Carbon stock assessment of Mangrove habitat using field measurements and using machine learning-based GIS and remote sensing techniques

Summary: Mangrove ecosystems are critical blue-carbon habitats that contribute to climate change mitigation through carbon sequestration. This study assessed aboveground biomass (AGB) and carbon stocks of Avicennia marina mangroves in a hyper-arid environment using integrated field and remote sensing approaches. Destructive sampling and field surveys were combined with high-resolution Pléiades Neo satellite imagery to calibrate biomass models. A newly developed allometric model achieved high accuracy (R2 = 0.93, MAPE = 42.9%), estimating carbon storage at 0.023 t C per tree for diameters of 4.5–13.0 cm. A machine learning-based remote sensing model also performed reliably (R2 = 0.643, MAPE = 17.3%), demonstrating the utility of spectral indices and canopy reflectance for biomass prediction. These results highlight the role of mangroves in national and global carbon accounting and provide practical insights for conservation planning, carbon offset frameworks, and standardized biomass assessment methodologies in arid coastal ecosystems.