#14 – Mike Zoeller and Adam Mosbrucker

Mike Zoeller and Adam Mosbrucker play key roles within the United States Geological Survey (USGS) Volcano Hazards Program. Both Mike and Adam rely heavily on the use of lidar and its derivative products to support applied science projects related to the specific volcanology of their respective regions, the Hawaiian Volcano Observatory (HVO) and the Cascades Volcano Observatory (CVO), respectively. In this episode, we chat about some of the exciting and novel applications of lidar within their work such as locating landing zones for helicopters and finding ideal placement locations for telemetry and solar power systems for GNSS instrumentation. We also discuss how lidar is utilized to constrain lava flow, develop geophysical models of underlying volcanic mechanics and get a handle on long-term sediment storage and basin yields.

Episode Transcript:

#14 – Mike Zoeller and Adam Mosbrucker

October 28th, 2024

Editor’s note: Any use of trade, product, or firm names by U.S. Geological Survey employees is for descriptive purposes only and does not imply endorsement by the U.S. Government.

Announcer: Welcome to the LIDAR Magazine Podcast, bringing measurement, positioning and imaging technologies to light. This event was made possible thanks to the generous support of rapidlasso, producer of the LAStools software suite.

Austin Madson: Hello everyone and welcome to the LIDAR magazine podcast series. My name is Austin Madson, and I’m an associate editor at LIDAR magazine. Today we’re kind of continuing down the road of exploring these many different and exciting applications of lidar remote sensing. And you may remember that I spoke with a couple of guests on a recent episode from the Geological and Geophysical Survey in Alaska. If you haven’t given that a listen I would recommend to do that.

That said, today we’re super happy to have the opportunity to chat with two guests from the United States Geological Survey Volcano Hazards Program, Mike Zoeller from the Hawaiian Volcano Observatory, or HVO and Adam Mosbrucker from the Cascades Volcano Observatory, or CVO. The HVO and CVO are two of the five US volcano observatories run by the USGS Volcano Hazards Program. And just as an FYI, the other three are the California Volcano Observatory, the Yellowstone Volcano Observatory and the Alaska Volcano Observatory.

Mike Zoeller is a geologist and GIS analyst at the USGS Hawaiian Volcano Observatory. His time at HVO has encompassed Kīlauea volcano’s 2018 flank eruption and associated summit collapse, a series of post-collapsed summit eruptions starting in 2020, and the 2022 eruption of neighboring Mauna Loa volcano. Mike’s work has historically been focused in digital cartography and data dissemination. But topographic modeling including the use of lidar has grown to play a more important role in HVO’s research and monitoring efforts. Recently HVO procured an airborne lidar system that has completed test flights and is expected to begin data collection in earnest starting this fall.

Adam Mosbrucker is a geologist and geomatics expert at the USGS Cascades Volcano Observatory or CVO. He also serves as the volcano hazards program lidar coordinator. Adam specializes in quantitative fluvial geomorphology, with most of his career focused on sediment transport and Mount St. Helen’s volcano after the 1980 eruption. Topographic modeling has been a passion of Adam’s since college with lidar was an acronym with capital letters that needed explanation. Working with this geomorphic eye candy for more than 15 years has rewarded Adam with great appreciation for its lasting value alongside structure for most photogrammetry and other geospatial tools. So I just want to extend a welcome and thanks to Adam and Mike for joining us today. We appreciate it.

Michael H. Zoeller: Thanks for having us Austin.

Adam R. Mosbrucker: Yeah, this is great. Thank you Austin.

Austin Madson: Of course. So Mike and Adam, to get our listeners better acquainted with the USGS Volcano Hazards Program and in particular the US volcano observatories of which you all are both a part of. Can one or both of you speak to the specific goals or tasks of the HVO and/or the CVO?

