Haines’ landslide maps unused, what comes next?

Juneau, other Southeast communities also dealing with issues related to how maps are used for policymaking, development

The Beach Road slide in Haines on, Thursday, May 8, 2025. (Will Steinfeld/Chilkat Valley News)

By Will Steinfeld

Chilkat Valley News

Around the Piedad area, there’s plenty of recent memory of the ground moving. That includes Geneva Randles’ household, where her mother and daughter were living in December 2020, when heavy rain caused landslides throughout Haines. Randles’ daughter Grace Comstock remembers the day — her grandmother seeing the unstable ground across the street, and the policeman knocking on the door. Then they evacuated. 

Through the rain, the slope held, and they came back. Five years later, they’re still there. 

The family has a long history in the area: both Comstock and a nearby road are named after her great-grandmother, whose bedroom she now sleeps in. “This is home,” Randles said recently. 

But even with their long history there, the family is now cautious about doing much in the yard. Randles doesn’t want to uproot vegetation, because it could destabilize the slope. 

“I want to be rooted in,” Randles said. “I have these fears all the time, especially with heavy rain.”

It’s not a major surprise to area residents that they live with some risk of landslide. It’s a steep slope, never very far from the sound of running water, with panoramic views of the river below, but also the mass of dirt and rock above. In other words, it’s a lot like many of the neighborhoods in town.  

There is one thing, though, that makes this one stand out, and that’s how it looks on landslide hazard mapping of the borough. On the maps, Randles’ 50-year-old home sits in the middle of a large blue splotch marking the estimated footprint of an old slide. 

The old, buried landslides are notable because scientists say even in an age of massive amounts of data and high-tech modeling, they’re still the most reliable indicator of risk. “Slides happen where slides have happened,” as one adage in the field goes. 

The old slide paths are one of three hazard indicators shown by the state-developed maps, the other two being locations where slopes could fail, and potential paths of future slides.  

It’s been about four years since scientists started the mapping project: digging soil samples, trucking them to Fairbanks, scanning the terrain with a helicopter-mounted, 400,000 pulse-per-second ground-penetrating radar, known as LiDAR. And it’s now been nine months since the finished maps were released to the public.

But as of two weeks ago, Randles hadn’t seen them. Neither had many of her neighbors, or residents of other at-risk neighborhoods where the Chilkat Valley News went door to door.

A few that had heard about the maps said they either didn’t understand them, or didn’t seek them out, because they felt they already understood the level of risk they faced. 

Resident Scott Hansen, for instance, said he was more inclined to rely on what he had seen, than on the maps. 

“I like the stability of this slope,” Hansen said of his property on Hooter Lane, even though the maps show risk. “I’ve been up and down a lot, and there’s a lot of rock here. Particularly in 2020, just watching what happened, I’m not sure if we’ll get any more significant than that.”

—

The maps’ lack of traction so far would have been unexpected news to the ears of decision-makers and experts in 2020, who described the mapping as the primary tool needed to better understand risk. 

“It’s going to take money, time and people resources to do [the mapping] right,” said one federal scientist at the time, De Anne Stevens, who coordinated the United States Geological Survey’s portion of the emergency response. But still, Stevens said, it needed to happen. “We’ve got to get this done,” she told the Chilkat Valley News in 2020, “or we’re going to be caught flat footed again and again and again.”

In 2022, a report on the Beach Road slide by University of Alaska Fairbanks geologist Margaret Darrow conducted the first analysis of that type done in the borough, collecting high-resolution LiDAR and soil samples for Beach Road. What it showed was evidence of old slides along the hillside, which could have indicated the slope’s vulnerability even before it slid in 2020, back when residents and planners still considered the area safe. 

The fact that residents haven’t seen the maps is not a complete surprise, given that they were mainly intended for municipal decision makers, state geologist and co-author of Haines’ maps Jill Nicolazzo said. But even in the borough administration building, and in the assembly chambers, there aren’t yet any definitive plans for how they might be used.

The borough’s agenda for landslide hazard mitigation is contained in its 2022 Hazard Mitigation Plan which lists four action items: completing the mapping project and Darrow’s study at UAF, implementing recommendations from those studies, conducting annual public education, and reviewing requirements for new construction. 

Those tasks were primarily assigned to the borough planner. But in recent years, there hasn’t always been a planner. Former planner Dave Long left the position in 2023, and it sat vacant until current planner Chen Wu was hired this summer. Wu said landslide-related planning and mitigation has not been assigned to him as an official task, and the maps are not a tool he is using. 

That’s not just a Haines story. Nicolazzo says it’s a common problem that borough staff turn over in the time it takes for the science to be completed. 

“Maybe they had a plan for it that didn’t make it to the next elected official,” she said. “It’s just the nature of it. It happens everywhere.” 

