Estimation of fractional vegetation cover and its relation with permafrostin the upstream regions of Shule River Basin

Fractional vegetation cover (FVC) is an important factor in researches of hydrology, ecology, and meteorology. Nevertheless, the principal way of getting FVC is to estimate visually in situ or from digital picture. In this study, we applied multispectral camera to take pictures of alpine grassland, and then analyzed normalized difference vegetation index (NDVI) of each pixel of pictures to distinguish vegetation and bare land, and calculated the FVC. We compared the FVCs from multispectral camera and those from visual estimation, and found: (1) the visual estimation of FVC was subjective, however, the average of these estimates was reasonable; (2) the estimation using multispectral pictures was accurate and rapid; (3) the value of 0.4 for NDVI could be used as a threshold to distinguish vegetation and bare land; (4) to apply the threshold of NDVI for estimating FVC, the spatial scale of a pixel in a multispectral picture should not be more than 0.2 cm. We also studied the relations between FVC and permafrost distribution, and found that the FVCs in the permafrost area were significantly higher than those in the seasonal frost and transitional areas.