Michael H. Zoeller: So Austin, to start by answering that question I’m going to just read off here the official mission statement of the USGS Volcano Hazards Program, which is that our mission is to enhance public safety and minimize social and economic disruption from volcanic unrest and eruption. We deliver forecasts, warnings and information about volcanic hazards based on scientific understanding of the volcano’s behavior. There’s a lot that goes into that, and I think you can kind of divide it along two lines that kind of merge together in some points. And those two lines are operational monitoring and research focused.

So a lot of our work, especially when the volcanoes are not erupting is studying their long-term behavior and building better understanding of eruptive patterns and things like that so that we can better inform hazard assessments and maybe even eruption predictions when short-term unrest happens. And when that does become the case and eruptions occur, then we’re of course tasked with monitoring those eruptions, passing information to emergency responders and local government officials to make decisions about evacuations and things like that in order to minimize the impacts of these eruptions.

Austin Madson: Adam, would you say that your volcano observatory, CVO shares similar goals and tasks? Are there any kind of inherent differences between HVO and CVO?

Adam R. Mosbrucker: Yeah, I think our primary goal is the same. We’re sister observatories. There’s of course differences in the volcanoes themselves that we monitor. But our primary mission is the same.

Michael H. Zoeller: Eruptive styles are very different between the two types of volcanoes. Like, we’re dealing with mostly effusive eruptions here in Hawaii. But at Cascades there and also at the Alaska and California observatories it’s primarily explosive hazards. So yeah, the impacts are often the same, but the actual hazards presenting themselves from the volcanoes can be a little bit different.

Austin Madson: Right, yeah, and that makes sense. The regions are really quite different, which alludes to five different volcano observatories.

Adam R. Mosbrucker: Mike is fortunate to get to see a lot more eruptions than we do.

Austin Madson: For better or for worse.

Adam R. Mosbrucker: Well, from a scientific perspective I suppose {laughs}.

Austin Madson: Right. So Adam, I know that the CVO uses lidar for lots of different applies projects to help kind of facilitate the really great science that you all do over there. Can you talk a bit about how you and your team utilize lidar in different kind of derivative products to help find helicopter landing zones and why, for example, that’s important?

Adam R. Mosbrucker: Yeah, we use helicopters quite a bit at CVO, AVO and HVO. It really increases the efficacy of our field work. So it’s usually to access volcano monitoring stations. But it’s also to transport research staff to remote areas. One of the ways that lidar can be used for that that I’ve used it for that and others is helicopters need a specific slope angle. They can’t land on steep slopes, for instance, and they can’t land in heavily forested areas. And so if you look at just your typical slope derivative and then combine that with say DSM, so you can see the vegetation, the canopy height. Put those together in your study area and determine where a helicopter might be able to land.

Austin Madson: Right, so you’re kind of doing this analysis. You have a general idea of where you want to get in, and then you kind of pinpoint locations that are actually feasible.

Adam R. Mosbrucker: That’s right, yeah. And that slope analysis can also be useful for field crews, especially those traversing terrain with kind of valuable information, see where they might be able to get out of a deeply incised canyon, for instance. Especially at Mount St. Helens and other Cascade locations. You can – it’s easy to hike miles trying to get up and out of a canyon, when if you would have known that you went a few hundred meters downstream you could have gone up this little side channel.

Austin Madson: Right, especially if you have 40 pounds of gear on your back too.

Adam R. Mosbrucker: Exactly.

Austin Madson: What kind of – as a follow on to that, I know that the volcano observatories, like CVO of which you’re a part of, rely really heavily on high precision GNSS instrumentation to kind of monitor volcanoes and underlying mechanics, and in particular the complex systems that are going on underneath. So that said, real time monitoring and dissemination of this data does require various power and telemetry systems. Can you talk a bit about how you use lidar data in those kind of derivative products to not only locate these systems, but which of these areas are going to provide the most amount of power or the amount of power that you need? And also kind of data downlink and uplink line of sight analyses and things?

Adam R. Mosbrucker: Yeah, it really becomes like a site suitability analysis. Some of those basic skills that many of us learned in our first or second GS class. Similar to slope. They’re sort of easy derivatives from lidar. But lidar gives us the accuracy and the ability to do that easily. So site suitability, we’re looking at both solar power potential. So typical solar radiation calculations in, for instance, ArcGIS, which the government has a site license too. So all have access to that.

