Monitoring vegetation edges in dynamic coastal zones is essential for understanding long-term shoreline change and supporting effective coastal management, particularly as climate change accelerates erosion, sea-level rise, and ecosystem shifts. This study provides the first validation of VedgeSat, an automated Vegetation Edge (VE) detection toolkit, in contrasting tropical coastal environments, with relevance for coastal monitoring worldwide. In Sumatra, Indonesia, fifteen sites were assessed, encompassing diverse vegetation and sediment types, a range of water clarity, and varying wave exposures in both open and sheltered coastal settings. Vegetation edge detection was conducted with high-resolution PlanetScope imagery and differential Global Positioning System (dGPS) field surveys. VedgeSat performed reliably in areas with dense vegetation, regardless of the type of vegetation or sediment, and water clarity achieving sub-pixel root mean square errors (RMSE) of less than 7 m, R 2 values up to 0.89 and minimal positional bias. Performance declined in areas with sparse and patchy vegetation, such as pioneer mangroves and grasses in sandy environments, resulting in higher RMSE and reduced R 2 values. A sensitivity analysis demonstrated that tuning the threshold of Normalized Difference Vegetation Index (NDVI) values can optimize edge detection across diverse vegetation types and environments. Overall, the results confirm the robustness of VedgeSat for scalable monitoring of vegetated coasts without retraining, while also identifying limitations in sparsely vegetated settings. This study provides the first benchmark for automated vegetation edge detection in tropical systems and demonstrates the potential of satellite-based approaches to enable large-scale, repeatable, and cost-effective coastal monitoring in data-scarce regions.
Nugraha et al. (2026) studied this question.