Spatial Temporal Analysis of Mangrove Cover Changes in Vanga Bay, Kenya (1985-2026) using Google Earth Engine

This study analyzes four decades of mangrove cover changes in the Vanga Bay region, Kenya, using Google Earth Engine (GEE), Landsat and Sentinel satellite imagery and machine learning techniques. Spatial analysis reveals an approximate 12.2% decline in mangrove cover dropping from 3,685 ha to 3,234 ha between 1991 and 2016, primarily driven by the over-exploitation of wood products for energy and building, transboundary conversion pressures and climate change impacts. This localized mangrove loss not only degraded critical fish nursery habitats but also threatened significant carbon pools, given that Vanga's dense Rhizophora mucronata stands hold average carbon stocks of 469 Mg C/ha. Since 2019, community-led initiatives such as the Vanga Blue Forest project have actively protected 460 hectares of mangrove forest, resulting in targeted recovery and emissions avoidance of over 106,929 tCO₂-eq over its crediting timeline. Natural regeneration is most prominent in low-energy estuarine areas characterized by high sediment deposition that outpaces current sea-level rise, highlighting the complex interplay between natural geomorphological processes and community forest management. Ultimately, these results demonstrate the power of integrating GEE and machine learning for accurate, long-term monitoring of transboundary coastal ecosystems. Such tools provide the precise data needed to scale up community-led restoration strategies and safeguard coastal resilience against accelerating climate and anthropogenic pressures