And then combining those power needs, say hots per square meter with radio frequency analysis. So different – most of our telemetry systems rely on radio links, and different frequencies are more line of sight than others. So having a clean line of sight is sort of the way to optimize those systems. If you use viewshed analyses and especially if you overlay those nested on top of each other because some of our telemetry paths require repeaters. So you have maybe two or three sights coming through a repeater, and then that repeater going out to a hub that’s hard wired into a network, for instance.

So multiple viewshed analyses of an area in line with that solar power potential can easily identify some of those sites. Of course no matter how good a GIS analysis is you have to go out to the site to see, right, before you haul 500 pounds of gear there in a helicopter expecting to set up a station.

Austin Madson: So you’re going to kind of run the suitability analysis and then go and get boots on the ground and get up there to make sure your analysis was correct.

Adam R. Mosbrucker: That’s right, yeah.

Austin Madson: Have you automated that process, where you’re just kind of going through and clicking through the different tools?

Adam R. Mosbrucker: Yeah, it’s really iterative. We don’t – we have a lot of volcano monitoring stations. But the rate at which we put up new ones is pretty low. We tend to keep them for years or decades even. It’s not frequent enough I guess to need automation. And yeah, it’s just clicking through the tools and then working with our field engineer and instrument teams kind of iteratively.

Austin Madson: Well, how many – you said that you don’t put too many out there or do any replacements I guess. But how many do you have approximately? Do you know off the top of your head how many stations?

Adam R. Mosbrucker: Oh, gosh.

Austin Madson: Is it, like, 500 or in the…

Adam R. Mosbrucker: No, I mean, across the Cascades…

Austin Madson: ‘ish.

Adam R. Mosbrucker: ‘ish, maybe 100. Maybe 100.

Austin Madson: Okay well, yeah, thanks. It’s a really interesting applied use case of lidar and derivatives that is really important but often overlooked I think, some of these kind of site analyses, questions and data questions and things that you ask. So thanks for filling us in there Adam. Mike, I know a lot of our listeners may recall that Kīlauea and Mauna Loa have had several notable eruptions over the last couple of years. Can you speak to how the HVO or Hawaiian Volcano Observatory uses lidar derived products to better constrain lava flow prediction models and why the accuracy of those models is important?

Michael H. Zoeller: Yeah, so I think at the most basic level in predicting lava flow directions, lava flows downhill. So can often use a lot of the same tools that people use for, like, watershed analyses to project flow directions. But because we’re talking about a three dimensional structure here and the viscosity is obviously a lot higher than water when it comes to lava, so yeah, morphologically it presents a little bit of a different structure coming down the hill.

The topography is really important, and also at the start of my career nine years ago we were doing what was called Favalli modeling for a lot of flow directions. Where we would just run one path of steepest descent down the slope, and then we would introduce slight perturbations into the digital elevation models to get a little bit of randomness to the potential directions. That was because we were using at that time, like, 1980s national elevation dataset 10 meter DEMs.

Now that lidar is available for most of the island of Hawaii, we’re able to use much more accurate lidar DEMs where we have greater confidence and don’t need to do all that iterative analysis of the different potential perturbations. And we have much more confidence. It also – and by eliminating that need for so many iterations, we’ve developed much more advanced tools that take into account the greater complexity of lava relative to something like water.

So we have a postdoc within the Volcano Science Center named Dave Hyman who has published this tool called Lava2d which takes into account all sorts of different morphological constraints of a lava flow coming down a hill, such as crystallinity, initial eruption temperature. And it’s able to predict run out distances of the flow. So it’s not like the old Favalli models, you started it at a point on the ground up on the top of the mountain, and it ran all the way to the ocean no matter what.

