It is easy to view the ground as stable, as fixed, as immovable, even when deep down we know that it’s not. Sometimes the earth seems to shudder, as with an earthquake, and sometimes it pops, as with a volcanic eruption. Other times the earth slips, bits of dirt, handfuls of pebbles, beads of water combining and shifting until they coalesce into a cascade that blocks roads, shears homes from their foundations, and claims precious lives.
This happened in Ecuador earlier this year, when heavy rains triggered a hillside collapse in Quito, killing at least 24 people. It happened in Montecito, California, in 2018, when a type of landslide called a debris flow killed 23 people. And it happened in the Indian state of Uttarakhand in 2013, when roughly 13 inches of rain caused a slope along the eastern snout of a nearby glacier to fail. That landslide, along with the floods that helped trigger it, killed an estimated 6,000 people.
Landslides happen for many reasons, set off by earthquakes, volcanic eruptions, or human behavior. But “probably the most common driver we see for landslides worldwide is rainfall,” Ben Leshchinsky, an associate professor in the College of Forestry at Oregon State University, told me. “Say you have lots of rain. What that effectively does is it reduces the strength of the soil. When that soil strength decreases, it can reach a point where it fails, and naturally just slides away.”
And climate change is creating more extreme rain events. The 13 inches of rain that triggered the landslide in Uttarakhand was a more than 400 percent increase over the daily norm of 2.5 inches. Rain is why landslide researchers are warning that climate change may make landslides more likely, and that we are not prepared for this growing risk.
In High Mountain Asia, a landslide-prone region that includes Uttarakhand, climate-related shifts in rainfall will increase landslide risk by as much as 50 percent in certain areas, a 2020 study in the journal Geophysical Research Letters found. “These places that were wet and would get the precipitation are now going to get more of it,” Sarah Kapnick, a co-author of the study, said. (Now a senior climate scientist with J.P. Morgan, Kapnick was a research physical scientist at the National Oceanic and Atmospheric Administration when the study was published.) More rain, on its own, might increase landslide risk, but that risk is amplified by the timing of the precipitation—much of which is happening in the summer, when it’s falling as rain, as opposed to earlier in the year, when it would fall as snow. These patterns set the stage not only for more landslides, but for cascading catastrophes.
You can “get a rainfall event that triggers a landslide that blocks a lake that causes an outburst flood,” Dalia Kirschbaum, a landslide researcher with NASA and a co-author of the study, told me. An outburst flood is a kind of megaflood, in which water previously held back by either a glacier or glacial deposits of rock and sediment is released. In High Mountain Asia, the large numbers of glaciers and glacial lakes, which form from retreating glaciers, amplify risk, but, according to Leshchinsky, landslides are “an issue anywhere there’s basically any kind of relief or pretty steep slopes.”
Landslides occur on all seven continents, and in the U.S., they happen in all 50 states. In 2019, the Hooskanaden Landslide in Oregon wiped out a portion of the state’s Highway 101. In 2018, spring was unusually wet across much of the country, so in “places like Pittsburgh, the Appalachians in West Virginia, and in Virginia, and North Carolina, there were a lot of landslides,” Jonathan Godt, the coordinator of the landslide-hazards program for the United States Geological Survey, told me.
Climate change is heightening other landslide risks too. Longer, more powerful wildfire seasons and rising sea levels both can lead to more landslides.
Take the 2018 Montecito landslide. Officially, it began early on January 9 when roughly half an inch of rain fell in less than 30 minutes, but the roots of the catastrophe were seeded earlier. In December 2017, the massive Thomas Fire, which at its peak was powerful enough to generate its own weather, burned through trees and other vegetation that could have held the soil back. It likely changed the structure of the soil as well.
“When the fires burn super, super hot, oils and other chemicals in your vegetation leave almost a kind of plastic-wrap layer on the soil,” Cara Farr, the national coordinator for the U.S. Forest Service Burned Area Emergency Response program, told me. This phenomenon, called hydrophobicity, keeps the soil from absorbing rain not just at the surface, but deeper.
