Federal grants help states build landslide inventories with lidar and machine learning

USGS awarded a new round of grants to support landslide inventories, hazard mapping and risk reduction. Pennsylvania and Nevada are developing their first landslide databases or statewide inventories, with plans to use geospatial data and, in Pennsylvania, machine learning to extend mapping capacity.

More complete inventories and susceptibility maps can help emergency planners and land-use officials identify exposed areas and prioritize further investigation. Consistent post-event data collection can also help researchers understand which slopes fail and how far displaced material travels.

New federal funding is helping geological surveys turn terrain measurements and scattered landslide records into hazard-mapping resources. Pennsylvania will use its grant for a two-year initiative to map landslides and create a susceptibility map for Washington County, supporting the state's first landslide database. It also plans to adapt machine-learning models developed with the Kentucky Survey to identify potential landslide features in lidar data. Nevada plans its first statewide landslide inventory, combining existing records with wildfire data and other information, alongside post-event collection protocols and a public dashboard. Researchers can compare drone imagery collected after slides with earlier lidar measurements, while satellite imagery can help establish timing and spatial extent. These efforts strengthen the evidence available for emergency planning and risk reduction. They remain inventory-building and mapping projects: the article does not establish operational forecasting capabilities or measured improvements in public safety.

Evidence supplied by StateScoop.