Application of geomorphons and TPI in the geomorphometric study of GCL landforms in the Drawsko Lakeland (NW Poland)

This study compares two automated geomorphometric methods – geomorphons and the Topographic Position Index (TPI) – for detecting glacial curvilineations (GCLs) within the Komorze tunnel valley, NW Poland. Using LiDAR-derived digital elevation model (DEM), outputs from each method were evaluated against a manually refined reference. Geomorphon-based classifications, particularly at a flatness threshold (FT) of 3, most closely matched the manual interpretation, preserving ridge structure and spatial coherence. TPI methods often over-fragment the ridges at smaller neighbourhood sizes. These results suggest that while manual delineation remains most reliable, automated tools such as geomorphons can meaningfully support large-scale GCL mapping with careful parameterisation.