Now we can gauge the run out distances, and obviously the slopes that we’re getting out of the high accuracy lidar DEMs are critical for determining those run out distances. If the slope is higher it has potential to move faster and get farther away from the source then. If the slopes are lower, if things tend to shorten and widen, it is the prevalent pattern. Now I’ll say that most of the island has one meter resolution, 3DEP lidar DEMs available. We are still down sampling that to ten meters just because the processing time on the one meter dataset is pretty long. And we’re not sure if there really is a pay off. But even at ten meter resolution the greater accuracy relative to those older DEMs, which were derived from contour lines, has made a huge difference in our ability to accurately predict things.

And the 20202 eruption of Mauna Loa is a great example of that. I think Dave is working on a paper that shows that the predictions that he came up with using the Lava2d model and lidar derived topography lined up really nicely with what eventually happened during that eruption.

Austin Madson: Right, so the kind of constraining variables, it sounded like someone needed to go out and grab those in situ in order to do some more real-time modeling. Like, the temperature and what was that, some viscosity measurement sorts of thing?

Michael H. Zoeller: And crystallinity. So we can get that from the lava samples that we take out of the channel. Eruption rates, so the effusion rate of a volcanic eruption is essentially the amount of lava in volume that is coming out of the ground per unit time. And that’s a critically important variable when it comes to all eruptions for many reasons. But it’s especially important for input to that Lava2d model because the higher effusion rates can force fluid lava much farther away from the source than lower effusion. Lower effusion tends to be stacked up closer to the source vent.

So yeah, the inputs that we need are things that are derived from actually visiting the eruption on the ground. Those more basic Favalli models that I started my career with, you can still run those with just a DEM, whether it’s from lidar or the National Elevation Dataset in a point to start the flow at. But to get these more accurate models it’s nice to have the eruption parameters better constrained from ground visits and to have that lidar topography to better model slopes.

Austin Madson: Playing off of that, what about – so you have a flow come out in August and then another flow come out in October. Are you able to kind of update your topographic inputs based on that kind of change in micro topography?

Michael H. Zoeller: Yes, because it doesn’t take much to deflect a lava flow. A feature on the ground that’s, like, a handful of meters tall and several hundred meters long if it’s placed in the right position and oriented the right way can deflect the lava flow. And certainly the previous lava flow that came out of the volcano is a great example of something that can do exactly that.

So a good example of this was during the 2018 eruption at Kīlauea. It lasted three months, and there was enough time during that eruption that we were able to contract a couple mid eruption lidar surveys. And we were able to use those digital elevation models that came out of those surveys to run more accurate flow modeling based on the stuff that had already been in place on the ground.

The Mauna Loa eruption in 2022 was much shorter duration. We didn’t quite have the time to do that. But we still had our photogrammetry tools. So we were trying to use photogrammetry to the same effect of updating topography during the eruption. But this is where I come to HVO’s most exciting development, that we now have a RIEGL Vux-120 system of our own that’s gone through testing. We’ve flown it once and yeah, we’re going to start collecting some real data with it here in the fall. The first thing we want to do with it is go get updated topography for that whole 2022 Mauna Loa eruption. It’s going to be really useful during future eruptions because then we’ll be able to keep going back for more updated data as the eruptions progress.

Austin Madson: That’s great. And now let’s have a quick word from our sponsor LAStools.

The LIDAR Magazine Podcast is brought to you by rapidlasso. Our LAStools software suite offers the fastest and most memory efficient solution for batch-scripted multi-core lidar processing. Watch as we turn billions of lidar points into useful products at blazing speeds with impossibly low memory requirements. For seamless processing of the largest datasets, we also offer our BLAST extension. Visit rapidlasso.de for details.

Austin Madson: So Mike, I have a kind of a follow up. It seems like in the past your team was kind of reliant on contracting out different acquisitions to update those models, like, during some of those longer eruption periods or to just update the models in general. So how has this new in house lidar scanner or I guess how will it affect things at HVO moving forward or any of your partners or anybody that you may be working with on some of these projects?

