Mangrove forest mapping through remote sensing imagery: study case for Buenaventura, Colombia

Abstract

Mangroves are plant communities of high ecological and economic importance for coastal regions. This investigation provides a methodology for mapping Mangrove forests through remote sensing images in a semidetail scale (1:25,000) in a sector of the municipality of Buenaventura, Colombia. A Sentinel 2 image and 2017 highresolution ortophotomosaic of the municipality were used for the mangrove cartography, using QGIS software, spectral analysis was performed and supervised classification was established using Maximum Likelihood algorithm. Results shown that mangrove is the most representative cover in the study area whit 7,264.21 ha in total extension (59.21% of total area), the development classification got a thematic accuracy of 80% and 0.70 in Kappa index. The used methodology can be used as an academic and research reference for mangrove semi-detail mapping in the world.

Keywords

• cartografía
• manglar
• sensores remotos
• firma espectral
• clasificación supervisada
• máxima verosimilitud