MAPPING MANGROVE DENSITY FOR CONSERVATION OF THE RAMSAR SITE IN PENINSULAR MALAYSIA

• Published in: International journal of conservation science 2020-01, Vol.11 (1), p.153-164
• Main Authors: Razali, Sheriza Mohd, Nuruddin, Ahmad Ainuddin, Kamarudin, Norfaryanti
• Format: Article
• Language: English
• Subjects: Biomass; Classification; Conservation; Conventions; Density; Feature extraction; Forests; Geographic information systems; Image resolution; Mangroves; Mapping; National parks; Nature conservation; Normalized difference vegetative index; Plantations; Remote sensing; Sand; Satellite imagery; Satellite observation; Satellites; Sea level; Sea level rise; Shadows; Unmanned aerial vehicles; Vegetation; Vegetation index; Wetlands
• ISSN: 2067-533X; 2067-8223

It is widely agreed that rapid development has led to mangroves being in urgent need of improved monitoring and assessment techniques for better conservation. In Malaysia, the Convention on Wetlands, otherwise known as the Ramsar Convention, came into being specifically to address this problem and protect this particular area. The rapidly rising sea-level at mangrove sites is currently impacting the depth of mangrove soil, so action must be extended using available technology to sustain mangrove lives. This study tested if the vegetation indices from recent high-resolution multispectral satellite images Satellite Pour l 'Observation de la Terre, known as SPOT, can map mangrove density and predict health area for the sites. An Unmanned Aerial Vehicle (UAV) was also used to fly at the nearer sites for mangrove density classification mapping based on feature extraction tools classification. This study employed Normalized Difference Vegetation Index (NDVI), Green Normalized Difference Vegetation Index (GNDVI), Greenness Vegetation Index (GVI), and Ratio Vegetation Index (RVI) to map mangrove density for the study site. The study showed, in terms of Maximum Likelihood Classifier (MLC) measurement, that 52% of the area was "high density", 10%-19% was "low density", and "low density mixed sand" and also "sand" area. The results also showed that NDVĪ responded higher for "high density" with 0.56, GNDVĪ with 0.27 for "high density", GDVI with 17.0, and RVĪ only classified two density areas. In addition, the UAV images were classified into shadow, obvious tree crown, vegetation in the water, sea area, and others. As expected, the study revealed that UAV (b) was presented in a very high percentage in obvious tree crowns but in a low percentage with 2% of other classes, making it a distinct class from others. In addition, UAV (a) showed 42% shadow with a small portion distributed in other vegetation features (sea area and others) class, 4% to 10°%. Mangrove density can also be used as an indicator of mangrove health status because low density mangroves are always found near to the risk areas.