Comparisons of alpine grassland fractional vegetation cover estimation using different digital cameras and different image analysis methods

Fractional vegetation cover (FVC) is an important indicator of grassland. Accurate estimation FVC has great emphasis for researches of soil erosion and the exchanges of water and CO2 between land surface and atmosphere. Estimating FVC with digital photos has become the most popular and efficient method in ground observation. Different types of cameras and different methods have been employed to estimate FVC. To best of our knowledge, none of studies compared the advantages and disadvantages among them. In the current study, photos from different types of alpine grassland in Shule River Basin on the Qinghai-Tibetan Plateau were taken using 3 different cameras, i.e. a multispectral camera (ADC, with red, green and near infrared rays), a modified multispectral camera (XNite, with blue, green, and near-infrared rays), and a conventional camera (COMMON, with blue, green, and red rays) . With the FVCs retrieved with WinCAM (a commercial software) as “real” values, accuracy and efficiency of the Digital Photos Processing System (DPPS) and L*a*b* method were compared. The performances of XNite and COMMON cameras were better than that of ADC. DPPS system was more accurate and flexible than L*a*b* method. With the combination of DPPS and conventional camera, FVCs of alpine grassland can be estimated accurately and efficiently.