SPATIOTEMPORAL ANALYSIS AND PREDICTION OF MANGROVE DYNAMICS FOR NATURE-BASED SOLUTIONS IN TANJUNG CARAT, SOUTH SUMATRA

Abstract

Mangrove ecosystems play a crucial role in coastal resilience, yet their long-term spatial and ecological dynamics remain insufficiently quantified in tropical delta environments. This study examines multi-class land-cover changes and mangrove canopy dynamics in Tanjung Carat, South Sumatra, Indonesia, from 2016 to 2024, using Landsat 8/9 imagery and NDVI-based classification. The results show that mangrove area declined from 96,630.95 ± 120 ha to 91,184.88 ± 140 ha, representing a 5.6 % (≈ 5,446 ha) net loss, primarily in Muara Sungsang and Sungsang I–III, where conversion to aquaculture, settlements, and agricultural fields was most intense. In contrast, Sungsang IV remained relatively resilient, with only a 1.4 % (≈ 1,171 ha) decline due to stable hydrology and limited accessibility. NDVI values (0.12–0.59) indicated canopy degradation in fragmented zones and dense cover in the core areas. The Artificial Neural Network (ANN) model achieved 98 % overall accuracy and a Kappa coefficient of 0.98049, validating the robustness of the predictive framework. The 2032 projection suggests an additional 4.6 % decline (≈ 4,195 ha; ≈ 0.58 % yr⁻¹)  under a business-as-usual scenario. These findings underscore the urgent need for Nature-based Solutions (NbS) integrating hydrological restoration, adaptive zoning, and community stewardship to balance coastal development and mangrove conservation in South Sumatra’s deltaic landscape.