Mangrove forest mapping using shortwave infrared reflectance and image subset techniques using proximity measure

Mangrove forests provide a habitat for many endangered sea water-tolerant plant species and coastal fauna. Despite their ecological importance and shoreline protection properties, mangroves are under serious threat, resulting in significant losses in mangrove forest cover. Their efficient management requires frequent mapping to detect degradation or regeneration. The methodology presented in this paper, identifies land areas with high moisture content based on their lower shortwave infrared (SWIR) backscatter that are then clustered in an area of interest based on proximity to the reference features typically associated with mangrove forests, such as coastlines and river edges. This method minimizes manual operation and use of paper maps, extensive fieldwork data, or heavy reliance on local knowledge. Validation of the methodology is provided based on a reliable inventory of mangrove forest in Iriomote Jima, Japan.