Episode 4: Debris Flows and the Fire-Flood Cycle

Julia Ehlert Nair: Welcome to After the Fires, a limited series from the Caltech Science Exchange. I'm Julia Ehlert Nair, a writer and science communicator here at Caltech. In this limited series, we're sharing what Caltech scientists on campus and at JPL are learning in the wake of the LA fires as they're learning it. And in this episode, I'm joined by two special guests, professor of Geology, Mike Lamb and graduate student Emily Geyman. Thank you so much for joining me.

Mike Lamb: My name's Michael Lamb and I'm a professor of geology here at Caltech.

Emily Geyman: I'm Emily Geyman. I'm a fourth year Ph.D. student working with Mike

Julia: Mike's research group at Caltech studies how water, sediment transport, and other processes are changing the earth's surface. Following the Eaton fire in January this year, Mike's group, like many at Caltech, turned into its attention to studying the fire's impact on the environment. More specifically, they have been exploring the science of post-fire debris flows. We'll hear first from Mike about how debris flows work and then from Emily about the research the team is conducting in the aftermath of the fires.

Mike: Debris flows are a natural occurring mixture of rocks and water and mud that come off of steep mountains and hillsides. They're kind of a mixture between a water flood and a landslide. So they move a bit like water floods in that they can go long distances and they can be quite devastating, but they carry almost as much rocks in them as they do water. And so, that makes them even more deadly and devastating than water flow.

Julia: I think most people use the term mudslide or landslides, but you use the word debris flows. Is there a reason for that?

Mike: These flows carry all kinds of things in them. They carry mud, which is small little particles. They carry sand, they carry gravel, but they carry even big boulders like the size of cars and they carry tree trunks. Depending on where they go, they pick up anything that's in their path. So debris is meant to be a sort of catchall term to describe that whole range of material that's being carried. I think mud flow doesn't quite do the hazard the justice that it deserves because you think of mud and it's squishy and maybe not that dangerous, but if you see boulders the size of cars coming down at you going 20 miles an hour, then it's a major hazard that can tear through homes and be quite devastating. So, I think debris flow is a better term.

Julia: It really captures the risk that they can present.

Mike: Yeah, that's right. So we have a whole bunch of debris basins that here in the San Gabriel Valley in front of the San Gabriel Mountains that protect a lot of the communities from these debris flows. So they're actually happening a lot more frequently than people realize because they really only get in the news when the debris basins get overwhelmed and this muddy and sediment rich water gets poured into neighborhoods. But debris flows are happening all the time every year where they're just not big enough to overwhelm the debris basins. That might lead to the perception that these debris flows are that muddy water, but it's just the material that's over topping the basin and the bigger, more dangerous material is being caught up there.

Julia: And can you describe what a debris basin is and looks like?

Mike: Well, they come in all different shapes and sizes and they were built at different times. Most of them are relatively small, maybe the size of a football field or smaller. Some are quite a bit smaller. They're usually just a concrete dam that then holds back whatever comes into it. They often have a drain in them. So because they're really designed to capture the sediment and not the water, so they want all of the sediment, the rocks and whatever to be delivered and caught there, but then the water would drain out as it goes. And so if you live in the San Gabriel Valley, there's probably somewhere through your neighborhood, there's a sort of a concrete canal that then delivers water down to a network of other concrete canals that ends up going out the LA River. I think it's easy to take for granted that all of this infrastructure is here. I mean, I also live right next to a debris basin and I love living here. We live in paradise in many ways, and that's afforded in large part because of these debris basins.

Julia: So now that we understand the risk and the infrastructure, can you tell us why there is a higher risk for debris flows after a fire?

Mike: Well, that's a great question. We know that there is a higher risk. Debris flows can happen at any time. They tend to happen during big rain events, but after fire, even moderate rainstorms can trigger debris flows. There is a debate about why actually they are more common after wildfires. There's different hypotheses out there.

Julia: Since this is Caltech and we are interested in some might say—obsessed with—figuring out how and why things work the way they do, we're going to take a detour into these hypotheses. Then, we'll get back to future debris flow risks from the Eaton fire. Okay. Hypothesis number one.

Mike: The plants are up there. They're on the hillsides and their roots are anchoring the soil. And so you might think, well, without the plants or without their roots there, when it rains on the soil, it's much more likely to give way and make a landslide. And this is certainly a mechanism that does cause landslides, but probably is not the reason why we have debris flows after fires. Most of the time the fires aren't severe to burn out the entire root network, the vegetation comes back pretty quickly, and that's because for the most part, a lot of those roots are still there. So the classic idea of burning the vegetation doesn't seem to work.

Julia: And for hypothesis number two.

