Examining impacts of the Belo Monte hydroelectric dam construction on land-cover changes using multitemporal Landsat imagery

Many hydroelectric dams in the Brazilian Amazon have been constructed, but how dam construction influences land-cover change has not been fully examined. For our research, we selected Belo Monte hydroelectric dam, the third-largest dam in the world, to explore its impacts on major land-cover change. Multitemporal Landsat images between 2006 and 2017 were used. The maximum likelihood classifier was used to classify these Landsat images into primary forest, secondary forest, agropasture, man-made bare land, natural bare land, and water. The land-cover change was examined using the post-classification comparison approach based on different stages of dam construction, and was further examined along the upstream and downstream river buffer. The results indicate that overall classification accuracies of 89.7% and 92.3% were obtained for the 2011 and 2015 land-cover classification results, respectively. Primary forest decreased continuously from 47.8% in 2006 to 35.3% in 2017. Different stages of dam construction had various impacts, that is, before dam construction, deforestation and agropasture expansion were the major land-cover change categories; during dam construction, the increased area of man-made bare lands, the canal construction zone, and the increased area of natural bare lands downstream were obvious, in addition to deforestation and agropasture dynamics; when dam construction was complete, water bodies increased considerably upstream and decreased downstream. These big changes in water bodies may have long-term impacts on ecosystem functions and environments. This research provides new insights on the impacts of dam construction on land-cover changes, which is valuable for making better decisions about water and land resources.