass. It started with grass and then it was California pines because they had grown all these pine farms around. And scarily in 1991, I was in Oakland when the Berkeley Oakland fire happened and it Australian eucalyptus and I started to worry that maybe there was some relationship there. I haven't yet. Where are you going next? Well, it is a question people people should be asking. So but I remember that fire in Canberra because it started and it was it was a wall of flame 30, 40, 50 feet in the air.

Craig Clements: Most of the understanding of grass fire dynamics is from Australia.

Vincent del Casino: That's not surprising, given and camber's interior, you know, and it's just surrounded by that sort of thing. So that's a little off topic, but talking about the grass stuff, I was thinking about that.

Craig Clements: I was in California, I remember seeing the plume from the Oakland Hills fire, the tunnel fire.

Vincent del Casino: I was, well, again, a little off topic, but it is interesting. I was teaching elementary school in Oakland, and I was a runner at the time, and I running up the mountain, and I saw that black smoke. I went, what, did a car explode up on the, and then, you know, then it all went. Yeah, that was. So, let me get back, I guess, to the topic, it's not being autobiographical. So. The work on the mobile, the work you've got. So you've taken these ideas of getting close to fire or in fire. And you literally, in some ways, for lack of a better layman's way, chase fire in a truck.

Craig Clements: Yeah. Now we have two trucks. We're getting a third. We are getting a third now.

Vincent del Casino: You're getting a third truck. Tell me a little bit about what you're trying to do. Because you've explained some of this to me in the past, and I find it quite fascinating of what you are measuring and why there's a value add for that work. So how you got to that sort of mobile strategy, what you were doing in that mobile strategy and what you've found from it that's helped kind of some of the larger discussions.

Craig Clements: Yeah, so we were doing these field campaigns and we'd fly stuff, you know, we'd go back east and they burn a lot in the southeast. So we'd have to go to the southeast and set up our towers and stuff. But I was always interested in measuring bigger fires. And so I think it was like 2012, we got, it was 2011, we got this MRI grant for a Doppler lidar and a bunch of other equipment. And we decided to, or I decided like, hey, let's take this stuff and try to measure some fires. So we built out the truck, put the lidar in the back, and the lidars is like a radar, but it's a laser-based system and we can point at a fire and get wind profiles or scan the plume. And I was lucky enough to have a post-doc here at the time. I said, this is what we're gonna do. We're gonna actually chase wildfires. We're going to go collect these observations because nobody's ever done it. It has not been done. And we were, you know, small outfit here, San Jose State, just one truck, a couple of people. And, but we had a lidar, which is quite unique. Now we have seven, but that's all another conversation. But we had one of the first lidars in California, for sure, and probably in the Western US. So we went to fires and we started sampling wildfire plumes. And part of Neil's post-doc research was taking those observations and putting them in context of plume dynamics. And so for the first time, we actually had observations of real wildfire plumes that were. Generating pyrocumulonimbus so very tall plumes and so the lidar wouldn't penetrate those because of the ice and the water Droplets and the cloud droplets so lidar doesn't penetrate too deep into these big plumes

Vincent del Casino: But it can measure the three dimensions.

Craig Clements: It can measure, it can if you can scan it fast enough. Radars are better for that, but we can actually get at least the plume height and structure and the winds around it. So what we found was that we can sample these wildfires in really high detail and then the modelers are like, wow, that's really cool because now we can see this. There's been times where people have flown aircraft with LIDAR and gotten smoke concentration or smoke structures, but not really. The plume core. And so that was kind of exciting. So we're like, okay, let's go bigger. Let's write a proposal and try to get more funding and more assets to do this. We submitted it a few times to NSF and we just couldn't get through the peer review. I don't think science is ready for it. I don't know, either that or people just, and it was well thought out. And now we're at a stage where everybody wants to do this. Now it's actually building a whole program to chase wildfires. And based off the work that we did. And so that's cool. So we sampled over 50 wildfires in California. And another exciting thing that leads to the next development of our program is that because I had an NSF career award, I had access to NSF assets. And so a colleague of ours reached out and said, hey, have you ever thought about flying an aircraft with a cloud radar? Like, that's kind of out of our, you know, that's not really my expertise, sure. And so we wrote to NSSF. We got some test hours on the Wyoming King area. The University of Wyoming has a NSF facility. And they flew it over a pyrocumulonimbus and threw it. And the cloud radar was amazing, like detail and structures that we had never seen. We saw updraft, we measured, we got a really nice paper of this, updrafts of megafires paper that a grad student of ours, of mine. 58 meters per second. So like 125 mile an hour updraft wind that's well above the ground a couple kilometers. And so these structures we didn't even know existed. This is why you don't fly through those.

