Multi-scale and multi-criteria mapping of mountain peaks as fuzzy entities

Mountain peaks are mapped as multi-scale entities with modifiable boundaries and variable contents. Four semantic meanings are imported and quantified to first characterize peaks at a range of spatial scales and then evaluate the multi-criteria 'peakness' at each scale. Peakness is defined as the prototypicality of identified summits and as the similarity of each point (cell) to summits. The procedure then summarizes the individual-scale peakness across considered spatial scales into a univariate membership surface. This allows mapping of vague peak entities as non-homogeneous peak regions whose boundaries depend on user-specified peakness thresholds. This procedure was applied in a case study to tackle several challenges in landform delineation, including boundary, spatial continuity, spatial scale, topographic context, and multi-criteria definition.