REMOTE SENSING APPLIED TO RECOGNITION OF SEAGRASS MEADOWS IN LOW VISIBILITIES ENVIRONMENTS: LA GUAJIRA, COLOMBIA

Abstract

Seagrass meadows are important ecosystems due to their high productivity and ecological value among tropical ecosystems, because of their high species diversity. In Colombia seagrasses are located around some islands, oceanic coral banks and along the Caribbean shelf, mainly in La Guajira Department, where more than 80% of the seagrass meadows of the country are present. In the world, the delimitation of this ecosystem has been successfully mapped during years, with assistance of remote sensing, using satellite image of different spatial scales. Nevertheless, the specific environmental conditions in La Guajira, such as high water turbidity and reduced light penetration restrict the use of traditional satellite images employed for those seascapes. With the aim of delimiting and establishing the extension of seagrass meadows in La Guajira, based on analyses between July 2013 and February 2014, a methodology of massive image interpretation that included fieldwork fast verification was applied, generating as a result one layer of seagrass habitats in Cabo de La Vela – Dibulla area at 1:100000 scale. Methodology included geometric correction, image fusion, fieldwork information, definition of thematic classes, determining of criteria for spatial delimitation, visual interpretation of images, thematic uncertainty qualification, and final cartography production. The process of cartographic production showed that Landsat 8 OLI satellite sensor images made easier the identification of seagrass meadows in deep areas (>10m). In total, 53621 ha of seagrass meadows were identified, and the largest meadows of Colombia were delimitated, which reach dimensions of up to 6018 ha.