Vincent del Casino: Yeah, it sounds incredibly crazy.

Craig Clements: Yeah, and so now there's a campaign. So then we went back to NSF and said, hey, look at these data. And we applied for another MRI and got the radar. And now that's our baby. We have the only mobile radar in the Western US. And it's a cloud radar, so it's specific for looking at cloud droplets and ice crystals. But it's amazing at ash and debris and wildfires. And that's really what we use it for. So that is a really cool asset. And as not being a radar meteorologist, because most people that use radars, they study radar meteorology. They come from that background, from severe storms and such. And I'm like, it's just another tool in my toolbox that I can now study phenomena that's very interesting. So that led us to that.

Vincent del Casino: Yeah, it's interesting. Clearly, in your career, what's the tool at hand that might go and answer a question? What's also interesting in that point, and I think a lot of people might be watching this or thinking about it, or students who are, like the idea that because you had a career award, which is a five-year award, it is pretty unique, they're hard to get, that the way that research is actually organized in the United States is that NSF has these sites that are meant to be used. By everybody. So the plane, we used to have an asset down in Moss Landing, we had a ship, right? So there are these kind of assets that help science move forward. And, and, and that's like an important part of this is even though you're doing this work at San Jose State, I know you've had partnerships across a number of different institutions over

Craig Clements: Yeah, yeah, no, I work with, you know, and actually wildfire meteorology is there. Fire weather community is pretty small. It's usually, in terms of research, it's not small anymore. I mean, it's crazy. Like, early part of my career, getting a small grant to do this type of work was like, oh my gosh. And now there's just a lot of funding and a lot money being thrown at fire, which it needs to, it has been neglected for decades. So finally we're getting to the stage where Agencies are putting money into the research, right? But yeah, so I've worked with a number of colleagues Still a small group and one of those colleagues was a modeler Adam Kochanski Who then joined us when we started work? Yeah, and so he is a the one of the world's top Fire modelers and you know very lucky to get him here. Yeah

Vincent del Casino: So work, I was going to go there next. So you perfect segue, because people won't know what it is. Yeah, so explain it. But I'm really interested in the I part, the nature of interdisciplinary

Craig Clements: Yeah, that's a...

Vincent del Casino: So explain the name, what the center is, what you're doing, and how you've brought some people together.

Craig Clements: Yeah, so what happened was I always said, you know, we gotta expand this. I can't do this myself. You know, I'm only an observationalist. We need modeling. We wanna, you know. I know folks that can really do, and I know the community well. I think we need to build something bigger. And so I remember I asked for, I think it was right when you came here.

Vincent del Casino: I was right in my first year.

Craig Clements: I said, hey, I really wanna get a cluster hire of wildfire scientists to start a center. And we were allowed to do that, thank you very much. And part of that is, the proposal was the Wildfire Interdisciplinary Research Center, or WIRC, W-I-R-C. And part that was getting the modelers, but also getting social science, remote sensing folks, engineering and fire dynamics, and fire ecology, and fire management. That's kind of the interdisciplinary aspect. Now, do we all work on the same projects together all the time? No, but we have a few that we all work on together, because it is still hard to mesh different fields together. But if we look at, so I was able to hire a number of people, including Kate Wilkin, Ali Tohidi in engineering, Kate Wilkins in biology, and a couple others, and Adam Kochanski. And that team, I mean, that was probably, people told me that the team that I hired is probably one of the strongest wildfire teams in the world. And Miguel Valero, who was here for the cluster hire, he's back now in Barcelona, but he's still part of the center. I'm Anastasiewicz, she's part of this center because the center became an NSF center. So it makes this national, it makes us, yeah, it's a San Jose State Center. But we're an NSF center, so the NSF wants it to be broad. We brought on- Which is great. Yeah, I mean the core is here. The heart is here, obviously. But, so those faculty, we wrote a lot of proposals together and my role now is, I'm still chasing fires, I'm trying to think about what radar I wanna get next and these kinds of things and what projects I wanna do, but I feel my role is really supporting junior faculty, not just in wildfire, but across campus. My startup was nothing because that was before research was really active on campus. And so I kind of had to start from scratch and build something from the ground up. And all faculty here have to do that. I mean, as you know, the CSU is a challenging place to build big research programs. But it happens. And we have amazing faculty across the CSUs, but particularly at San Jose State. And I think it's great. And I share their frustrations and what they've gone through and stuff. There is, you can make it here, and I think I've been lucky because of the support from the university. And honestly, it's the timing of wildfire science. Right. Since we started work, most universities around the West have started centers, are trying to build centers. And that's fine, you know how I am about all that, and then there's lots of competition, that's great, because we can't do everything. Other centers are doing other things, other groups are doing other things in the wildfire problem, that that's just not what we're doing. We're doing a lot of many things, but I think one of our big things is model development, forecasting the fire environment with real tools that agencies use, which is amazing, and building those tools and making them publicly available. So anyways, there's a lot things going on and the center, I'm just lucky that folks wanted to trust me and come here and. Move to the Bay Area and build a research up. So, and it's like Ali Tweedie's works amazing. Adam is just, he's got an army of researchers just cranking out new models that are, you know, the state of the science and that's exciting.

