After 2014 Oso landslide, what did we learn about preventing disaster?

In the decade since the Oso landslide, technicians have meticulously mapped the state of Washington, scanning thousands of square miles of terrain, precise nearly to the inch, part of efforts to forestall another tragedy like the one that killed 43 people when the hillside above Steelhead Haven collapsed 10 years ago this week.

The maps, created using planes that fly over the state taking laser measurements of the ground below, help geologists analyze where landslides have happened, whether a year ago, or 10,000 years ago. Past landslides are the best predictor of future landslides.

Geologists with the state Department of Natural Resources have mapped 34,683 slides over the past decade.

And yet …

While they now have laser-created lidar maps of nearly the entire state, they’ve pinpointed past landslides in only 14% of it. And they haven’t taken the next step, recommended by a number of outside experts, of plugging that data into models to try to map the potential hazard zones of future landslides, and then highlighting where a landslide would have the gravest consequences.

“We live in a steep, wet state, so landslides are an inevitability, but I don’t think many people really know how significant the risk is and where those risks lie,” said state Public Lands Commissioner Hilary Franz. “Oso served as the major wake-up call to everyone.”

There has been considerable action at nearly all levels of government — county, state, national — in the decade since Oso, the deadliest landslide in American history.

Snohomish County, where the landslide happened, is one of the few areas in the state, and the nation, to develop accessible, easy-to-read maps showing the most vulnerable areas. Building on the state’s lidar data, they’ve identified more than 12,000 buildings, housing some 60,000 people, in landslide hazard zones in Snohomish County. Tens of thousands more people almost certainly live in landslide zones in other counties that haven’t done the mapping and projections Snohomish County has.

Franz’s statewide agency now has five full-time geologists analyzing landslide risks, compared to before Oso, when one person spent only about half their time on landslides.

In 2020, U.S. Rep. Suzan DelBene and U.S. Sen. Maria Cantwell spearheaded national legislation, spurred by Oso. The National Landslide Preparedness Act helps fund mapping and analysis and led to the nation’s first ever national landslide strategy. The strategy identifies three broad needs to make landslides less deadly and damaging: filling basic information gaps about landslide risk, mapping and forecasting landslide hazards, and coordinating among the many jurisdictions and agencies responsible. That law is currently up for reauthorization.

All of the efforts are a race against time. The right combination of geology, topography, hydrology, gravity and time will cause the ground to give way, regardless of human action. And climate change will, in a variety of ways, make landslides more frequent. Hotter, dryer weather leads to more wildfires. Fires kill trees, and in the process, weaken hillsides. And when storms do come, they are heavier and more intense, drenching hillsides that may not be able to handle the added moisture.

“One of my grave concerns is that we are going in the right direction and making progress, but I’m afraid that we are being outpaced by the changing climate,” said Joseph Wartman, a University of Washington engineer who investigated Oso and now runs UW’s Natural Hazard and Disaster Reconnaissance Facility, also known as RAPID. “We’re really beginning to see the practical implications, which leads to these cascading, devastating effects.”

But deadly landslides can be avoided. Mapping and analysis can tell us where landslides are likely and, thus, where we should avoid living and working. But knowing where not to live is only part of the battle. Warnings about Oso went back decades, but they were either ignored or not sufficiently communicated to those who lived beneath the hillside that crashed across the North Fork of the Stillaguamish River.

There were noteworthy slides at Oso in 1951, 1967, 1988, 1996 and 2006. The 1967 slide flooded 48 lots of the new Steelhead Haven neighborhood that had been planned less than a decade earlier.

After the 2006 slide, county officials held community meetings and discussed potentially buying out homeowners in the neighborhood. At the same time, “a new homebuyer was not told about the landslide, when it had occurred only a week prior,” a local environmental group wrote in a meeting summary.

Seven new homes were built between the 2006 slide and the catastrophic 2014 slide. Owners or family members of six of the seven homes would die in the slide.

Even now, after a decade of data collection to help predict where the next big landslide might happen, fears of liability and plunging home values can make jurisdictions hesitant to make good use of that data.

The precise causes of the Oso landslide remain in dispute. A lawsuit that blamed the state and a private timber company for logging above the slide a decade earlier settled for a total of $60 million just before a scheduled trial.

There had also been weeks of heavy rain before the slide.

The state’s expert witnesses prepared a report ahead of trial, concluding the recent logging did not play a role.

“Whether or not timber harvest really affected it is debatable, but one could say without hesitation the harvest of the trees certainly didn’t help stabilize the slope,” said former Public Lands Commissioner Peter Goldmark, who led DNR at the time of the landslide and when the lawsuit settled. “The real fault was having that development there. In my mind it was a matter of time, not a question of if it was going to happen but when.”

Maps are not enough

A traditional topographic map will show you where a hillside is and how steep it is. An aerial map (think Google Earth) will show you where houses, roads, mountains, rivers and trees are located.

Neither will show you what the actual ground beneath the tree cover looks like, where it is slumped, piled, split or scarred.