Mike: When you burn vegetation, some of that plant material becomes vaporized and creates a sort of sticky layer inside the soil. And that layer is water repellent. So when you rain on the soil instead of that rain soaking into the soil like it normally would, it runs across the surface. And because these surfaces are steep, they're steep hillsides, the water goes quite fast, and then it can erode material and pick it up. And somehow that then builds on itself and makes a debris flow that probably is happening in some places.

Julia: And finally, Mike's favorite hypothesis, number three.

Mike: The third hypothesis is one that I like and I think is probably the dominant mechanism in our local mountains. I call it the dry hypothesis because actually the mountains lose a lot of their soil without any rainfall. So they're so steep that just by removing the plants, the loose sand and gravel that's up on top of the hillsides just rolls and bounces its way down the hill. As the fire is going on, this happens because the plants are kind of acting like little dams holding the sand back. And when you lose those little dams, it's just free to roll down the hillside. It collects in the channels because those are not as steep and it sits there until it rains and then gets flushed out of the channels. Some of the measurements that we've done after the eaten fire have showed that this is indeed happened during the fire and during the storms that we've had this past winter.

Julia: So those are the three reasons scientists think debris flows, follow fires. But back to those measurements, Mike just referenced his team has been on the ground and in the sky since January when the fires occurred, collecting ephemeral data in real time. What have they seen?

Mike: We had a few light rain events that happened after the fire, and then we had one pretty good sized rainstorm that happened February 13th and 14th. And that rainstorm was a typical big rain event that we would have here about every year, and it produced quite a lot of debris flows. Most of the debris basins did their job again and captured those flows. But we did see, even though it wasn't all over the news because the debris basins did their job, there was quite a lot of material that got moved in the mountains. Emily can tell you some of the specific numbers.

Julia: Yeah. Emily, could you tell me a little bit more about what you've been finding in your research?

Emily: There's about 22 of these debris basins that LA County maintains along Altadena and Pasadena and Sierra Madre along that mountain front that got burned and 21 of the 22 debris basins did see major debris flows during this mid-February range. So there are debris flows basically everywhere, and it's sort of amazing. It seems like the size of these debris basins was just big enough, but barely to hold the material that came out of the mountains. And there's one that overtopped a little bit Sierra Madre dam. It was sort of in some ways like the perfect design storm event that removed a ton of the hazard. It cleared out a lot of the sediment from the mountains, but then it stopped at the right time once we had filled up our capacity and the infrastructure and it stopped. So it was really lucky in some ways.

Julia: It's really great to hear that those basins are doing their job just as designed. Can you share a little bit more about what you've been looking into to understand how this debris flow activity is happening?

Emily: So Mike mentioned this when he talked about the dry hypothesis, but one of the predictions from this hypothesis is that during the fire when all of the vegetation is getting incinerated on the hill slopes, that's the time when you see the sand and the rocks and the boulders kind of come out of the woodwork and bounce down the mountain and form these big piles of loose sediment. It can be often like weeks or months before the rainstorm comes, which would produce the debris flows. You already get the signal that you can go measure and ask how much volume of this material is there ready to go in the catchment. You can sum up all that volume and ask is it more or less than the capacity that we have with the debris basin at the base? And so that was the sort of the most simple exercise that we worked on.

And it turns out that if all of the material that came out from during the fire, if all of that did come down in this one storm, it would overtop almost all of the debris basins, but the size of the storm was such that it only cleared out about 60% of that material. So there's still some up there, but it's probably a little bit less than what came out in the last storm. And that's sort of perfect for us. We already know that the debris basin infrastructure was the right size to meet the hazard last time, and it'll probably be about the right size to meet the hazard when we get another big rainstorm next year. That's sort of the current working model.

Julia: So it sounds like if the rainstorms had been worse in that first round, we might've seen a lot more overflow than what we actually did.

Emily: It could be, yeah. Some of the, I think areas for ongoing research is trying to understand what is it, what are the critical conditions that are explaining these times when the material will hop halfway down versus come all the way out all at once when I think it poses the most risk for our infrastructure.

Julia: How exactly are you measuring the amount of sediment and material?

Emily: So most of what we've been working on is with lidar, so that's light detection and ranging. It's sort of like the light equivalent of radar. So you send some device above the mountains, it could be a drone like we've been doing a lot of our measurements from A UAV or it could be a small low-flying aircraft, but something goes over the mountains and it sends little laser pulses down to the ground and you measure the time for that laser pulse to return and you can get really detailed topography information. In many ways, we're kind of lucky about the experimental design that LA County has given us in the form of these debris basins because we are both lucky in the timing of the sequence of events of the storm. And when we were able to acquire data, the debris basin, it's this amazing funnel that gives us a chance to kind of measure the volume of the sediment very precisely at the bottom.