Michael H. Zoeller: Yeah, the system that we have is fortunately pretty easily deployable. I know the VUX-120 is capable of flying on UAS platforms. I can’t say that we have enough confidence in our drones to stay in the sky with $150,000 instrument attached to it. So what we’ve arranged, we’ve built a mounting apparatus to hook it up to our preferred helicopter, which is an MD 500, and we have a vendor that we’ve worked with for decades here who knows the island like the back of his hand. And we trust him flying the thing.

So yeah, we’ve done our test flights on his aircraft, and we’ll be able to deploy this thing on short notice on the order of, like, hours to a couple of days to get out there and fly it. Even if we have a week long eruption we could potentially fly the thing multiple times during that eruption and get updated topography with each of those collects.

Austin Madson: Right. Are you working with any partners on the RIEGL hardware? Or is this just kind of an HVO in house only kind of thing?

Michael H. Zoeller: We do have some really important partners on this. In addition to RIEGL themselves who have been very helpful, we’ve been working a lot with the US Army Corps of Engineers, Cold Regions Research and Engineering Lab based out of Hanover, New Hampshire. They’ve been a long time partner with us on lidar projects. They actually loaned us, I think it was a VZ-2000 terrestrial scanner for a couple of years that we were using out at the crater during some of the Kīlauea summit eruptions of late.

They have extensive experience in lidar technology. They mostly use it for monitoring glaciers in Greenland and stuff like that. But they collaborated with us on work out here. They’ve visited a couple of times in addition to loaning us their instrument for a little while. They helped enormously on the design and integration of the whole system with the aircraft and the mounting apparatus. So yeah, they are our trusted comrades in this, and we have a lot to thank for them. Specifically Adam LeWinter there at the Cold Regions Lab, his supervisor Dave Finnegan and also (sounds like: Don Filiana) who was just out here helping us set up the instrument.

Austin Madson: Yes, exciting things are happening over there. It’s always fun to get nice, new toys.

Michael H. Zoeller: Um-hum, absolutely.

Austin Madson: Well, switching gears a bit over to Adam, Adam I know you’re kind of a geomorphologist at heart, and maybe that’s your bread and butter so to speak. But that said, can you talk a little bit about your work on morphometric sediment budgeting using lidar, and I think you use photogrammetry and maybe some cameras? Something that you’ve dubbed said cam and things. Can you talk a little bit about that ongoing work you’ve done for a long time?

Adam R. Mosbrucker: Yeah, I can talk about this topic for about four to five hours if you want. {Laughs} Okay, I’ll try to break it down. Volume is of great interest to myself and many of my colleagues, from calculating infusion rates like Mike discussed or volcanic dome growth, glacier mass balance. Even an entire edifice to understand an eruptive history.

But for my work, fluvial sediment transport primarily at Mount St. Helens, which I guess I should provide your listeners with a little bit of background on Mount St. Helens briefly. So the 1980 eruption of Mount St. Helens generated one of the largest landslides or debris avalanche that we have in recorded history. It buried the Upper Toutle River. That’s T-O-U-T-L-E. People get confused on how that’s pronounced, Toutle. It buried the Upper Toutle River Valley in three point three billion cubic yards of sediment. So like our national debt, three point three billion cubic yards, can either of you imagine what that would look like?

Austin Madson: It’s a lot.

Adam R. Mosbrucker: It’s a lot. It’s a big number. Tammy Christianson, one of my colleagues at CVO, she calculated this out for a fact sheet that we put out. So it’s enough to fill one million Olympic sized swimming pools. Maybe that’s easier to picture. So what Mount St. Helens really is, it’s a very unique natural laboratory for geomorphologists. We’ve reset an entire landscape and then watched – been able to watch over the past four decades river systems and sediment transport do its magic.

And we also, throughout that entire four decades, we have a huge stack of DEMs. A huge stack of elevation data from traditional photogrammetry to lidar. I think our earliest lidar there was 2003. Our most recent is 2022. But we have 1980 photogrammetry. We have a suite of SfM photogrammetry sets of different areas around there. So calculating long-term sediment storage and basin yield with DEMs of difference, one minus the other, some in time step, it’s a really great tool to understand transport processes there. As long as there gets to be some issues with horizontal, but more vertical alignment. You got to be really careful when you’re working with so many different datasets from different sources across different time, right. There’s – datas have changed, geoids have changed. Accuracy and uncertainty, assessments get kind of interesting, but it works great.

