Eyes on Earth Episode 122 – EROS Interns, 2024

TOM ADAMSON:
Hello everyone and welcome to another episode of Eyes on Earth, a podcast produced at the USGS EROS Center. Our podcast focuses on our ever-changing planet and on the people here at EROS and across the globe who use remote sensing to monitor and study the health of Earth. My name is Tom Adamson.

We had a handful of interns working at EROS this summer. Some undergraduate college students, some recent grads, and some were graduate students. They got a variety of valuable experience, and we talked with some of them to find out what their experience was like, what they learned, and how they used their unique skillsets to help with the mission at EROS.

We talked to two groups separately about their work. First, we talked to Carson Price, Hunter Hagedorn, and Ryker Peede. So let’s start with Carson. Where do you go to college?

CARSON PRICE:
I go to school at South Dakota School of Mines. I just graduated there this last spring, and now I’m gonna continue my Masters degree out there.

ADAMSON:
Hey reminded us, where is the South Dakota School of Mines?

PRICE:
That’s out in Rapid City, South Dakota. So out in the Hills over there.

ADAMSON:
All right, yeah, over in the Black Hills. And Hunter, where do you go to college?

HUNTER HAGEDORN:
I go to school at South Dakota State University, which is in Brookings, and so I’ll be a rising sophomore this year.

ADAMSON:
And Ryker, where do you go to college?

RYKER PEEDE:
I just completed my Masters program up at SDSU this summer, so I will be done with college after this.

ADAMSON:
Now it’s time to talk about the work that you all did while you were at EROS this summer. Let’s start with Carson.

PRICE:
So I worked on a project called cloud mask closely with Hunter where we were using AI and machine learning to mask out different parts of satellite imagery from both MSS and Landsat images to help just mask out parts of the image that would be troublesome for different algorithms.

ADAMSON:
OK. Yeah, clouds kind of get in the way of us studying Landsat imagery. So you’re working on, and you mentioned AI, so you’re working on some artificial intelligence that will identify those clouds. Is that how that works?

PRICE:
Yeah, convolutional neural network. Just an adaptation on a normal neural network that can then grab different parts of that image and classify them as either clouds or water or many other things.

ADAMSON:
OK. So you mentioned that Hunter is also working with you on this project. Hunter, do you have anything to add or is there any other perspectives that you have on that work?

HAGEDORN:
Yeah, so as Carson said, I’ve been working with him on the cloud mask project. Throughout the process, we’ve been doing a lot of work, especially migrating to Landsat 9 data, which was a big headache that they had since the MSS data was a lot more difficult to use. And we’ve also created more of a CI/CD pipeline to use with our project, which is another big thing that I’ve been working on this summer.

ADAMSON:
Describe that pipeline again. I think there was an acronym in there. What does that mean?

HAGEDORN:
So CI/CD is, I think it stands for continuous integration, continuous development. It’s basically just making sure that the work we’re doing actually works when we’re trying to use it and it kind of makes an ease of the process.

ADAMSON:
Both Carson and Hunter mentioned MSS data, which is the sensor that was on Landsats 1 through 3, and it sounds like you’re making sure that works, kind of in coordination with current Landsat data. Does that sound about right?

HAGEDORN:
Yeah.

ADAMSON:
Either one of you can answer this. Do you have ideas on how people will benefit from that work?

PRICE:
Any algorithm that uses satellite imagery could benefit from this, so anything that can identify different things like land fires or city development. Then all of them can take advantage of masking out different parts of the image that are troublesome.

ADAMSON:
Let’s turn to Ryker. What did you work on here this summer?

PEEDE:
I was actually a part of a couple of different teams, one of them being the Dev OPS team. I was in charge, or tasked with, monitoring the costs of all of the accounts of different projects that we have that are hosted in AWS up in the cloud. So every job that is done in the cloud has a cost, and so I was kind of tasked with monitoring that as well as creating presentations for, I guess stakeholders. So some government customers as well as account owners, just to see how they’re spending the funds in their account.

The other team I was a part of was the data management team. There I learned SQL and some new skills over there, but my main task with them was to help automate existing reports so that would help cut down the time it took to do these reports either every week or every month or whenever they had to be performed.

ADAMSON:
Let’s turn back to Carson. What was something surprising that you learned during your internship?

PRICE:
I thought it was really surprising to learn a lot more about the cloud because it’s so large. There’s a lot of different things that you can do with it. So diving into all of the different ways that you can do one thing is really daunting at first, but after working with couple different people then it makes it a little easier to work through.

ADAMSON:
How about you, Hunter? Is there something surprising that you learned during your internship?

