Embracing imperfection: how an unvalidated remote-sensing based land degradation workflow helped Caribbean experts in disaster management

Caribbean islands face significant soil erosion and landslide challenges driven by extreme events such as tropical storms and hurricanes. These processes have substantial impacts on agriculture, ecosystems, infrastructure, and local communities. Despite the availability of Earth Observation (EO) and geospatial datasets, their potential for land degradation monitoring and management remains underutilized due to the lack of accessible methodologies and technical barriers to implementation.
To address this gap, we developed a multi-scale geospatial framework integrating open-access satellite data and national LiDAR datasets for monitoring surface processes and soil erosion in Saint Lucia and Dominica. The methodology aims to build capacity among local officers, equipping them to identify erosion hotspots and implement mitigation measures effectively, with potential for replication in similar contexts. At the island scale, global Digital Elevation Models (DEMs) were employed to compute hydrological and geomorphological indicators, delineating areas at risk of erosion and sediment transport. At the local scale, high-resolution LiDAR data facilitated detailed analyses, including: (1) identifying erosion-prone areas and sediment transport pathways; (2) assessing road infrastructure to detect drainage inefficiencies and processes contributing to slope destabilization; (3) analyzing runoff dynamics from upslope regions to coastal zones.
Our framework does not aim to improve model accuracy for purely research purposes. Instead, it adopts a simplified, yet science-based modeling approach tailored to the unique challenges of small-island settings. By focusing on actionable insights, it provides local experts with practical, science-based strategies to address land degradation issues and contributes to building resilience within local communities.