Spectral Remote Sensing and Geospatial Analysis of Mangrove Dynamics: A Case Study of the Vietnamese Southern Coastline

The thesis addressed four primary objectives: (i) critically reviewing spectral indices used in mangrove remote sensing, (ii) mapping mangrove distribution across the Vietnamese Southern Coastline (VSC), (iii) examining associations between mangrove dynamics and land use land cover change, and (iv) characterising spatiotemporal patterns of mangrove change in relation to managed and protected coastal environments. Multi-sensor satellite data from Landsat (1988–2023) and Sentinel-2 (2016–2023) were processed using Google Earth Engine to generate outputs at 10–30 m spatial resolution. The analytical approach combined multi-spectral indices, deep learning using a U-Net model, machine-learning classification, and spatial and trend-based regression techniques to support analysis across differing spatial and temporal scales. The thesis is organised into six chapters, comprising an introduction, a comprehensive literature review published in a peer-reviewed journal, three research chapters based on peer-reviewed publications, and a concluding synthesis.