Short Wave - Predicting Landslides: After Disaster, Alaska Town Turns To Science

Episode Date: October 3, 2022

On August 18, 2015, in Sitka, Alaska, a slope above a subdivision of homes under construction gave way. This landslide demolished a building and killed three people. Today on the show, host Emily Kwong recounts the story of the Kramer Avenue landslide and talks about how scientists and residents implemented an early warning system for landslides to prevent a future disaster. See pcm.adswizz.com for information about our collection and use of personal data for sponsorship and to manage your podcast sponsorship preferences.NPR Privacy Policy

You're listening to Shortwave from NPR. Okay, Aaron, I want to tell you a little bit about Sitka. So this is the town in Alaska where I used to live and report. And there was one day that honestly changed people's relationship to the weather. Emily, I love your Sitka stories. So I'm all yours. Yeah, Sykka's on an island on the edge of the Pacific Ocean. In the Tongass National Forest.

It's beautiful there. And it rains a lot over 100 inches a year. year. We have beautiful rivers with salmon, and the salmon need rain. Our ecosystems need a lot of rain. This is Lisa Bush, executive director of the Sickest Sound Science Center. You know, all of this, it's like, we are rain people. We are people of the rain. And Lisa never feared the rain before, but the morning of August 18th, 2015 was different. I remember my pants getting wet, like all the way up to my knees, just walking from my car to the airport. So I remember, thinking this is a lot of rain, a lot of rain. Rivers in town began to rise and the land

started to slide. 40 landslides were documented on Baranoff and Chichikov Islands that day. A slope above a subdivision of new homes under construction gave way. This landslide, the Kramer Avenue landslide, demolished a building and took the lives of three Sitkins, brothers Elmer and Ulysses Diaz and Sikka's building inspector, William Stortz. I heard the news at the radio station. Lisa heard it on the plane. I mean, I knew all the people who died. I knew all the people who were affected.

I knew lots of people who were searching for these people. It's hard to get your head around. And the feeling in town that day was just one of precarity. The sky was so gray. There were these thick clouds that were almost pressing down. And the mountains, you don't think about mountains moving, right? But looking around, the geology suddenly didn't feel so, static. For days, Sitkins were shoveling debris, cooking casseroles, keeping vigil with the families

of those lost, and coordinating shelters for those evacuated. I can stay with a friend, so my whole house is open. Charter is his room. These are locals reading Facebook posts written at the time. My home is very small, but I can offer food, blankets. Keith Perkins of Sitka wrote this, the day of the last missing person was recovered. The sunset tonight in Sitka, seven days from a horrific moment, a day of warm sun, the day Sitkans helped bring out our third friend home to his family,

the day that marks closure, the day that begins a healing process for the community. Emily, no matter how much you read or hear about natural disasters in other communities, nothing prepares you for when it happens to your own community. Absolutely. And I wanted to take you back to this moment so you could understand everything that happened after. Because the thing for Lisa, the Sickest Sound Science Center, to do was to get answers. Why did this happen? Why did this rain tip the scale the way it did? Why Cramer Avenue? And how do you stop a tragedy like this from ever happening again? Yeah, I mean, those are huge questions. Who do you call for something like that?

You call scientists. Of course you do. You know, my boss at the time, news director, Rob Walsy, said, the answer is literally a puzzle and everyone's holding a different piece. And Lisa dialed everyone, NASA, the National Forest Service, the National Park Service, the U.S. Geological Survey. They responded so quickly, yes, how can we help? All these scientists said, yes, we can help you with our expertise. With the goal of creating something Sitka and really the United States had never seen before, a community-based early warning system for landslides.

This is the story of how scientists and locals found common ground to protect a community from disaster. I'm Emily Kwong. I'm Erin Scott, and you're listening to Shortwave, the Daily Science Podcasts from NPR. So, Emily, introduce us to Sica's landslide early warning system. I am going to do you one better, Aaron. It just launched last month, so can I show it to you? Yes, please. Website app, what should I look at?