Vincent del Casino: No, it's fabulous and I mean, and I totally agree. I mean I grew up in the system. My first job was at Long Beach State and you know, I won't even tell you what my startup package was, you know. But you know I ended up in a research center doing work on the HIV crisis, I'm a medical and health geographer by training and it turned out like it became an amazing place for me to think creatively because I didn't have the same pressures of like here's your R1 and here's the five journals I want you to be in. So you can go out and you can play around, I think, a little bit more. So you may not have had much resources on the front end, but the flexibility to kind of be creative, I think is sometimes.

Craig Clements: That's what I say. I think like, you know, you come to an institution that may not, you don't have the pressure. And I was actually lucky. My career was based the fact that Allison Bridger, a former chair, she had just become chair when I was her first hire. And she kind of changed the dynamic of the department. Even though the department, meteorology was always a research heavy department because of its size and stuff. And traditionally that's what they kind of justified.

Vincent del Casino: Sure.

Craig Clements: Which is great, but that really helped. And I was told, do as much research as possible because students, they wanna get hands-on.

Vincent del Casino: Well, that's the other piece of it. I think that's important, too, and I'm interested in that. I mean, you obviously trained a lot of graduate students who've gone on and probably done all kinds of many stuff, either in industry or on for doctoral education. But also, the undergrads are getting involved. I mean there isn't a picture I see of you in the mobile environment where there's not three or four students at the same time. How important are students to the kind of work, and how do they inform how you even think about it, or sometimes even the questions you're asking.

Craig Clements: Of course, they're super important. I mean, obviously, we're an academic institution. I'm a professor, and that's who I interact with, or recruiting students. I used to have to say, hey, I've got money. Is fire weather a research topic you're willing to maybe dabble in, because there's a funded research project here, and students are like, yeah, yeah. Then they get really interested. Through the fire, we started the fire weather lab, fire weather research laboratory, People started coming and getting out and getting into the industry. Now, there's a huge need for fire meteorologists. San Jose State is the number one university that pumps fire meteorologist into industry. The utilities, all utilities are dealing with what happened in Hawaii, what happened PG&E, what happened California. You know, it was California and then now it's happening in Colorado, it's gonna happen elsewhere. It's happened in Oregon, so. That training that we offer here, which is quite unique, these students are becoming fire meteorologists and they're getting jobs, right away. They're getting poached. They can't even finish everything on time because they're a job offer. And I'm like, well, this is great. This is our goal, is to get you, to move you up and get you into a career path. So that's exciting. Now, working on the research, some students like to do the modeling, and some students like to. Being outside and everybody likes to launch a weather balloon and work in the field. I've had students travel to New Zealand, all around the U.S. Setting up equipment and they like it and most of it's part of their thesis but sometimes the way I work my lab is like okay this is your thesis, you're funded on this project and this student over here is on another project, funded on another and this one's on another but they all go to the field, they also up the equipment. Even if it's for one other student's thesis, they all interact and get the hands-on experience. So, because it's a team and you have to work as a team and so they all, it's great. And now with Adam in our department, we have more students coming in that are doing more of the modeling and technical computing aspects. But they all work together and they all studying the same phenomena. So, it is great to get students out in the field. Undergraduate students, so one thing you might know is every year. All the students and myself and any faculty that are interested in going out, we get FireLine trained. So we pay for it out of pocket to a private firm based in the Bay Area that goes through our training. It's a one day training. All new students have to do a 40 hour online course. It's basically wildland firefighter training. So your basic firefighter, training. That's awesome. So they go out and they get that. Now. When we deploy to wildfires, we're super conservative because we are not firefighters. Although I've had a bunch of students come back, come to San Jose State after a career as a wildland firefighter. And now because they're like, ah, I've been, you know, carrying shovel and lugging hoses for a long time. I wanna do the science and they move into a different career path after that. But having students that have a firefighter background, wildland fire fighter background is so important for me because we go on a fire. They know everything. You know, I'm still, I can't program the radios and stuff and it's like, okay, what are we doing here? How do we clone this? But our reputation on fires is really great, not because, because of our students are always so professional, but because we just continuously go to fires and we have a really good relationship with Cal Fire and an amazing agency to work with. So yeah, the students obviously play a big role, but they're also getting to take a radar out on a fire.