This doesn’t typically pose a problem if the first rainfalls after a wildfire are light: Those drops break up that plasticlike layer, allowing the rain to penetrate the earth. But “if you get a heavy rainstorm after a fire, that’ll drive an increased flooding,” Farr said—and an increased risk of the type of landslide loosed in Montecito. Heavy rains can essentially separate the upper layers of the soil from the hydrophobic layer below.
These debris-flow slides are fast-moving, Fausto Guzzetti, who heads the Office of Technical and Scientific Activities for Risk Prediction and Prevention at the Italian National Department of Civil Protection, told me. They start off with as little as a few cubic feet of dirt and other debris. As that mixture of dirt and debris starts to move downhill, it transforms into thick waves of rocks, boulders, and still more debris, traveling up to 35 miles an hour and covering distances of more than 50 miles in some cases.
“You cannot outrun them,” Guzzetti said. Debris flows made up of coarser materials—big rocks and boulders—turn into wrecking balls that can knock cars out of parking spaces and homes off foundations. Debris flows made up of finer material are just as dangerous: They go everywhere, “into buildings and vehicles and drowning everything in there, including people,” Guzzetti said.
After a wildfire, “sites are susceptible to debris flows [for] anywhere from three to five years,” Drew Coe, the watershed-protection-program manager at the California Department of Forestry and Fire Protection, told me. “If [in] your first year after the wildfire you don’t get enough rainfall, or if you don’t get an intense enough rainfall, you may not get the debris flow until the second year.” Drought years also mean that the plants that could help anchor the soils don’t grow. And any new fire resets the clock.
Places that abut the sea, as Oregon does, are susceptible to yet another risk. Oregon in particular has a lot of landslides—more than any other state—partly because, geologically speaking, it’s a young state, with young rocks that aren’t as cemented or as strong as older rocks. “And what that leads to is more landslides,” Bill Burns, an engineering geologist with the Oregon Department of Geology and Mineral Industries, told me. Oregon also has a landslide “toe” problem.
The term landslide actually has two meanings. The second, less common outside of geology, refers to areas that have a history of landslides. Those past landslides reshape the slope, making it more susceptible to landslides in the future. One feature, called the toe, which marks the end point of a landslide’s moving material, has researchers concerned. Over time, the toe can act as a sort of cork, helping stabilize the rest of the slope. But in places such as the Arizona Landslide in coastal southern Oregon, the toe is exposed to the ocean. And the fear is that, as sea-level rise increases, these exposed toes “become increasingly susceptible to increased total erosion through combination of sea level and storms,” Jonathan Allan, a coastal geomorphologist at the Oregon Department of Geology and Mineral Industries, told me. Rising seas might grind the toe down, making it more likely to fail, and reduce the area’s stability, making future landslides more likely.
Rain, wildfire, and sea levels can on their own increase the risk of landslides, but all of these factors can compound. A wildfire could break out on an ocean-facing slide in a region that is also experiencing increased precipitation. And landslides simply aren’t as well understood as other natural phenomena, which is why researchers are now scrambling to better study them.
“We did not have until very recently ways of mapping landslides,” said Guzzetti, noting that the lack of landslide data stands in marked contrast to the global network of seismometers for monitoring earthquakes and floods. Landslides can be relatively small, too, so absent a fatality or property damage, many go unremarked. This is why, in addition to investing in satellite imaging, NASA, for example, has launched Landslide Reporter, which is designed to crowdsource landslide data. If you see a landslide, say something.
One more factor, divorced from climate, is making these dangers all the more pressing to address: “where people are building,” Leshchinsky said. He noted that the greatest risk tends to be in developing countries where standards and enforcement on buildings in landslide zones may be sparse. But “the fact that people are starting to push further and further into what they call the wildland–urban interface puts more people possibly—I don’t want to fearmonger—into these areas that may or may not be stable,” places where the ground might literally slip from beneath their feet.