Temporal analysis of land cover and land use changes in a Brazilian Amazon urban watershed

Urban expansion has impacted watersheds that supply water to cities. This study analyzes land use and land cover changes from 1984-2023 in the Belém Metropolitan Region in Amazon Region, using Geographic Object-Based Image Analysis (GEOBIA) applied to Landsat imageries. Five land cover classes were mapped: lakes, aquatic macrophytes, urban areas, forest, and pasture. Results show a 13.4% increase in forest due to protected areas, a 3.7% reduction in lake surface area from macrophyte proliferation, and a 3.2% increase in urban expansion. Major transitions occurred in the boundaries of the watershed. The conversions of pasture to urban areas were most intense between 1984-1989, with 138.1 ha converted. The most significant land-use change in the study area was from pasture to forest, totaling 528 ha and to urban areas (274 ha) between 1984-2023. The overall accuracy of classification exceeded 88%. The study demonstrates GEOBIA’s effectiveness for long-term watershed monitoring using Landsat imagery.