Integrated machine learning segmentation and 3D change detection for a scalable coastal cliff monitoring workflow
Published in Computers & Geosciences, 2026
A workflow combining machine learning classification and 3D change detection to extract detailed coastal change measurements from LiDAR point cloud data. The method can detect changes on complex surfaces with overhanging topography, validated using 410 sequential LiDAR surveys along a 2.3 km coastal stretch in Del Mar, California. Includes a new ML application for cliff toe identification and beach/cliff separation, scalable to regional monitoring efforts.
Recommended citation: Mack, C.J., Maclay, M., Krier-Mariani, R., & Young, A.P. (2026). "Integrated machine learning segmentation and 3D change detection for a scalable coastal cliff monitoring workflow." Computers & Geosciences, 106165.
Download Paper