Long-term mangrove dynamics and safety in coastal areas: An emergency Remote Sensing-based analysis using Landsat and Sentinel-2 Data

Mangroves are essential to coastal ecosystem health, providing natural defenses against erosion and serving as key habitats. This study investigates long-term mangrove cover and fragmentation trends in Pakistan's Indus Delta and Sandspit regions using Landsat imagery (1988-2024) and Sentinel-2 datasets (2019, 2024). Classification results show significant mangrove expansion, with Sandspit increasing from 10.22 km² in 1988 to 15.64 km² in 2024, and the Indus Delta expanding from 305.54 km² to 1002.82 km² over the same period. Sentinel-2 data showed slightly higher cover estimates, with values of 15.24 km² and 15.98 km² at Sandspit and 895.18 km² and 1009.76 km² for the Indus Delta in 2019 and 2024, respectively. Canopy density analysis from 1988 to 2024 categorized mangroves based on Normalized difference vegetation index (NDVI) values into sparse, moderate, and thick-density classes. NDVI thresholds (sparse: 0.1-0.3; moderate: 0.3-0.5; thick: >0.5) were used to assess canopy density. Results indicated densification of mangrove forests and a shift from sparse to dense vegetation, reflecting ecological recovery. Classification accuracy exceeded 96 % with high kappa values (>0.93). Due to field access limitations, accuracy assessments were conducted using visual validation via high-resolution Google Earth imagery. Fragmentation and transition analyses further revealed dynamic patterns and resilience of these ecosystems under environmental stressors. Comparative analysis of Landsat and Sentinel-2 data revealed high classification accuracy, with overall accuracies exceeding 96 % and kappa values above 0.93. The findings also provide actionable insights for national and provincial forest departments, supporting policy recommendations aimed at enhancing mitigation and adaptation strategies in response to landform changes and climate variability.