Authors: Lien Pham*, Virginia Institute of Marine Science, Tuan Quoc Vo, Can Tho University, Department of Land Resources, 3/2 Street, Ninh Kieu Dist, Can Tho, Viet Nam, Thanh Duc Dang, Singapore University of Technology and Design, Engineering Systems and Design Pillar, 8 Somapah Road, Tampines, Singapore
Topics: Remote Sensing, Coastal and Marine, Physical Geography
Keywords: Support Vector Machine, Object-based image analysis, Mangrove associations,SPOT imagery, Human activities
Session Type: Poster
Start / End Time: 1:10 PM / 2:50 PM
Room: Lincoln 2, Marriott, Exhibition Level
Presentation File: Download
Mangrove forests are well-known for their provision of ecosystem services and capacity to reduce carbon dioxide concentrations in the atmosphere, but they are among the top threatened habitats worldwide. Effective management of mangrove forests requires accurate quantitative and spatial information of these forests. This study aims to investigate the changes of mangrove forest associations in tropical regions and possible reasons of changes via the case study of the Can Gio biosphere reserve in Viet Nam by using remote sensing datasets (SPOT imagery), object-based image analysis and the Support Vector Machine classifier. The classification results showed that the highest overall accuracy of this combination was 81.6%, confirming the applicability of using remote sensing on assessing the development of mangrove forests. We found that there was a decrease of the Avicennia alba – Sonneratia alba area by 20.1% caused by the development of aquaculture and other human activities, but there was an increase of the Rhizophora apiculata area by 34.8%. These trends may increase risks of soil erosion and lead to the region more vulnerable to climate change and tropical storm events.