| Wetlands provide a range of environmental and socio-economic benefits, which range from their ability to store floodwater and improve water quality to providing habitats for wildlife to support biodiversity and aesthetic values. The loss of wetlands, which caused by the effects of urbanization, sprawl, land cover changes has gained considerable attention now. The utilization of satellite remote sensing and GIS technology for wetland information extracting and dynamic monitoring has proven to be a useful application. The objective of this study was to find more efficient way to extract wetland information from remote sensing data.Taking Minjiang river estuary as study area, Wetlands information was extracted from two different kinds of RS data source using different methods supported by the RS and GIS technology. Then wetland landscape ecological theory was applied to analyze the extracted results.As for SPOT5 image acquired at 2003.12.04, high spatial resolution fusion data was generated using SFIM method. The present situation of wetland information was extracted from this fused image, and landscape pattern analysis results shows the diversity of the whole landscape is relative high and there exists dominant landscape pattern type.Decision tree models were designed and carried out for extracting wetland dynamic change information from TM/ETM+ image, which acquired at 1988.04.09and 2001.03.04 respectively. In order to analyze the dynamic change of landscape pattern in study area, a database which integrated spatial, property and change process information was constructed based on the information extracted above. The analysis results show that during the focused period the diversity index, evenness index has decreased while the dominance index, the fragmentation index, and fractal dimension index has increased, which indicating the decreasing of spatial heterogeneity in study area besides the landscape stability was also decreased.
|
PDF Full Text Request