Austin Madson: Yeah, I was going to ask about your uncertainty and how you’re deriving some of those figures. Are you just propagating the uncertainty through each surface model that you have?

Adam R. Mosbrucker: There’s different ways to do it of course. I tend to just use a propagation route, some squares equation or so, looking at the individual accuracies of each set and then propagating those together. Another advantage we have at St. Helens—we’re completely spoiled there, I’m spoiled—is we have a huge network of stream channel cross section surveys. There’s – I think there’s over 300 surveys that were established right after the eruption. With the intention of understanding this transport issue kind of before geospatial tools really became efficient.

But what that means is, there’s seven or eight hundred monuments across the area, and not all of those have high quality say, like an RTK point on them. But a lot of them do. So we have a built in large network of ground check points that we can use to evaluate that uncertainty also.

Austin Madson: I see. What about – so I read one of your AGU, American Geophysical Union fall meeting abstracts. I think it was from last year. And you were talking about your SedCam. Do you to enlighten us here on what that is and how you’re using that?

Adam R. Mosbrucker: Yeah, SedCam. Not lidar related, but it’s super fun. Besides DEMs and the cross section that work, we also have – there’s a long-term series of stream gauges, USGS stream gauges that monitor mostly suspended sediment, but some bed load back in the ‘80s to understand that transport issue. And one of the tools that’s been used a long time is turbidity, which many people are familiar with in water quality world.

But a turbidity sensor is – it’s in situ. It can foul. It has a certain range. They’re expensive. They’re very well established, and they work great. So my SedCam is something that I came up with a number of years ago, and it’s just been slowly, slowly in R&D and it’s really starting to take (inaudible), which is great. So I think five states across the US have a SedCam system that I’ve built now.

Austin Madson: Oh cool. That’s awesome.

Adam R. Mosbrucker: What it is, is basically a – there’s two cameras in a box, and they’re mounted above a river surface. Take a picture say every 15 minutes and then the color profile, the spectral signature in those images gets evaluated and regressed against physical samples. So you basically get an estimate of suspended sediment with a picture of the river surface.

Austin Madson: So you can only use SedCam where you have some of the institute data to at least derive those initial relationships.

Adam R. Mosbrucker: That’s right, yeah. Yeah, it requires that yeah, and every site is a little – just like turbidity and other sediment surrogates. Every site is a little bit different. So you have to build that relationship at different sites.

Austin Madson: Right, yeah, and mostly that stems from the kind of difference in material that’s being transported or the optics or the angles and – or a combination of those I guess?

Adam R. Mosbrucker: I think it’s mostly the mineralogy and the different materials, yes.

Austin Madson: Yeah, Adam, at the beginning of this question you had mentioned some pretty cool applications that we haven’t touched on. Things like monitoring dome growth or edifice volume calculations and things. Can you one, tell our users and myself what an edifice is for those of us that aren’t super familiar? And how do you go about actually monitoring dome growth using lidar and other methodologies as well as this edifice volume calculation?

Adam R. Mosbrucker: Yes, so we’ll split it. Dome growth is something that mostly my colleagues at CVO are concerned about, say during an eruption. So I started at CVO in 2009. The last dome building eruption at Mount St. Helens occurred between 2004 and 2008. So I haven’t personally had the pleasure of experiencing that yet. But I have helped, and I know that I’m familiar with the topic.

So during say a Cascades eruption, like St. Helens, effusion rate, like Mike was talking about, which is basically the same thing, as the volcanic dome is growing and we want to know the rate that it’s growing so we can understand that eruption process. So it’s down to volume again. We either collect a lidar scan, which is infrequent but very high quality. And then that’s often supplemented by either traditional photogrammetry, there was – we had a lot of that during the last St. Helens eruption. There’s a lot of aerial photogrammetry shot and then some other colleagues at CVO have used SfM photogrammetry or a very old school version of SfM to do just repeat surveys of – frequent surveys of that dome growth.

