Analysis of optical-SAR data fusion and merged data set procedures in highly developed urbanized coastal ecosystems

摘要

Although optical remote sensing systems have proven to be of considerable use for much land cover mapping in coastal environments, some limitations have also become apparent. Investigations of the potential of imaging radar systems for land cover and coastal area investigations have been reported for some time, satellite imaging radar systems have generated considerable interest in their unique and complementary role in mapping coastal environments for some time, but image availability and the extent of image coverage has been limited. The focus lies not only on the contribution of SAR as a single sensor but as part of multisensor combinations and methods of improving accuracy and precision. The NOAA Coastal Change Analysis Program (C-CAP) has developed a protocol and classification system for mapping coastal wetlands and adjacent upland land cover for any coastal region of the United States based primarily on analysis of Landsat TM data. In an expanded C-CAP classification wetlands are grouped into five categories and the upland land cover into eight. Research to date has also proven that considerable optical spectral confusion exists among the categories. This study is part of a broader evaluation of the utility and role of SAR for coastal resource management data needs. The purpose of this investigation is to evaluate the potential of SAR-based multi-sensor data combinations for digital mapping and classification of highly developed urbanized coastal ecosystems.