Experimental Studies On Hydraulic Parameters Of Water Repellent Soil And Spectra Of Soil

As the basis of terrestrial ecosystem, soil not only can provide essential water andnutrients for plants, but also is the main habitat of animals. The abnormal behavior of soilalways affects crop growth severely and reduces crop yields. Soil salinization andhydrophobicity are two hot fields of soil research in the domestic and overseas. Afterprecisely estimating degree of salinity and hydrophobicity of studied soil, this type of soil canbe analyzed and then a proper method to improve soil can be proposed. The main contents ofthis study include two sections: analyzing water infiltration characteristics of different soilsunder different degree of hydrophobicity; estimating soil properties using remote sensingtechnology and evaluating the prediction accuracy. Main contents and results of this study areas follows:(1) Soil water retention curves under different water repellent levels were determined toanalyze unsaturated water movement features. Models of van Genuchten and Brooks-Coreywere contrasted for fitting soil water retention curves. Infiltration data of horizontalone-dimensional water infiltration experiment under various water repellent levels weresimulated by using Philip and Kostiakov formulas. After simulation of Philip formula, therelationship between soil permeability and WDPT (Water Drop Penetration Time) wasanalyzed. Soil water diffusivity was calculated by using horizontal infiltration method and anexponential relationship was introduced to correlate soil unsaturated diffusivity andvolumetric water content. Results show that van Genuchten and Brooks-Corey models have agood suitability for soils of different water repellent levels. During infiltration, a turning pointalways appears at certain infiltration time for water repellent soils. Before turning, fitness ofKostiakov formula is better than Philip model. When WDPT>40s, variation of soilpermeability tends towards stability and changes between0-0.1cm·min-0.5. Exponentialrelationship is feasible for simulation of soil unsaturated diffusivity and volumetricwater content and the effect for water non-repellent soils is better than for waterrepellent soils. Unsaturated water movement parameters of water repellent soils areapparently different from water non-repellent soils. The results are a good reference forfurther analyzing of water repellent soils. (2) Four types of soils were selected and light reflectance (λ) was measured at differentsoil water content (θ). Results showed that parabolic functions fit θ with measured λ very wellbut only for individual wavelengths. Multivariate linear functions of θ with measured λ atvisually selected characteristic wavelengths led to improved predictions, but the coefficientsof determination between the soil water content and measured reflectance (R2ranged from0.788to0.925for the four soils) were still not high for the studied soils. Stepwise multiplelinear regressions between θ and measured λ showed higher coefficients of determination (R2increased to0.99when the number of the statistically selected wavelengths increased) thanthe multiple regression, but had lower coefficients of determination than the stepwise multiplelinear regressions between θ and the normalized band depths (Dn). The multi-variable linearfunctions fitted the measured θ vs. Dnbest with much higher R2values, even when a singlewavelength was used. Re-sampling wavelengths of less than20nm preserved the mainfeatures of the original reflectance for the studied soils.(3) Spectral reflectance of soil samples were obtained under controlled laboratoryconditions using a portable spectrometer. A total of211samples were divided into a trainingset and a validating set for modeling soil properties. Derived from the originally observedreflectance (λ) data, six quantitative λ-related indices were applied to establish models for SCand SOMC by stepwise linear regression analysis using the training data. Adjusted coefficientof determination for each model was used for evaluating model stability. The establishedmodels were evaluated for the accuracy of predicting SC and SOMC using the validation data.Results showed that: Among the established models, the model relating λ to SC (the root meansquare error (RMSE)=2.99, the correlation coefficient (r)=0.73) had a small difference withthe model relating continuum-removed reflectance to SC (RMSE=2.94, r=0.76). Consideringthe convenience of utilization for the model and the statistical analysis of model outputs, themodel relating λ to SC was selected as the best model for prediction. With the smallest RMSE(0.84) and the largest r (0.96) values among the6models, the model relating reciprocal valuesof λ to SOMC was selected as the best model for prediction. This work supplies a feasiblemethod to estimate SC and SOMC. It has potential for remote sensing applications todetermine soil properties.