Retrieval Of Properties Of Different Genetic Soil Horizons Based On Hyperspectral Remote Sensing Data

This thesis provided spectral properties of several genetic soil horizons measured by ASD Fro2500 spectroradiometers, and soil properties were retrieved from the soil reflectance.The thesis focused on retrieving soil properties from the hyperspectral data. The first chapter mainly introduced the background, applications and potentials of hyperspectral remote sensing and precision agriculture. In the second chapter, soil survey results of the National Experimental Station for Precision Agriculture were provided. In the third chapter, the spectrum of different genetic soil horizons were compared and soil properties were related to soil spectrum. In the fourth chapter, soil organic matter and soil moisture were retrieved from soil reflectance. In the fifth chapter, the major conclusions of the thesis were summarized and future research topics were mentioned. The main results are as follows:1. Based on soil survey, the major genetic soil horizons of National experimental station for precision agriculture were Ap, At,, and B_k. The soils of the station were divided into three orders and nine soil families.2. Based on a large number of soil spectrum data, the critical wavelengths for definition of soil spectrum curves were identified. The critical wavelengths were 600, 800, 1000, 1350, 1800, 2100, and 2400nm. This is useful for soil spectral data compression and band selection.3. The relationship between soil organic matter content and soil reflectance was studied, prediction models were established. The soil reflectance and logarithm of reciprocal of soil reflectance (A value) at 447 ran was strongly correlated with soil organic matter content. The first derivative of soil reflectance at 516nm provided the best prediction for soil organic matter than those at other wavelengths.4. The relationship between normalized soil reflectance and moisture was investigated. Four prediction models were established. The model based on the logarithm of reciprocal of soil reflectance (A) at the wavelength of 1406nm provided the best prediction of soil moisture.