Mapping gullies, dunes, lava fields, and landslides via surface roughness

摘要: Abstract Gully erosion is a widespread and significant process involved in soil land degradation. Mapping gullies helps to quantify past, anticipate future, losses. Digital terrain models offer promising data for automatically detecting mapping especially vegetated areas, although methods vary widely measures of local roughness are the most varied debated among these methods. Rarely do studies test performance metrics gullies, limiting their applicability small training areas. To this end, we systematically explored how derived from high-resolution Light Detection And Ranging (LiDAR) can aid unsupervised detection over large area. We also tested expanding method other landforms diagnostic similarly abrupt land-surface changes, including lava fields, dunes, landslides, as well investigating influence different thresholds, resolutions kernels, input resolution, comparing our with previously published algorithms. Our results show that total curvature suitable metric recognising analysed fields LiDAR data, comparable success more sophisticated metrics. Tested dunes or landslides remain difficult distinguish surrounding landscape, partly because they not easily defined terms topographic signature.