Hydro-Functional Catchment Typology via K-Means Clustering on the Hungarian Great Plain -- Analysis Scripts

Analysis scripts for developing a hydro-functional typology of 2,910 subcatchments in the Hajdusagi (HAHO) region of the Hungarian Great Plain using K-means clustering on 12 morphometric, land cover, soil, and infrastructure descriptors. Optimal k=6 was selected based on silhouette analysis (0.287). Stability validated through bootstrap resampling (ARI > 0.95) and multi-seed analysis. Includes sensitivity analyses (feature removal, GMM comparison, land-cover-only clustering, Moran's I spatial autocorrelation), dominant cluster sub-analysis, k=6 vs k=8 justification, and external validation against MODIS NDVI/NDMI vegetation indices. Catchment delineation was performed in ArcGIS Pro 3.6 using the ArcHydro toolset.