A Spectral Mixture Analysis and Landscape Metrics Based Framework for Monitoring Spatiotemporal Forest Cover Changes: A Case Study in Mato Grosso, Brazil

An increasing amount of Brazilian rainforest is being lost or degraded for various reasons, both anthropogenic and natural, leading to a loss biodiversity further global consequences. Especially in the state Mato Grosso, soy production large-scale cattle farms led extensive losses recent years. We used spectral mixture approach followed by decision tree classification based on more than 30 years Landsat data quantify these losses. Research has shown that current methods assessing forest degradation are lacking accuracy. Therefore, we generated classifications determine land cover changes each year, focusing cleared land. The analyses showed decrease area Grosso 28.8% between 1986 2020. In order measure changed structures selected period, fragmentation diverse landscape metrics were carried out municipality Colniza Grosso. It was found areas experienced also high degree over study with an increase 83.3% number patches mean patch 86.1% time resulting altered habitats flora fauna.