Vincent del Casino: So I have one last question. I think I could probably talk to you for another couple hours on this stuff and actually dig in on it. But one of the things that is interesting and obviously a big impact is computing power, the needs that we have now. And I made an argument in a few places, sometimes on deaf ears and other times not, that the next great divide in equity in higher education and training is gonna be in high performance computing because if you don't have access to that. You can't do anything that Adam's doing. And even the data sets you're bringing in, I can only imagine how significantly large they're getting now. So how important is that computing infrastructure, the future of computing, to our ability to do the work you're doing, just the research work, but even train students to give them environments, the simulation environments? We're talking really, really big models.

Craig Clements: Yeah, I know. I'm a hardware geek, right? I need trucks and space and radars and tools like that and wrenches, right. But the computing facilities, a lot of faculty need that. And as we know, cloud computing can work for a lot cases. A lot of research can be done on the cloud. And it tends to be cheaper, because you don't need to run data back and forth. So you can do it all there, keep it there, then you just download it when you need it. That works for probably a large proportion of computer scientists at most universities. They can get away with that. But we're doing real-time forecasts. So having a high-performance HPC cluster here, it gets used continuously. And we might get a call from CALFIRE. Adam gets a call. From CALFIR, it's like we're spinning up a run on a wildfire, and it takes a thousand core, a thousand processors. And so he definitely needs hardware here But he also needs cloud, but the hardware is important. For students, and I'm not in this field, but for students that are learning, needing to do HPC, because there's cloud, but then the HPC is the way you run your codes. It's specific to that system. And having that experience is very important, not just for our students, but all the students that are working in a high-performance computing environment. So that's engineering, math, and computer engineering, computer science, physics. So they're all using these computing tools. Unique at work and for wildfire is we're forecasting. We're actually running forecasts, real time, here at San Jose State and publishing those forecasts for agencies, which is unique. And it's tough to do because there was a one time, like the Dixie fire, it's the largest fire in California history, a million acres, and we were running that for almost a month and a half twice daily, and the grid got bigger and bigger because the fire got bigger and bigger and we're running at very high resolution. So it's. It's terabytes of data and just the computational need to run that one case operationally, meaning it's gonna spit out in six hours and it's going to spit out another six hours, et cetera. That's huge. And then you're running three or four fires here at San Jose State. So that's really cool. And I mean, Adam has this. And the unique thing is that Adam works really well. Like the team there, there's a wildfire ticket now for IT. So the support is there and that's really helped the group get the computing resources they need, even though we're buying the computing, but having it managed and being able to do that. And what's really exciting is NVIDIA, we've got now four more orders of four more of these NVIDia GPU clusters.

Vincent del Casino: I was just there yesterday. Okay, yeah.

Craig Clements: Adam got two this last year and apparently some of the first two anywhere. So it's pretty exciting. So they're trying to move their code to the GPU because some of their codes, they're running a fuel moisture model for the Conus for the nation that runs five times faster on the GPU. So it exciting.

Vincent del Casino: It's amazing. It's just so important now like you're not doing this work without access to those sorts of

Craig Clements: Yeah, no, it's important for universities to be able to invest in these types of facilities, whether it's a grant or internal or what have you, but having the ability to have PIs have those systems in place is important.

Vincent del Casino: Well, thank you so much, I really appreciate the time, this has been fantastic.

Craig Clements: Yep, thanks for having me. Great talking with you.