For that you need lidar, an acronym for light detection and ranging, a half-century old technique that has only gained widespread use in the past 25 years or so.

To make a lidar map, a plane or, more recently, a drone, flies back and forth while firing lasers at the ground. A computer system measures how long it takes the lasers to hit the ground and bounce back.

The first returning lasers show the surfaces closest to the sky — trees, bushes, buildings. The last returning lasers, providing billions of measurements, can be digitally extracted to form a precise model of the bare earth below.

It shows faults, cracks and fissures. It shows sag ponds, scarps and hillocks. It shows geologists where the earth has moved, how and where it might move again.

“The single most important thing they’ve done is all that lidar mapping,” David Montgomery, a UW geomorphologist who served on a governor-appointed landslide commission in the wake of Oso, said of the state’s work since the slide. “It really is the best foundation for trying to assess those kinds of hazards. So they deserve major kudos for that program. Have they done enough with it is a very legit question.”

At the time of the Oso landslide, roughly a quarter of the state had been mapped with lidar. Now nearly all of it has.

Oso, and the Stillaguamish River valley, had been mapped with lidar before the landslide. The images showed the scars and terrestrial detritus of thousands of years of landslides in the valley. They showed that the deadly landslide March 22, 2014, essentially was par for the course.

And they showed, with heartbreaking hindsight, that simply having the maps is not enough.

“This lidar data is not really being used to full advantage,” said Dan Miller, a geomorphologist who co-wrote a report 15 years before the landslide warning of “catastrophic failure” at Oso. “If you go to the DNR’s landslide hazard page there’s nothing that I can find that suggests that they’re using lidar now to do anything except map where they’ve seen past landslides. I haven’t found anything that it’s being use to predict where landslides will occur and how far they’re going to run out.”

Casey Hanell, the state geologist at DNR, said it was all a work in progress. While the lidar maps are essentially complete, the state’s geologists have only, so far, analyzed 14% of them, focusing on areas with steep terrain and big population centers. They’ve completed Whatcom, Snohomish, King and Pierce counties and areas around the Columbia River Gorge.

“We need to get that base data first and then that allows us to build on some of that predictive modeling for the future,” Hanell said.

Knowing the hazards

What the state has done is make its lidar repository available to all — to counties, cities, planners, the public.

Snohomish County has taken lidar data and made an easy to find (search Google for “Snohomish County Hazard Viewer”), legible map showing where steep slopes and soil types combine to form landslide hazards.

Teal splotches on a gray map make clear the areas the county sees as hazardous. You can enter an address and see where it lies. Hazard areas largely follow the county’s river valleys — the Stillaguamish, the Snohomish and the Skykomish — as well as some coastal areas, like Mukilteo.

The county uses the maps in permitting — where to allow more development — and in emergency management planning, said County Executive Dave Somers.

The county expanded the definition of “landslide hazard area” to include much larger areas both above and below steep hillsides, although still not nearly large enough that the neighborhood beneath the Oso landslide would have been covered.

For people already living in hazard areas, “We go out and tell communities about them and share that information,” Somers said.

They tell people how to spot early earth movement, the warning signs of a possible slide: cracks in a hillside; bent trees, fences or utility poles; difficulty opening doors or windows.

The county now monitors moisture in the ground. After heavy rain or during high winds, the county sends alerts to residents living in hazard zones to look for landslide activity.

“There are specific conditions where slopes are more likely to fail, so we much more proactively reach out,” said Jason Biermann, a senior policy adviser to Somers who worked in emergency management during the Oso disaster and was on the scene for a month afterward.

Generally, Biermann said, people already know they’re living in a landslide zone.

“There’s a little bit of empowerment of, if we say, ‘Hey, if you start noticing these things here’s the number to call,’” he said.

Miller, the geomorphologist, said he’d like to see the maps incorporate landslide probability and size, how far runouts could go and the potential consequences.

King County has a similar online map, which provides more information, but it is both harder to find and more difficult to comprehend.

It also comes with a lengthy legal disclaimer, emphasizing it is not mapping landslide hazards, but “potential” landslide hazards. The disclaimer denies all liability for “errors, omissions, or inaccuracies” and for any harm resulting from the map, including to potential property values.

It is indicative of one of the reasons counties are reluctant to make hazard maps. It’s not just lack of resources. Counties fear liability from making the maps, and residents often resent the maps, which can cause property values to nosedive.

One of the difficulties in moving away from a landslide hazard zone is that as soon as your home is designated as being in such a zone, it becomes more difficult to sell.

Juneau, Alaska, updated its landslide and avalanche hazard maps last year after a contentious process dominated by fears the maps would lead to plunging home values and difficulty finding insurance.

Ultimately, the maps were approved, but included a disclaimer that says “The City and Borough of Juneau did not adopt these landslide hazard assessment maps.”

“Agencies are famously loathe to commit to paper on landslide runout boundaries, because the one thing you can be sure of is whatever happens isn’t going to match your prediction,” Montgomery, the UW geomorphologist, said. “You make your best call and then reality comes in and does what it does.”