Land Cover Mapping in Cable Bay with Sentinel 2 from 2018 to 2024

This study mapped and analyzed land cover changes across six classes—Broadleaf, Conifer, Cropland, Developed, Meadow, and Water—from 2018 to 2024. We applied a Random Forest classifier to high-resolution Sentinel-2 imagery within a hybrid Python-R workflow, achieving classification accuracies of 97.0-98.6%.

Spatiotemporal analyses, including transition matrices and hotspot detection, revealed a significant decline in Broadleaf forests and a net increase in Developed land. These shifts reflect urbanization’s pressure on natural habitats, threatening biodiversity and ecosystem services. The findings underscore the urgency of conservation efforts and informed land-use planning to preserve Cable Bay’s ecological integrity.