Landsat TM optimal bands selection for freshwater lake international importance wetland interpretation and monitoring

摘要

Wetland identification and dynamic monitoring is an important application for remote sensing technology due to the increasing relevance of environmental considerations. Using Landsat TM/ETM data, we analyze the image brightness curves of typical objects in the largest freshwater lake area---the Poyang Lake, which is one of the seven international importance wetlands in China. The result shows that the optimal combination bands for TM image interpretation are TM4, TM3, and TM2 assigned with R, G and B primitive color respectively. Based on image interpretation results, we establish the wetland classification system for Poyang International Importance Wetland and classify TM image of 1994 and ETM image of 2000 using supervised maximum likelihood algorithm. The wetland changes of Poyang Lake natural conservation area before and after 1998 Great Flood are analyzed and compared.