Quantitative inversion of soil sodium content and pH by hyperspectral remote sensing

Taking debris flow area as an example, this paper studied retrieval of soil Sodium content and pH by hyperspectral remote sensing, which provided a new method for estimating soil dispersion. Based on preprocessing, the authors extracted four spectral indices, including reflectance(R), inverse reflectance(1/R), inverse-log reflectance(log(1/R)) and band depth(BD), to establish the prediction model for Sodium content and pH using stepwise multiple regression method. Results indicated that reflectance spectra and inverse-log reflectance were the optimum parameters for inverting soil sodium ions content and pH, respectively. Determination coefficients R² of prediction samples were 0.690 and 0.641 respectively, and R² of test samples were 0.523 and 0.438, which showed that soil spectra with high spectral resolution had the potential for the rapid prediction of Sodium content and pH, thus, providing reliable detection method for soil dispersion using hyper-spectral technology.