The December 2021 eruption of Mount Semeru caused massive ecological damage, initiating primary succession across landscapes buried by volcanic material. While satellite monitoring provides macro-scale trends, the micro-scale spatiotemporal dynamics and specific topographic drivers of early recovery remain poorly understood. This study quantifies the rate of vegetation recovery and statistically evaluates its micro-topographic constraints. We conducted a time-series analysis using ultra-high-resolution (2.7 cm/pixel) Unmanned Aerial Vehicle (UAV) RGB imagery acquired in 2022, 2023, and 2025 across a 127.08-hectare Area of Interest within Volcanic Hazard Zone III. Vegetation cover was extracted using the Excess Green (ExG) index with dynamic Otsu thresholding, validated by rigorous accuracy assessments (Overall Accuracy: 89%–95%). A 3-bit spatial coding logic tracked pixel-level successional trajectories, while topographic controls were evaluated using a non-parametric Mann-Whitney U test based on spatial random sampling. The results revealed a net vegetation recovery rate of +1.88 hectares/year, forming concentrated succession islands. Crucially, statistical analysis confirmed that micro-topographic slope acts as the primary abiotic control (p < 0.001); pioneer vegetation preferentially colonized steeper slopes (mean 43.20°) that offer structural refuge from active lahar flows, whereas flatter depositional beds (mean 23.64°) remained largely barren. Conversely, elevation (~798 masl) showed no significant correlation (p = 0.68), indicating that secondary fluvial disturbances override broad altitudinal gradients within this active corridor. This study demonstrates a validated, reproducible UAV framework for high-precision ecological monitoring, providing fundamental insights for targeted post-eruption restoration strategies.
Copyrights © 2026