Parameter Estimation of Fractal and Experimental Models of Soil Water Retention Curve Using Pedotransfer and Spectrotransfer Functions

Over the last decades, soil spectral data as a rapid, low-cost, and non-destructive method has been widely applied to estimate basic soil properties. In this study, the feasibility of using spectrotransfer functions (STFs) and pedotransfer functions (PTFs) was explored to estimate the parameters of fractal and experimental models of Soil Water Retention Curve (SWRC). For this purpose, a number of 100 soil samples were collected and their spectral reflectance over 350-2500 nm region were measured using a handheld spectroradiometer apparatus. Some soil physical properties and parameters obtained from fitting fractal and experimental models of SWRC to the measured data were determined. After spectral preprocessing, stepwise multiple linear regression was applied to derive PTFs and STFs using basic soil properties and soil spectral reflectance as input, respectively. According to the results, the parametric PTFs had high accuracy in estimating the fractal dimension of the soil particle size distribution (Dpsd) (R2 = 0.96), while the derived parametric functions had moderate predictive accuracy in estimating other studied fractal and hydraulic parameters including DSWRC-TW, DSWRC-B, λBC, nvG and bC (R2 = 0.40 – 0.59). The results also showed that the proposed spectral transfer functions (STFs) had moderately accuracy in estimating Dpsd (RPD = 1.40) and had poor accuracy in estimating DSWRC-TW, DSWRC-B, λBC, nvG and bC (RPD = 1.13 – 1.37). Overall, the results of this study showed that despite of the relatively lower accuracy of spectral parametric functions compared to pedotransfer functions, the use of soil spectral data due to simultaneous estimation of several parameters, lower cost, less time and field data (especially with development of soil information databases and spectral libraries), can be used as an indirect, rapid and novel method in estimating parameters of fractal and experimental models of SWRC.