Downscaling Switzerland Land Use/Land Cover Data Using Nearest Neighbors and an Expert System

High spatial and thematic resolution of Land Use/Cover (LU/LC) maps are central for accurate watershed analyses, improved species, habitat distribution modeling as well ecosystem services assessment, robust assessments LU/LC changes, calculation indices. Downscaled Switzerland were obtained three time periods by blending two inputs: the Swiss topographic base map at a 1:25,000 scale national statistics from aerial photointerpretation on 100 m regular lattice points. The resulting was factor 16 to reach 25 m, while increased 29 (in map) 62 land use categories. method combines simple inverse distance weighting 36 nearest neighbors’ information an expert system correspondence between input categories possible output types. developed algorithm, written in Python, reads writes gridded layers more than 64 million pixels. Given size analyzed area, High-Performance Computing (HPC) cluster used parallelize data analysis obtain results efficiently. presented this study is generalizable approach that can be downscale different types geographic information.