And it just so happened that in the week or two weeks after the Eaton fire, we knew that we wanted to collect these airborne LIDAR data after the first storm. And so there's a national NSF funded national Center for Airborne Laser mapping that has one of these aircraft and the pilots and the data engineers who are the experts of running these lidar systems and acquiring airborne lidar data sets. And so we applied to them for funding to come after the first storm and survey the catchment. And it just so happened that we got the aircraft to fly about a week after the rainstorm producing all these debris flows, which was the perfect amount of time for the debris basins to run out of water. So all that you were measuring when you flew over, it was basically just the dry sediment.

Julia: So is this related to the modeling work that you've been doing?

Emily: I think we're pursuing two parallel paths. One is a observation based sort of exercise where we have been lucky to be able to, in this perishable time window between the storm and between the fire and the first storm, we got to collect a lot of data. Data that hasn't been collected very often before just because there's only a couple weeks and it's a time when emergency responders are really busy doing very important things, but maybe a more useful thing to be able to ask the question before a catchment even burns, should we be able to know what the hazard is that's going to come out of it? How much material will come out of the hill slopes? What kind of rain event will produce debris flows?

Julia: Emily uses all this data she's collecting and creates a model, a computer program that predicts what's going to happen. Hers is a physical model. It takes all the inputs from her measurements to give a prediction, but there's kind of model that's also useful, an empirical or historical model.

Emily: And so the US Geological Survey, they have a huge team that works on debris flow modeling and they've collected these great historical data sets. They have incredible records about the sizes of debris flows, what the rain events were that produced, those debris flows. And so the USGS kind of trains a hind cast empirical model to look at the past data and try to learn what are the factors that cause these debris flows. And those are the straight of the art models and they're actually pretty good. But one thing that they can't do very well is adjust to changing conditions. I think a lot of opportunity from these new data to build sort of physically based models that are more flexible to be able to account for changing conditions that we should probably expect in the coming decades.

Julia: What do you find most interesting about those models?

Emily: I think what I like about them is that I think if you feel like they have the right processes in them, you feel a little more comfortable extrapolating them to other places. And so I think you have less trust that the empirical models are going to do well because they're just trained on less data or they're trained from data from a different place. And so if you get a physical model, I think you have more confidence when you transfer it to another setting that you're going to get the right answer.

Julia: That makes sense. How exactly does your field work inform the modeling work? Mike, do you want to speak to this one?

Mike: Coming back to the hypotheses we have about where the debris flows are coming from, they could be coming from the rainfall hitting the soil on the hillsides, or they could be coming from this material 10 feet of material. I said that's already piled up and loaded in the channels and the physics of how those two different processes would operate are quite different. And so if you want to build a physical model like Emily's talking about with the right mechanisms in it, you need to know what those mechanisms are. And at this point, really until we've had the sort of data that Emily's talking about in the recent fires, we have no idea what the mechanisms are. I mean, the hypotheses are hypotheses but not tested and they're not tested because we don't have or we have very little observations up inside of these steep terrains when the debris flows are starting.

So these surveys, airborne lidar surveys that Emily was talking about are allowing us now to really make a high resolution digital model of the mountains exactly what they looked like before the fire happened, exactly what they looked like after the fire happened, but before any rain and then exactly what they look like after each rain event. And by differencing those we're able to see what changed. Did the hillsides go down in elevation because they lost some of their soil? Did the channels go up in elevation because they became a place where that soil temporarily accumulated? So those observations are really pivotal to even starting to think about making physical models. Emily's been working hard at pushing this forward and like she was saying, the precision that you need to make the measurements is really high. If you're off by a centimeter, which is very small averaged over the whole mountains, you end up with a huge extra volume of material that shouldn't be there.

So you really have to dial this in. And this is Emily's expertise. She's very good at working with these data sets and getting the noise down so you can see the signal, even if the signal is quite small. So these laser airborne systems are a lot better than they used to be. And in part that there's new ways of processing the data, some of which Emily has been developing herself that really allow you to beat down the noise. So we haven't been able to do this partitioning between how much is coming from the hills and how much is coming from the channels. And when we tried before and we added them together, it didn't equal what came out in the basin. So we knew we had something wrong, but we weren't sure what it was. So in the hot off the press data that Emily has just been working on this week, it looks like we can solve that mass balance problem. We add up what the differencing between these laser scans show and the sediment that was lost from the mountain range is what accumulated in the debris basin. So we have good confidence that the noise is gone and we're getting the right answer. So I was guessing that more would come from the channels than the hillside, but I was still expecting it to be a mix. But the hot off the press data show that the hillsides are basically negligible and that nearly 90 some, 95 or more percent is coming from the channels. The favorite hypothesis.