Austin Madson: Can you talk a little bit about the edifice?

Adam R. Mosbrucker: So an edifice is basically a mountain. You can think of it like the large structure that a mountain is. We call it the edifice from – if you were hiking you’d be hiking in the foothills, and you’d start to climb the mountain and then you’re on the edifice. You’re on the flank of the mountain. And so to understand eruptive histories a lot of Cascade volcanoes are strata volcanoes. So they’re a composite of different eruptions over thousands of years. So you have an eruption build up, and then some gets eroded away or glaciated, et cetera. And then another eruption happens, builds up again.

So you have this complex history. So one really fun way to understand say an eruptive volume with really, really old eruption is to use geologic map units. The whole country pretty much has a – we have a geologic map of everywhere. But our volcanoes, most of them have pretty detailed geologic maps. So we show a particular lava flow or a particular pyroclastic flow deposit or other things.

So if you know the dates, the approximate dates or eruptive period of those geologic units, you go into GIS. You clip those out. Extract the – say a lidar dataset of terrain on those units and then interpret terrain between them and calculate a volume. And you can do that a number of times and understand the sort of volume of each or volume of change. You have to make big estimates about glacier erosion and other things. This is like – lidar is almost a – it’s too high precision. It’s more precision than it needs to be.

Austin Madson: Totally, yeah. It’s nice that you bring up these kind of historic pyroclastic flows. One of the reasons that I got into earth science and remote sensing to begin with was this animation of the shrinking and swelling of Mount Etna, this classic InSAR-based animation from a couple of decades ago. And then one of my brothers went out to Sicily and brought back this really beautiful map of the historic pyroclastic flows of Mount Etna. And I have – I spent way too much money framing it. So it’s on my wall right now {laughs}. It’s really beautiful.

Thanks for enlightening us there Adam. Let’s switch back over to Mike. Mike, can you talk a little bit about how HVO uses lidar to kind of better constrain some of these geophysical models of the eruption and some of those underlying volcanic mechanics? I know we talked a little bit about effusion rates and Adam just touched on dome growth and edifice volume. But I wanted to kind of get your line of thinking on this as well.

Michael H. Zoeller: It’s similar here at HVO in that high resolution topography datasets allow us to get pretty accurate numbers on the amount of lava that’s been erupted from the entirety of an eruption. And if we have mid-eruption collects of high resolution topography, whether it be from photogrammetry or lidar, then we can kind of use those as time stamps to get the amount of lava erupted over unit time between those different collects.

And looking at the total for the entirety of an eruption, that’s often compared with geophysical datasets such as InSAR. Also, like our network of telemeters and GPS instruments, which those instruments, they can provide an approximation of the amount that the magma chamber has deflated during an eruption. And if our numbers from the topographic modeling, checking volumes that way, if they line up with what the geophysics is telling us, that’s a really good indication that the system is in some degree of equilibrium. And all of the stuff that’s exiting the magma chamber is coming out of the ground where we think it’s coming out of the ground.

Now if there’s a mismatch, that might be an indication that it’s entering into some sort of storage reservoir elsewhere in the volcano and to keep an eye out for potential seismicity or deformation associated with another region that could eventually have the potential for eruption.

So that’s kind of where the topographic modeling comes into play. A really good example of that at very fine scale occurred during the series of eruptions that we had in Kīlauea summit caldera, between 2020 and 2023. That was an area where we had contracted a lidar survey from – it was an aerial lidar survey from a helicopter that occurred in, I think it was July of 2019. So we had a really nice one meter resolution DEM of the whole caldera. And when this started filling in with eruptions, the first one beginning on December 20th of 2020, the bottom of the crater started rising up as lava started to stack there down in the bottom.

