关键词: 温性草原, 植被光谱, 光谱特征分析, 利用方式

Abstract: Grassland is the most important and widely spread terrestrial ecosystem. The use patterns and degree of grassland influence grassland's growth and therefore affect the spectrum of the grassland. In this study, hyper-spectral data of temperate steppes in different use patterns and degree are collected from Inner Mongolia in 2011. Spectral characteristics of the data including reflectance, first derivation of reflectance, continuum removed reflectance and vegetation indices are analyzed. The analysis is helpful for the remote sensing classification of temperate steppe and provides a strong support for remote sensing monitoring of grassland degeneration. The analysis result shows that the combination of the spectral reflectance between 750～950 nm, the first derivative reflectance at 730 nm, the red edge slope, NDVI and EVI is able to clearly classify the vegetation types of Setaria italica, Stipa grandis and Leymus chinensis. The red edge slope, NDVI and EVI can differentiate enclosed Stipa grandis and grazed Stipa grandis, and distinguish cut Leymus chinensis from uncut Leymus chinensis respectively. The long-term round enclosing will lead to grassland degeneration. Conservative grazing is useful to keep grassland ecosystem healthy. In temperate steppe of Inner Mongolia, vegetation fraction and aboveground biomass with NDVI has a high correlation, and the R² value is 0.601 and 0.818 which is higher than the R² value of 0.281 and 0.675 in the correlation with EVI, respectively. So NDVI is a better choice than EVI to estimate the grassland biomass or vegetation coverage of the temperate steppe in Inner Mongolia. Estimation of aboveground biomass is more reliable than the estimation of vegetation coverage based on NDVI.

Key words: Temperate steppe, Vegetation spectrum, Spectral characteristics analysis, Use patterns of grassland