HAGEDORN:
Yeah. So I was really surprised with all the work that actually is going on out here at EROS. I’ve always heard about EROS and I’ve seen the signs that when you drive on the Interstate and you can see the signs for EROS, but I never really knew what work was going on here and how big of an impact it really has on the world and what I guess the mission is here.

ADAMSON:
And Ryker. What’s something surprising that you learned?

PEEDE:
I would say similar to what Hunter said, but just the mission that we have out here at EROS and how it’s integrated with everyone who uses, I guess just the data that EROS produces and the products we produce from science to, you know, universities, just everyone who consumes it. It’s a product that everyone throughout the whole world uses, and to kind of take a step back and see that that’s our mission out here and it’s to provide that, and it’s the longest running, I think, repository of public accessible free data since 1972 is just something I think that is cool and to be proud of.

ADAMSON:
What was the biggest benefit for you? The big take away from this internship?

PRICE:
Yeah. So I had always had an interest in AI and machine learning, but this really cemented my idea of what that means in an enterprise situation. I’m going to be continuing my Masters project as the project that we’ve worked on all summer. So then, not as a thesis, but as a Masters project I’ll get to continue working on this.

HAGEDORN:
I think it’s really nice that I got this internship as early as I did because it kind of helps me point myself in more of a direction with the rest of my undergraduate career. Same with Carson. I’ve always been interested with AI, but I didn’t really know what it meant in the workplace, and so this upcoming semester I’ll be doing an exchange program in Minneapolis at the University of Minnesota and I’m seeking out a lot more AI classes since working with it here.

PEEDE:
Yeah. So since I’m kind of, I was towards the end of my academic career when I was hiring for this internship, they kind of took my interests and my background and kind of aligned me with what they thought I would be interested in. Like the cost reports and monitoring the cost with the cloud accounts as well as the data management, which is really interesting because my background is math and data science, so just to be able to see large databases and how they work and then interact with them on a day-to-day and you know write queries that help people get the information that they need and to be able to tell a story. I think that was best aligned with what my interests and my background was, and I think they did a good job at positioning me because I’m very happy with where I am and the teams I’m on.

ADAMSON:
Now we’re going to talk to three more interns who were part of the Wildland Fuels Research project. Grace Parrott is a recent graduate from Luther College. Hazel Mebius is a student at Dakota State University, and Katelyn Woolfrey is a student at the University of Virginia. Let’s start with Grace. Tell us more about Luther College. Where is that?

GRACE PARROTT:
Luther College is in Decorah, Iowa, in a—it’s a small northeastern Iowa town in the Driftless Region, and it’s pretty small. I mean there’s about 1500 students that go there and it’s a very tight knit environment. So I really like going there.

ADAMSON:
That’s great. And what was your degree in?

PARROTT:
My degree was in environmental science and I got a minor in biology.

ADAMSON:
Hazel, where’s Dakota State University?

HAZEL MEBIUS:
Dakota State is in Madison, South Dakota, about an hour north of Sioux Falls. And I went to Dakota State to study cyber leadership and intelligence with an emphasis on digital forensics.

ADAMSON:
Katelyn, where’s the University of Virginia?

KATELYN WOOLFREY:
It’s in Charlottesville, Virginia, which is about 19 hours southeast of here.

ADAMSON:
Only 19 hours. Excellent. And what is your degree program?

WOOLFREY:
I’m getting a Bachelor of Science in Biology with a minor in statistics.

ADAMSON:
OK, let’s go back to Grace. What did you work on here this summer?

PARROTT:
So this summer I was part of the Wildland Fuels Research project, which involves doing a lot of field work in the Black Hills and Sequoia National Park this summer on mixed severity fires, setting kind of fuels, the fire severity, fire behavior and how the ecosystem is recovering after those fires.

ADAMSON:
Hazel, do you have anything to add to what the project was?

MEBIUS:
Yes, we had a focus on terrestrial lidar scanning when we were there and we, yes, we scanned 20-meter circle plots, and we wanted to see how the TLS scan compared to regular tree metrics. So we did both traditional forestry measurements and then we also did scanning and then we’re going to see if scanning is somewhat equal to traditional forest measurements.

ADAMSON:
What’s a traditional forest measurement?

WOOLFREY:
There’s a lot of them. The easiest might be just tree height, you know height from top to bottom, but also things like diameter at breast height and then stuff like canopy base height, which is how tall the canopy is from the ground and canopy radius. Other things like that.

ADAMSON:
OK, so that’s the traditional field work, but you’re also using lidar as part of what you were working on too?

PARROTT:
And we also focus, we focus on the trees, but we also focus on the fuel bed. So we measured the live vegetation, but also the litter at each site along transect lines to see how the fire would burn and did burn.

WOOLFREY:
That would be stuff like live grass, dead grass, logs and sticks and stuff. Anything on the ground that’s not a tree.

