Advertisement

NASA landslide map estimates risk in real time

"The model has been able to help us understand immediate potential landslide hazards in a matter of minutes," said landslide expert Thomas Stanley.

By Brooks Hays
The new model updates the risk of landslides around the world every 30 minutes. Photo by NASA's Goddard Space Flight Center / Scientific Visualization Studio
The new model updates the risk of landslides around the world every 30 minutes. Photo by NASA's Goddard Space Flight Center / Scientific Visualization Studio

March 22 (UPI) -- NASA can now assess the risk of a landslide in real time at any location on the planet.

The new model, designed by scientists at NASA's Goddard Space Flight Center, combines satellite imagery with predictive analytics to offer landslide risk assessments on the fly.

Advertisement

Machine learning allows the model to analyze historical landslide trends in different locations, while real-time precipitation data provided by a constellation of satellites helps the algorithms formulate up-to-date risk levels.

Researchers detailed their predictive model this week in the journal Earth's Future.

"Landslides can cause widespread destruction and fatalities, but we really don't have a complete sense of where and when landslides may be happening to inform disaster response and mitigation," Dalia Kirschbaum, a landslide expert at Goddard, said in a news release. "This model helps pinpoint the time, location and severity of potential landslide hazards in near real-time all over the globe. Nothing has been done like this before."

NASA and Japan Aerospace Exploration Agency satellites provide precipitation data. The data helps the model identify places experiencing especially large amounts of rain or prolonged periods of precipitation. The model then taps into historical trends to determine whether the location is prone to landslides.

Advertisement

In addition to historical trends, the model also analyzes local physical characteristics to assess risk, including the presence of roads, fault lines and removed trees, as well the area's topography and the nature of underlying bedrock. This combination of factors determines whether or not the model decides to elevate a specific location's risk.

If the model finds an area has a moderate to high risk of experiencing a landslide, it will produce a "nowcast." The model spits out new nowcasts every half-hour.

When tested against previous landslide data, the model successfully predicted elevated risks in areas affected by monsoons and hurricanes.

"The model has been able to help us understand immediate potential landslide hazards in a matter of minutes," said Thomas Stanley, landslide expert at Goddard. "It also can be used to retroactively look at how potential landslide activity varies on the global scale seasonally, annually or even on decadal scales in a way that hasn't been possible before."

Latest Headlines