معرفی
Stefan Leyk is a Professor of Geography at the University of Colorado Boulder within the Department of Geography in the College of Arts and Sciences. His research focuses on GIScience, spatial uncertainty modeling, and historical landscape analysis, with significant contributions to cartographic pattern recognition from historical maps and spatial dynamic modeling in public health. He holds a Ph.D. from the University of Zurich and the Federal Research Institute for Forest, Snow and Landscape (2005).
His primary research interests include uncertainty in GIScience and spatial uncertainty modeling, land cover change modeling using historical spatial information, cartographic pattern recognition from historical maps, and spatial dynamic modeling approaches in public health. His work bridges historical geography with advanced computational methods, particularly in extracting settlement patterns from historical map archives and developing spatiotemporal datasets spanning centuries.
Leyk's recent publications demonstrate strong trends in historical settlement analysis, with major projects like CHRONEX-US and HISDAC-US creating century-long datasets of urban infrastructure and settlement evolution. His work increasingly integrates machine learning with historical map processing, focusing on uncertainty quantification, built-up land validation, and environmental justice applications related to flood risk and coastal hazards. Key thematic areas include long-term urban growth patterns, rural poverty dynamics, and wildfire risk assessment at the wildland-urban interface.
Leyk has received significant research funding through collaborative grants including 'HNDS-I: Building Long-term, National-scale Spatiotemporal Data Collections from Historical Map Archives' (2025) and 'HNDS-I: Data Infrastructure for Research on Historical Settlement and Population Growth in the United States' (2021). He actively mentors graduate students including Alek Berg, Caitlin McShane, and Yuying Ren, and teaches advanced GIS courses such as GEOG 4303/5303 GIS: Spatial Programming and GEOG 4103/5303 GIS: Spatial Analytics.
His laboratory work centers on geospatial modeling of historical settlement and landscape analysis, with a focus on developing automated methods for processing historical map archives and creating linked spatiotemporal data. Current projects involve machine learning applications for feature extraction from historical maps, uncertainty prediction in built-up land layers, and the development of fine-grained datasets measuring 200 years of land development in the United States.