Julia: On the topic of safety, obviously this possibility of a debris flow is a pretty big disaster that could be very harmful. How does somebody find out if they're in a dangerous area for debris flows?

Mike: It's a good question. I don't think there is a great resource at this point for having debris flow hazards for fires. There's fire ratings that happen to different properties. There isn't the same thing that happens for debris flows as far as I know. If you buy a home, you don't necessarily get a disclaimer about whether you're in a debris flow zone or not. But that said, the US Geological survey does issue these reports for debris flow warnings and evacuation zones, and they're all public and you can go look and see what they have issued for in the past.

Julia: If somebody is in a location near where there might be a debris flow, is there anything they can try to do to protect their property?

Mike: The best thing to do is to heed these evacuation warnings that people have. There is quite a bit of effort that people do and cities and local officials do to try to put out sandbags or cement barriers or these sort of things. And they can work for small events like when one of the debris basins over spills, if it only over spills a little and you're just getting muddy water, then sandbags can help to keep that muddy water going down the street and not coming into your basement or your driveway or something like that. But if it's a full on debris flow that's moving cars and boulders the size of cars, sandbag is not going to do much. It's just going to pick those things up. So for little events, these kinds of measures can help, but if there's a big event and you get a warning, it's best just to get out of there.

Julia: Are there alerts that you can sign up for about debris flows?

Mike: Well, alerts, they do issue. So the warning system comes through the National Weather Service. This is currently how it's done. So using the US geological survey model that Emily was talking about, the US Geological Survey identifies rainfall thresholds that they think will trigger debris flows of certain size, and then that information goes to the National Weather Service who monitors storm events as they come in to the area. And if they think they're going to surpass the thresholds for triggering large debris flows, then they'll issue a warning. Now that has to be coordinated across with LA County because it could be that there could be a debris flow, but if the debris basin is capable of handling it, they may not evacuate people there.

Julia: So you mentioned it's more about intensity than the actual volume of freeing.

Mike: That's right. And I'm not sure that I really know why, but it is the case that the debris flows respond to high intensity downpours of water. So even if you had a brief cloudburst that only lasted 10 or 15 minutes, but if it dumps rain, then that could trigger debris flows. Whereas the same amount of material, same amount of water that was spread out over several hours may not trigger any debris flows. So there's something about how hard it rains that's really important for this debris flow initiation process.

Julia: And how does it feel to be a geologist in this moment and to know that you're working on something that has really closely impacted the community that you're part of?

Emily: I think most of all, I think this experience has really made me how many levels of local government are working together from, I think I knew about the work that the USGS was doing at the national level, but when you go below that Cal Fire and California Geological Survey, and then when you go below that, there's LA County. And then when you go below that, there's the particular towns like Sierra Madre that are the ones who are communicating with the residents when there's a storm coming. And then there are police officers who are knocking on the doors in the morning of a storm to tell the few sets of neighborhoods that are below the debris basins that are vulnerable to overtop telling them about the hazards. So there's just a huge chain, and it's been amazing to see the network that comes together.

Mike: A lot of what we do as Geoscientists is trying to piece together the clues of what's happened in the past because we don't get to see it happen, but we get to try to make sense of some of the evidence that's left behind. And so when we do have an event that we seeth something happen like these debris flows, thankfully were, as we said, mostly captured by the debris basin. There wasn't a tremendous destruction from this particular debris flow event. Of course, the fires were super devastating, but the debris flows afterwards pretty mild. It still was a major amount of material that moved off these mountains. And coming from the perspective of trying to understand why are these mountains here in the first place and why are they this height and what shapes them, it's an opportunity for us to really see an event that does shape the mountains. And so that data we've been talking about that we're collecting is going to be very useful I think, for making these predictive models and doing a better job at mitigating hazards in the future, but it's also giving us an important piece of the puzzle about how the earth works.

Julia: Well, Mike, Emily, thank you so much for joining me today, and I think that's a wrap. We also wanted to acknowledge the funding that made this work possible on a short timeline. It came from the Caltech Division of Geological and Planetary Sciences and donors to the Eaton Fire Response Fund, as well as the National Science Foundation's National Center for Airborne Laser Mapping. Thank you for listening. I'm Julia Ehlert Nier signing off for after the Fires, a limited series from the Caltech Science Exchange. Visit science exchange.caltech.edu for explainers on the LA fires and other topics like ai, quantum science, and earthquakes. This episode was produced by the Caltech Office of Communications and External Relations and Caltech Academic Media Technologies.