Deforestation in the Special Nature Reserve Gornje Podunavlje: Insights from PlanetScope, Sentinel-2 and Landsat-8 remote sensing data

Abstract

Deforestation is a pressing environmental issue, and assessing forest loss with precision is crucial for effective conservation strategies. This study evaluates forest area loss in the Special Nature Reserve "Gornje Podunavlje" over a seven-year period from July 2017 to July 2024. Utilizing remote sensing data from PlanetScope, Sentinel-2, and Landsat-8 satellites, the study employs the XGBoost machine learning classifier to classify forest cover. Results indicate a consistent decline in forest area across all datasets, with reductions from approximately 43 km² in July 2017 to around 33 km² by July 2024. Despite variations in reported forest area due to differences in spatial resolution-PlanetScope (3 meters), Sentinel-2 (10 meters), and Landsat-8 (30 meters)-the overall trend of forest loss is evident. Landsat-8 consistently reported higher forest area compared to PlanetScope and Sentinel-2, attributed to its coarser resolution which may include more edge effects. The high-resolution PlanetScope data allowed for more precise delineation of forest boundaries, enhancing the accuracy of forest cover assessments.

Keywords

• drought
• sentinel - 1 sar
• sentinel - 2 msi
• random forest