Remote sensing of forested wetlands: Application of multitemporal and multispectral satellite imagery to determine plant community composition and structure in southeastern USA

摘要: A hierarchical classification of forested wetland communities was developed for the lower Roanoke River floodplain northeastern North Carolina, USA, through use multitemporal and multispectral satellite digital data. Landsat Thematic Mapper (TM) images from different seasons (March–April, May–June, July–August) throughout a single year were used to exploit phenological variability forest generating landcover ecologically important vegetation types within floodplain. scheme that relied upon customized spectral `feature sets' TM bands their transformations generate classified each level community scheme. The objective enhance discrimination at subsequent levels inputs assembled time series in conjunction with detailed floristic information collected though in-situ methods. As such, general classes iteratively reclassified into more correspondingly `deeper' or nodes hierarchy. Vegetation included 21 several other study area. integration field data permitted spatially-explicit highly descriptive definitions occurring Additional validate compositional structural characteristics mapped plant described by