Hyperspectral remote sensing estimation models for the grassland biomass

In order to improve the research and application of hyperspectral remote sensing in the quantification of biophysical indices and biochemical indices of grassland, and estimate the grassland biomass at the canopy scale, an ASD FieldSpec Pro FRTM spectroradiometer was used for the spectral measurements of natural grassland in Xilin Gol League, Inner Mongolia. First, correlation between original spectral, hyperspectral feature variables and above-ground biomass of natural grassland was analysed. Second, the basic experiment data including biomass and canopy reflectance of natural grassland were classified into two groups. One group was used as the training sample to build the regression models with the one-sample linear method, the nonlinear method and the stepwise analysis method; the other group was used as the testing sample to test the precision of regression models. Results show that the stepwise regression estimation model using the five hypeespectral reflectances of 840, 1132, 1579, 1769 nm and 2012 nm was the best, the estimation standard deviation was 0.404 kg/m², the estimation precision was 91.62%. The results of this paper indicate that the grassland biomass can be estimated at the canopy level using the hyperspectral reflectance.