抄録
The TM-based Normalized Difference Vegetation Index (NDVI) data set for each month of a year was made in the Kamafusa Lake catchment by using of 11 Landsat Thematic Mapper imagery and one image-based atmospheric correction algorithm-COST method. It showed more precise variation compared with the present NOAA/AVHRR-based 1-km NDVI data set, and is practically possible to detect the vegetation cover changes in small basin area by using of high resolution TM data. The attempts to apply this data set in analysis of vegetation cover changes and their effects on suspended solids of stream water were conducted. It showed that SS concentration was correlated with the vegetation cover changes to some extent. The study also showed that two different mechanisms, i.e., snowmelt process and rainfall-runoff process, should be considered for different season when incorporating the NDVI values to analyze the stream water quality variation
