Research

Predicting Landslides

Aerial image of the aftermath of the 2017 Mud Creek landslide in California.

Landslides can cause catastrophic damage, but historically it’s been difficult to monitor susceptible slopes and predict when they’ll fail. But a recent study looking at the 2017 Mud Creek landslide in California shows that new methods could provide a heads up.

The researchers used satellite data from the months preceding the landslide to study how areas on the slope moved relative to one another. Within their survey region, they found sub-regions where ground locations largely moved together. These sub-regions, called communities in the researchers’ parlance, were remarkably persistent, showing little variation over long periods. But 56 days before the landslide, the researchers saw a sudden change between the communities on the slope. They believe their methodology could help identify slopes in danger of imminent slides.

So far, though, they’ve only applied this method to the Mud Creek landslide. It’s a promising start, but they’ll need to show that the technique works for other slides as well. If so, it will be a major step forward in landslide prediction. (Image credit: USGS; research credit: V. Desai et al.; via APS Physics)

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.