Determination of the Factors Governing Soil Erodibility Using Hyperspectral Visible and Near Infrared Reflectance Spectroscopy

Analysis of soil hyperspectral reflectance has been used by many researchers for soil chemical and physical property determination. In the study, we focus on the soil property which influence the soil erosion and to discuss the feasibility of hyperspectral remote sensing in soil erosion. Here, laboratory simulation experiments were conducted to analyze the change of soil physical and chemical properties parameters in the erosion process and which is the key parameters influencing the soil erosion. The rainfall simulation experiment showed that showed that the soil erosion is mainly is mainly subjected to the soil organic content and water stable aggregate under the condition of variable rainfall intensity rainfall intensity. The analysis of the above soil erosion showed that the calculation of soil erosion were significant correlated with soil organic matter content （R2 = 0.933, p < 0.01）and > 0.01 mm water stable aggregate（R2 = 0.960, p < 0.01）. In this study, wavelet analysis was applied to determine the hyperspectral reflectance data. The results show that the wavelet coefficients from spectral bands 562-658 nm, 699-839 nm, 859-1070 nm, 1727-1966 nm, 2158-2258 nm, 1350-1450 nm, 1900-205 nm are correlated with organic matter content and spectral bands from 859-1070nm,1350-1450nm,1727-1865nm,1900-2050nm are correlated with Water stable aggregate. According to the simulated rainfall experiments and hyperspectral reflectance analysis, the soil erosion should express by hyperspectral reflectance data.