The only borough elected body or administrative department that says they are using the maps is the planning commission. There, chair Patty Brown has advocated for using the maps to trigger slope-stability reviews when development plans come up for permitting. 

According to Brown’s plan, commissioners can use the maps to ask for increased scrutiny on projects that show up in areas marked as high-risk zones. That kind of review is important not just for the safety of the property owners, but for the community’s future, she said. 

“You’re not making a decision just for yourself, to say, ‘I don’t mind if my house goes, that’s God’s will, or whatever it is,” Brown said. “The decision you make on your land affects things around you, and in the future, when others buy the property.”

Using the maps to review development does have precedent elsewhere in Southeast. After Sitka’s fatal landslide in 2015, FEMA funded landslide hazard mapping in the area. The community’s assembly then put stricter requirements into place for new construction in mapped hazard zones. 

But it’s only a rough comparison. In Haines, the planning commission hasn’t put the map-based review process into written regulation, or suggested any map-related ordinances to the assembly. Planning commissioners say the “increased scrutiny” triggered by the maps will mean asking developers in high-risk areas to conduct a geotechnical study. The study would collect data about soil, rock, and slope stability properties likely more specific to the site than data used in the mapping model.

But at least for now, that process is just a principle agreed upon among some current commissioners — not borough policy. 

—

Codifying the process, too, still wouldn’t solve a number of road blocks the commission faces in putting the maps to use. 

Talk to Brown and you’ll hear a lot about gravel pits, which makes sense: gravel pits dig into slopes, sometimes in slide-prone areas.  Perhaps more importantly, a number have recently come before the commission during conditional use permit hearings, meaning they’re a category of development the commission has clear jurisdiction over. Other areas of jurisdiction for the planning commission include review of many of the town’s major builds, like the new Hilltop Subdivision or the proposed SEARHC hospital. 

There’s plenty of construction, however, that doesn’t come in front of the commission. In the most common permitting process — for instance to build a home in a residentially zoned area — borough staff simply check construction against a set of general approval criteria, and the planning commission never sees it. Landslide risk, planner Chen Wu said, isn’t a part of the approval criteria. And then there are the hundreds and hundreds of structures that have already been built. 

There’s another problem, too. Even if the planning commission, or assembly, were to exercise stronger authority over how and where people could build, there’s reason to believe residents would not want strict regulation. That was the case in Sitka, where when the assembly tied new regulations to landslide hazard maps, they quickly came under fire. As it turned out, the qualities of the maps that were useful to borough decision makers — best-available data, advanced modeling and a quantifiable display of risk — were also useful to banks and insurers. 

Residents in Sitka’s high-risk zones said the borough’s use of the maps was preventing them from financing and insuring their homes. By 2021, references to the maps in Sitka code were off the books, repealed by its assembly. 

The same story played out in Juneau, two years later. There, the assembly repealed building regulations tied to landslide mapping in 2023, due to the same set of resident concerns. 

The financial implications of the other communities’ maps have left fingerprints on the ones Haines now has. The report that came with them contains a disclaimer, that the maps are “not regulatory and are not suitable for legal, engineering, or surveying purposes.”

Between the written disclaimer and a presentation from Nicolazzo and co-author Martin Larsen, borough officials say they were made to understand certain limitations. As planning commissioner Rachel Saitzyk put it, “we have this LiDAR data, but it’s very clear that we’re not allowed to use it for planning.” 

Part of the limitations are scientific, planning commissioners and scientists say. Something as complex as future landslide risk is impossible to determine with complete certainty. One shortcoming of Haines’ mapping may be soil data that goes into the formula for slope stability. Currently, soil data is based on samples collected during the project, dug up at various points around the valley. That gives hard data about each location sampled. But in between those points, where data is estimated from nearby samples, there’s uncertainty introduced.

“The numbers we (input into the model), we’re making our best judgment, but they’re generalized for a large area, or soil type,” Nicolazzo said. “Change the number by a little bit and the output changes.”

There’s also reason to be uncertain about the back end of the model, too, where the numbers get crunched, Nicolazzo said. The model used for the Haines maps was developed in Oregon, and though it uses “pretty well established methods,” she said, they “might not fit Alaska perfectly, given that we keep having these fatal landslides that are a more complicated landslide that are not quite a debris flow, not quite a translational [slide].” 

In light of that, planning commissioners say their understanding is that the model is not accurate enough to rely on for judgments about the hazard level of specific pieces of property.

“Not accurate enough,” however, is itself a matter of judgment. Though there could be more data and more specific inputs into the model, having some amount of estimation is inherent to any model, said Josh Roehring, a landslide expert who does work in the region. “We have these generalized models that try to account for steepness and convergence, but the site-specific factors are just so hard to predict. That is an inevitability. Every model in the world identifies lots of areas that in any given storm don’t fail.” 