ADAMSON:
That’s fuel, potentially for a fire.

WOOLFREY:
For a surface fire, specifically.

PARROTT:
And we also looked at soil characteristics as well with that.

ADAMSON:
Dry soil versus moist soil that —

PARROTT:
We were looking at where — so duff specifically. So where the layer between where there’s litter on top of the ground and organic soil. So kind of like partially decomposed litter because if there’s a lot of duff present, then it shows that the fire wasn’t as severe because really severe fires really incinerate that duff layer, which is actually really beneficial for the soil.

ADAMSON:
How will others benefit from this work?

PARROTT:
Scientists will for sure benefit from this work because we are using the terrestrial lidar scanner as well as doing traditional forestry measurements and comparing them and seeing how accurate the terrestrial lidar scanner is to doing these forestry measurements, we are able to possibly quicken this process if it is as accurate because doing the manual measurements takes sometimes up to two or three four hours, depending on how many trees are in the plot, but the scanner can scan the plot in 20 minutes, so it would really quicken data collection and make it so scientists can collect more data on more areas and get a better representation of wildfires. So it could really quicken the field work process, and the quality of the science.

WOOLFREY:
I would also add that it also just removes a layer of error with the data because we took field measurements the entire summer and now we’ve come back and have to enter all of that data and we’re looking at these little notebooks that have mud and dirt and ash all over them. You know, it’s all of this stuff that comes with handwriting your data every single day, and removing that with lidar is so much easier.

MEBIUS:
Like you go from, I don’t think this tree was 72 meters away or like 7.2, but ohh, we’ll just add a comma in here and hope that fixes the problem.

WOOLFREY:
Yeah, because you can’t go back and measure it again. You know, it’s five hours out there. Yeah, doing it with a scanner is just removing a little bit of that human aspect that might add a little bit more error to it.

ADAMSON:
Did you all go to the Black Hills for this field study?

PARROTT:
Yeah. Actually, Hazel and Katy were out there earlier than I was, and they did a whole trip before I even came. But yeah, June was focused on the Black Hills, and we were primarily there looking at the Jasper wildfire that happened in 2000 in the Black Hills National Forest.

WOOLFREY:
And then July was for Sequoia National Park, and we were looking at the KNP Fire that happened there in 2021.

ADAMSON:
You all went to Sequoia too. OK, you all got to see the big trees?

VARIOUS VOICES:
Yes.

MEBIUS:
The big trees are very cool.

PARROTT:
Yes, every single day. It was like I had seen them for the first time and my draw just dropped. And you try to look at it and then you almost fall over because your head is tipping back so much and you’re just—It’s just amazing.

MEBIUS:
We had to walk up really steep slopes. And then I’m, like, trying to take a picture to show back home. Look what I just climbed up, this 20° slope, now and I’m looking at the picture. That’s just doesn’t even show what it was, when you’re having to actually climb it up.

PARROTT:
I mean, it would be like 600 feet of elevation gain in 0.2 miles. And so it’s just—you’re crawling on your hands and knees.

ADAMSON:
Let’s go back to Grace. What was something surprising you learned during your internship?

PARROTT:
I didn’t know much about wildfire. I mean, I haven’t studied it that much. I was on the burn crew in college and that was really cool to administer prescribed burns on the natural areas that Luther has. It was very interesting to me to see the different severities of wildfire and the effects. So when we were out in Jasper, the fire happened 24 years ago, back in 2000, and some of these areas just had downed logs covering hillsides and no regrowth at all, and it had been 24 years. And that just shows how negative an effect that the wildfire had in this area, that there wasn’t even a little bit of regeneration. So it was just very striking to me to see that.

ADAMSON:
Hazel, what’s something—

MEBIUS:
I guess for me was mostly learning about the after effect of fires. You always hear, OK, there’s a wildfire. Houses might get burned down or something. And I mean, that’s definitely bad. But just learning the effects afterwards. I learned all about erosion with the whole BAER team and landslides after a fire, which I don’t even really thought about till now. So that was very surprising.

ADAMSON:
Right. You mentioned BAER team. That’s actually B-A-E-R.

MEBIUS:
Yes, Burned Area Emergency Response.

ADAMSON:
Those are the teams that are studying those things like landslides, erosion. OK.

MEBIUS:
Yeah, yeah.

ADAMSON:
Katelyn, what is something surprising you learned?

WOOLFREY:
I think I just learned so much about South Dakota that I didn’t know. Just little cultural things. I’ve been to South Dakota before, but I get here and at lunch every day I learn something new. Like you guys have scotcharoos that taste really, really good. I got to try a scotcharoo the other day. I heard about Puppy Chow.

PARROTT:
We still have yet to make Puppy Chow for Katy, and that will happen next week.

MEBIUS:
Rhubarb crisp.

