Search

Vegetation green aboveground biomass is an important parameter of terrestrial ecosystem process-based models, and is closely related to photosynthetic activity. In this paper, the hyperspectral predictors for estimating green aboveground biomass were evaluated based on field spectral and corresponding biophysical parameter measurements during the growing seasons of 2009 and 2010 in the desert steppe of Inner Mongolia. Results showed that the performance of red-edge reflectance curve area (between normalized reflectance curve and wavelength in 680–780 nm region) was better than that of traditional vegetation indices (VIs) and red-edge position (REP). The model based on red-edge reflectance curve area produced a lower standard error of prediction (26.4 g m−2) compared with the optimal narrow-band ratio vegetation index (RVI) (37.4 g m−2), normalized difference vegetation index (NDVI) (33.5 g m−2), soil-adjusted vegetation index (SAVI) (37.5 g m−2) and modified soil-adjusted vegetation index (MSAVI) (38.6 g m−2). Statistical analyses performed on the slopes and intercepts of the fitted lines between actual and predicted values demonstrated that no statistical differences (P > 0.05) were found in regards to slope = 1 and intercept = 0 for red-edge reflectance curve area and optimal SAVI and MSAVI, whereas the slope and intercept based on optimal RVI were different (P < 0.05) from 1 to 0 respectively, and the slope based on optimal NDVI was different (P < 0.05) from 1. In conclusion, the red-edge reflectance curve area method is a practical and suitable technique for complementing the existing methods on estimation of green aboveground biomass in arid and semi-arid areas. [Copyright &y& Elsevier]