It's a good website. It's like a digital dashboard. So take out your phone or your laptop. Type in sitka landslide.org. Tell me what you see. Okay. So I'm looking at a clean page. There's a green checkmark that says the current risk of landslide is low. And the 24-hour forecast is also low. So this is great. This is reassuring. It kind of works like a traffic light system. Right, right, right. And you saw for yourself. It's really simple to look at. But developing a system that's both science-backed and user-friendly took seven years, a $2.1 million grant from the National Science Foundation and the involvement of an entire town.

Yeah, a town that had just gone through a massive tragedy. Yeah. I mean, parents, a year after the landslide, were still calling the school district when it rained asking, is it safe to send my kid to school? And the superintendent, she was used to determining a snow day based on road conditions, but she had no frame of reference for. when to call a rain day. So if you're surrounded by these hillsides, how do you know which ones are at risk of sliding? Yeah, this is the wear piece of the puzzle. Where do landslides happen? I'm going to let Jason Schmidt, a local geologist, show us that.

Hey, Emily, I'm here in the field. Jason is the geoscience coordinator at the Sick of Sound Science Center, heavily involved in this project. I asked her to hike us to the origin point of an old landslide. A debris flow happened here a couple years ago. And I'm at the headscarp, so it's the place where it all started. A debris flow, that's the type of landslide we're talking about here. It happens when you get a lot of water in the system. Water that travels down through gravity and transforms layers of earth into a slurry of mud, water, and other debris, taking on the consistency of wet cement.

It can move up to 25 miles per hour, giving you very little time to get out of the way. It almost looks like the mountain has liquefied and has turned into a river coming towards you. And so this is what happened on Kramer Avenue in 2015? It is. And Jason says a part of what makes Sica landslide prone is that the land is young. It hasn't had a lot of time to pack together and consolidate. And you can see the layers that are making up the geology of slopes here. And if you read the layers, you learn the story of how Sitka came to be.

So the top layer is the Tongass Forest, the pretty stuff. where these grasses and mosses and berries, blueberries are ripe. Underneath that is the volcanic ash, which is called tephra. It's a little bit orange in color. And when I rub it between my fingers, it's super fine material. And this tephra settled out of the air from blankets of ash, spewed up by Mount Edgecombe, the local volcano, about 10,000 years ago. And then underneath the tephra, there's some glacial till.

It's this gray, unconsolidated. when you run your hand over the rock, it just falls off. And it's only below all that you finally reach the bedrock. So the layers of land, they're just not packed together super tight. It's more like a loose lasagna. Emily, the fact that she can rub the teffre between her fingers and that the glacial till like breaks off when she just brushes against it with her hand. Like it's scary, how unstable that sounds.

It is typically very stable, by the way. It's just when it rains. Okay. But the soil, it does contain a clue. There are scars of old landslides all around Sitka. And geology tells us that new landslides are likely to form in the footprint of these old landslides, which is telling. So when Sitka's Geotask Force started discussing that, the Forest Service, Lisa says, was like, hold on. Well, we have a landslide inventory that we've been keeping track of for the last 50 years in the area.

Would that be helpful? And we were all like, what? Historical data, yes. Hello, we didn't even know that they were doing that. I love it when discovering a database is like discovering hidden treasure. Yes, and that trove of data wound up in the hands of Annette Patton, a postdoc at the University of Oregon and now lead geologist on this project. So with a sense of how slopes have failed before, the ridge above Kramer Avenue being one of them, Annette, along with Josh Roaring at the University of Oregon, wanted to know what amount of rain tips the balance. So like, do certain amounts of rainfall predictably lead the landslide risk?