And we were able to track meter scale fluctuations in the surface from a combination of photogrammetry modeling and lidar surveys from that borrowed terrestrial laser scanner from the Cold Regions Laboratory. So we were able to track effusion rate fluctuations on temporal scales of, like, you know, usually a few days. We found that the lidar collects were always better than the photogrammetry because as that eruption wore on and a typical pattern for Hawaiian eruptions is really high effusion rates at the start, and then it kind of tails off.

As it tails off the surface change might only be the order of, like, a meter or two between our different collects. And those might occur a week apart, and it might only – the crater floor that’s rising up with new lava might only change by a meter or two over that time scale of a week. That’s pretty close to the measurement accuracy of our photogrammetry methods, like over an area that’s maybe two to three square kilometers. The thing might be tilted slightly. So it doesn’t line up perfectly with the previous collection’s topography.

But with lidar we have a high degree of confidence in the accuracy of those X, Y, Z measurements coming out of the laser scanner. So it was really nice having that terrestrial laser scanner for a few years there. And like I was saying earlier, we’ll be able to do the same with the airborne system that we now have going forward.

Austin Madson: That sounds like you’ll have a lot on your plate with the VUX-120 and some really cool ideas on what you can do with it. To kind of wrap things up here I want to share a really funny, embarrassing story that’s related to volcanoes and HVO in kind of particular I guess. But my parents used to travel once a year when we were younger. I was around eight. And they said, okay, well, what do you want? We’re going from Hawaii. So my favorite color at the time and still is, kind of this, like, burnt orange, umber color.

So I told them or asked them politely, “Could you bring me back a jar of lava? {Laughs} I want lava.” And so I didn’t know, and at the time they came back and they bring me this lava rock. And so okay, I was a little bummed, and I kept the lava rock. It’s not orange. It’s not the lava that I had anticipated. And then years later I was watching something, who knows, on TV or something. And apparently it’s bad luck to bring these lava rocks back. And once I realized that I had no idea at that point where that lava rock went. And so I can’t decide if I’m running through my life right now with bad luck because I didn’t ship it back, or good luck because I lost it.

Michael H. Zoeller: I think you’re doing all right Austin, but yes, you are correct. We don’t encourage people to take rocks off the island. {Laughs}

Austin Madson: I guess I just called out my parents for doing something illegal. Maybe it wasn’t illegal back in the day. Maybe they weren’t even my parents, you never know.

Adam R. Mosbrucker: Is it different if you’re a geologist, Mike? Because I will not admit to taking any rocks from Hawaii.

Michael H. Zoeller: I mean, we ship rocks off to many labs around the country for analysis. To my knowledge none of them have had major fires or bad explosions. So yeah, if it’s being used for scientific reasons that’s okay. But yeah, the tourists here are discouraged from taking rocks. In fact I know that the post office in my home town of Volcano, over the course of a year they’ll get, like, kilograms worth of rocks shipped back from – to them from people who have taken them off the island and I guess some bad luck has befallen them. {Laughs}

Austin Madson: Okay, well, so the moral of the story, I guess, is, don’t take the rocks.

Michael H. Zoeller: Correct, yes.

Austin Madson: All right, well, I appreciate you all’s time today, Adam and Mike. That’s all we have for this episode. Thanks again for chatting with us, and thanks to everyone online for tuning in. I hope you were able to learn something new. And if you haven’t already, make sure to go ahead and subscribe to receive episodes automatically via our website or Spotify or Apple Podcasts or pick your poison. Stay tuned for other exciting podcast episodes in the coming weeks. Thanks again to Adam and Mike and take care out there.

Announcer:           Thanks for tuning in. Be sure to visit lidarmag.com/podcast/ to arrange automated notification of new podcast episodes, subscribe to newsletters, our print publication and more. If you have a suggestion for a future episode, contact us. Thanks again for listening.

This edition of the LIDAR Magazine Podcast is brought to you by rapidlasso. Our flagship product, the LAStools software suite is a collection of highly efficient, multicore command line tools to classify, tile, convert, filter, raster, triangulate, contour, clip, and polygonize lidar data. Visit rapidlasso.de for details.