That includes models and landslide hazard maps used in Italy, France, Norway, Sweden and Switzerland. And in those countries, the uncertainty hasn’t prevented the maps from being tied to legally-binding legislation. In Switzerland, for instance, almost no development can occur in the highest risk areas and development is heavily regulated in moderate risk zones — including required mitigation measures and surveying. The Swiss maps aren’t leaps and bounds ahead of what is now available in Haines, Roehring said, describing them as “pretty close to the level of sophistication” of their Swiss counterparts. 

But still, they can’t be used to guide regulation, and Nicolazzo acknowledged that isn’t just because of scientific limitations — there’s also legal liability. After the Sitka landslide, there were lawsuits from homeowners holding the City and Borough of Sitka responsible for damages. The state is hoping to avoid similar lawsuits, Nicolazzo said. 

 “People are concerned about property values, about insurance,” she added. “It is a very sensitive topic. We don’t want somebody to be able to take that map and say ‘Look, your property line is in a hazard zone.’ We’re trying to stay one step back from that.”

—

On hillsides in Haines, residents said they aren’t looking for the borough to put in strict regulations. Instead, there was a different idea — one that came up again and again: an early warning system.

That was particularly true among residents who already felt they lived in high risk areas. Randles, for instance, or Beach Road resident Jon Hirsh. In 2020, a massive slide on Beach Road destroyed homes and killed two residents. 

Hirsh, who has a geology background, said he thinks it’s an inevitability that the hillside above his home will come down again, on some day, at some time. He just doesn’t know when.  

“In my mind, I already know it’s going to go,” Hirsh said. “The main thing for my own life is to know how to make a decision. If it’s really raining and I’m there, I want to know how to decide to leave.” 

In the years since the slide, as people have moved back into homes on the hillside, Hirsh said neighbors call back and forth when the rain starts really falling. Hirsh asks neighbors if they’re monitoring the water or seeing any signs of over-saturation. If it ever looks anything like it did in December 2020, everyone needs to get out of the neighborhood, he said.

But just looking at water flow, things like how filled up culverts are, is, for Hirsh, not a clear metric to make a decision from. He wants some kind of benchmark, built by people with official expertise, telling him when it’s time to leave.   

“I want to have some number to point to. None of that exists, and it’s just between neighbor and neighbor,” Hirsh said. 

Early warning systems like the one Hirsh describes do exist, including nearby. In Klukwan, one is being developed by a team that includes Roehring, the regional landslide expert.

And in Sitka, an early warning system has been operational since 2019. There, soil sensors and weather data feed into a model, which then spits out a three-day forecast, each day labeled high, medium, or low risk of slide. The forecast is displayed on a website run by the non-profit Sitka Sound Science Center. 

Roehring, who also helped develop Sitka’s warning system, said it offers a straightforward way to put the maps to use. “We’ve developed this rainfall-based forecasting system in Sitka, and the question is, who needs to pay attention to that?” Roehring said. “And our answer is, folks that live on old deposits, or that live in these potential [mapped risk zones].”

How could Haines stand up a similar system? So far, there haven’t been obvious answers. 

Local geologist John Norton, for one, sees the project as a tall task in Haines, where the size of the community “makes it hard to muster the manpower needed for these efforts.” 

He’s right, in describing how much work went into the Sitka system. At the time of the Sitka slide, there weren’t trained geologists or geomorphologists working in the municipal government, or at the science center, said Lisa Busch, former director of the science center. 

That meant the expertise had to come from the outside, especially because the municipal government didn’t want to be involved, or responsible for the system, Busch said. The science center ended up “bringing the expertise to town.” The final cost for the early warning website included a $2.1 million grant, and more than four years of labor from scientists at multiple universities, government agencies, non-profits, and a major think tank. Plus annual costs for sensor upkeep and staff.

—

Based on that example, it’s not clear whether Hirsh’s wish will be fulfilled anytime soon. In the meantime, Haines residents will continue relying on their own personal observations, individual expertise, and levels of risk tolerance.

It’s not a straightforward calculus. Not everyone was fully informed of the risk when they moved into homes on hillsides, like Brad Jensen, who moved on short notice last year to Cathedral View Road. By now, Jensen has invested labor and money into the home, and he said he loves the neighborhood. The attachments make it hard to sell, to move, and could weigh into evacuation decisions — though Jensen did say he plans to evacuate in a bad storm. 

But the individual calculus is what’s currently available, and preparation will continue for the next big-enough rainstorm, in ways big and small, with or without the maps. Like Brown, the planning commissioner, who lives in a hazard zone herself. In addition to the working groups and brainstorming sessions, each time she walks down Beach Road during the growing season, she picks fireweed off the roadside and tosses it onto the bare dirt and rock above. There’s a hope, at least, that even though the scar of the slide is north-facing, and not much has taken hold but alders and fireweed — some of it Brown’s, perhaps — someday something might grow back.

• This story was originally published by the Chilkat Valley News.