Something like that, yep. Like if it rains really hard for a day, is that what triggers a landslide? If it rains really hard for an hour. So we didn't know for sure exactly what timescale of heavy rainfall would trigger a landslide. So that's where we wanted to start. So Annette combined this landslide inventory that the Forest Service happened to have with Sitka's rain record, 20 years of data gathered at the airport. And after a lot of number crunching and statistics, she started to see a pattern. Five of the most catastrophic landslides in the last decade,

ones that blocked roads, destroyed human life and property, they all saw a certain amount of rain in a three-hour interval. Oh, this is like a rain mystery. So how much, how much rain? It's not an absolute because the model is more designed to calculate probability. So a high-risk probability warning is triggered around 1.3 inches of rain in a three-hour interval. Before the Kramer Avenue landslide, 1.7 inches fell between 6 and 9 a.m. So it was like a whole day's worth of rain happened in that three-hour time period. That's a hard rain. So what's great about this, though, is that it's taking this unpredictable thing,

rainfall and landslides, and then putting some order on it. And we can then start to decode what might lead to these disasters, right? That's the question. And the pieces started coming together. We had where, we had why. Annette built an algorithm that pulled all of that information into one place. And that's the basis of Sica's early warning system. The Science Center, they also installed soil sensors in the three slopes most at risk,

Gavin, Harbor, and Verstovia, and 10 of these tipping buckets in people's backyards. These are just designed to measure local rainfall totals. And so they're building up infrastructure on the ground. together even more data in the years to come and kind of move into the future of this. Yeah, it's very Alaskan. You know, if you give people the chance to get involved, they will get involved. So the Science Center, they consulted the community as much as possible. Sika Tribe of Alaska got involved.

They contributed traditional knowledge about landslides and human movement. That's on the dashboard. And to make sure the warning system actually reaches everyone in town, the Science Center decided to map the social networks. Robert Lembert led this part. He's a senior scientist at the RAND Corporation and did all these co-design workshops. We ran a game, exercise, where we asked everybody in the room, and there's about 20 people or so, to fill out a little form and say, whom in the room would they take shopping for clothes to get good advice on, you know, what to buy?

I love this, Emily. So the people we trust most are the people we know will give us honest feedback on what we're wearing. Yeah, and certain names kept coming up again and again. Social network mapping identified 50 super connectors in Sitka, 50 people who should know about the landslide early warning system. And this idea that you've got individuals who have worked through the process who now, you know, trust this body of information, trust each other. Basically, any innovation in geohazard science is meaningless if there aren't innovations in human preparedness.

And scientists taking the time to do this work thoughtfully and people willing to show up for that. This kind of collaboration, it's only becoming more important, right? I mean, these landslides, they're connected to climate change. Southeast Alaska is going to see more extreme rainfall. Yeah, I mean, a lot of places are going to see more rainfall, which raises the question are other communities taking note of this project. Yes, yes, communities who also worry about landslides and other things, which is why the Science Center hopes to bring this predictive modeling to six other rural and tribal communities in southeast Alaska. The NSF-funded project, which is called Kutte, which is the sling it word for weather,

hopes to create a regional system for warning people about flooding, avalanches, and landslides. These natural hazards can't be stopped, at least not yet, but Lisa Bush says people can learn to live with them. To me, it's heartening to see a community adapt and, move on. Yes, we have to live among landslides. We have to live in a changed world. And that's not easy. But it's heartening when people do it. Especially, Aaron, I think, at the local level, to keep people safe. I can't think of a better use for science. Emily, thank you for bringing us this story from Sitka. My pleasure, Aaron. This episode was produced by Chloe Weiner and edited and fact-checked by Gabriel Spitzer.

The audio engineer was Co. Takasugi-Shernevin. Special thanks to KCAW. Sitka, especially Robert Woolsey, and Catherine Rose. And to the following sitcom. Megan Pastrannock, Brooke Schaefer, Paul Norwood, Keith Perkins, and Jason Schmidt. Beth Donovan is our senior director, and Anya Grunman is our senior vice president of programming. I'm Emily Kwong. I'm Erin Scott. Thanks, as always, for listening to Shortwave from NPR.