Vegetation Interpretation And Biomass Estimation Based On 3S Technology In Luoyuan Bay

This article was conducted by vegetation of Fujian Province Luoyuan bay coastwise area as research subjects. Adopts the traditional line investigation, sample survey methods and 3S technology, combined with ALOS satellite remote sensing image of Japan as data sources, discussesed the Luoyuan bay vegetational diversity and the ground biomass characteristics by ERDAS IMAGINE9.2,ArcView and ArcGIS9.0. Whereas this study can provides basic research data for port construction, protecting biodiversity and the sustainable development of ecological environment in Luoyuan bay.The results showed that there were 584 species of vascular plants which belong to 127 families and 381 genera. The vegetation of Luoyuan bay can be divided into four vegetation type, including evergreen coniferous forest, evergreen broad-leaved forest, subtropical shrubby grassland and grassy salt marsh. According to the differences of community constructive species, the community is consisted of pinus massoniana community, Acacia confusa. community, miscanthus floridulus shrubby grassland, grassy salt marsh and so on. The biological diversity of Luoyuan bay is relatively low and the ecological stability is relatively weak.furthermore, the spartina alterniflora have the trend of further spread.Conducted remote sensing interpretation on Luoyuan bay vegetation, the results showed that among the three classification methods, supervised classification and visual interpretation are more closer, which is keeping with the results we surveyed on the spot. The statistics precision of kappa coefficient is relatively high.We analysed biomass and each kind of remote sensing factor by relevant analysis, t-test and regression analysis, getting the equation of linear regression.we estimated the ground biomass of Luoyuan bay vegetation, getting the ground biomass by remote sensing. The result shows that both remote sensing for estimating and the vegetation with measured are similar, the error is around 8.5%, so vegetation biomass estimation by remote sensing is a simple and effective method.