This study analyzes changes in forest cover in the Acacia forest of Khartoum (2021–2026) using GIS, remote sensing, Sentinel-2 imagery, and an artificial intelligence-based Random Forest classification model. Random Forest is often an appropriate baseline for tabular data and for traditional pixel-based change detection (e.g., comparing multi-temporal bands, NDVI, or SAVI), and it is used in this study.

Spectral indices (NDVI and SAVI) derived from bands B2, B3, B4, and B8 were used as model inputs. The classifier was trained on 2023 baseline data, using NDVI thresholding to distinguish forest from non-forest areas. The results indicate substantial forest degradation over the five-year monitoring period, primarily caused by unauthorized logging. According to the NDVI values used to assess deforestation in the forest, the NDVI values for Year 2021 (0.4151), Year 2022 (0.5480), Year 2023 (0.5962), Year 2024 (0.5864), Year 2025 (0.56226), and Year 2026 (0.35181) indicate deforestation across the study years.