PARROTT:
Rhubarb crisp—yes, that’s iconic!

WOOLFREY:
Little stuff like that I learned every day, or little pronunciation stuff, and that’s not why I came here. But it’s very interesting.

ADAMSON:
How has your perception of remote sensing changed since you started your internship?

PARROTT:
I am not familiar with remote sensing. When I think of remote sensing, at least before this summer, I just thought of satellites and Landsat. But we started using this lidar system and I realized, oh, there’s terrestrial forms. And there’s also, you know, drone footage that we can take and satellites. So there’s many different ways of doing remote sensing, which is cool because then you can apply to a lot of different things.

ADAMSON:
Hazel, your turn.

MEBIUS:
Yeah, well kind of bouncing off of Grace is the applications for remote sensing. You think satellite imagery and you’re like, OK, that’s cool. What can you do with that, until you realize you learned all the different science teams we even have here, like famine watching, wildfire, what we were working on, coastal monitoring, and droughts and whatnot, so that was pretty cool.

ADAMSON:
OK, Katelyn, your turn.

WOOLFREY:
I think the most interesting thing, and this is kind of leading into the more final parts of our project is combining remote sensing sources together because we’ve collected all this lidar data now and our next step is to combine that with other sources of lidar like GEDI, which is currently on the ISS, or airborne lidar and also using that in tandem with Landsat and all this other stuff. So it’s not just looking at one source and seeing what that can tell you but also using it in combination with other stuff.

ADAMSON:
We also asked our interns if they had any advice for future interns at EROS. Here’s what they said.

PRICE:
I would definitely give the advice that you should always be open to asking questions. Because everyone will at least try to answer it, or they’ll direct you to whoever knows a better answer, so it doesn’t hurt to ask anything.

WOOLFREY:
My advice would be not to pigeonhole yourself because I never thought of myself as going into wildfire research. When I started my biology major, I thought evolutionary biology or computational biology. Those were kind of things that stood out to me, and then I saw this random posting and I was like, that would be interesting. And I applied. Didn’t think I would get it. I did. I met this great group of people. I’m in South Dakota now. I didn’t think I would travel halfway across the nation. And I really like it. I think it’s really interesting, and it’s something I would have never, ever looked at otherwise, because it just didn’t occur to me.

HAGEDORN:
Exactly what Carson said. I mean, don’t be afraid to ask questions. That’s kind of for any situation, but I think also being open to learning and also open to advice and don’t take criticism too hard because as an intern, you’re not really—you don’t really have all the answers, and there’s people here who have worked here for like 20 plus years. I would also say don’t keep your head down, and get involved. There’s a lot of things around here and it makes it a lot easier to just be working when you know people around here and actually feel like a part of the team.

PARROTT:
My piece of advice for future interns is don’t be intimidated by new things. I remember being out on my first day in the Black Hills trying to learn all the protocols, and we had practiced them before we had gone out, but I’d only gotten, you know, a first glance and so actually applying them in the field and using them to collect actual data that for this project was a bit nerve wracking, but after a couple of days I got it down and so that was very encouraging. So I just say give it time, and right now we’re getting to a lot of different softwares to try to process and analyze the data and that’s also a bit intimidating because some of them seem very archaic and haven’t been updated for years and years and years, and are just a bit finicky to use, but to just keep at things and try to not get discouraged would be my advice.

MEBIUS:
My advice is talk to people, which people hear a lot, but everybody here is so passionate about what they do. We were looking for maps to decorate our cubicle for, and so then we end up getting to talk with Roger and then that goes like 2 hours by about learning about different topographic maps and whatnot—he’s so interesting. But there was like, Oh well, thank you. And so just everyone’s story is interesting and knowledgeable. Just don’t be afraid to talk to people. They’re usually more than happy to talk to you about what they’re working on and why they find it interesting.

PEEDE:
You basically get out what you put into an internship, meaning, if you come in every day kind of keep your head down and only stick to the work that you’re tasked with, you’re not going to learn a whole lot. And so what Carson and Hunter said as well, I would, you know, encourage interns to talk to as many people as you can, as many departments as you can. Learn about the mission. See how everyone works together. I know they try to do this every summer when you have interns, but try and get on a tour of EROS because it’s kind of hard to understand just the scope of what’s going on out here, and I think that can best be understood with an hour tour, or however long it is that that they do provide.

ADAMSON:
Thank you to all the interns for joining us on this episode of Eyes on Earth. It was great to get to know you and hear you talk about your time here at EROS. Check out our social media accounts to watch for all future episodes. You can also subscribe to us on Apple and YouTube podcasts.

VARIOUS VOICES:
This podcast, this podcast, this podcast, this podcast, this podcast is a product of the U.S. Geological Survey, Department of Interior.