| Soil water movement parameters were key problems in the research of soil moisture migration. Spatial variability of water movement parameters had also been a great deal of concern and attention. On the experiment basis in Yangling and its surrounding area, the paper was constructing a single parameter model of soil water characteristic curve or soil unsaturated soil water diffusivity in 0-20 cm and 20-40 cm soil layers. Using classical statistical methods, multifractal method, joint multifractal method and pedo-transfer functions analyzed the parameter in single-parameter model on a single scale or multi-scale spatial variability in different soil layers, investigated the parameter in different soil layers of multi-scale spatial variability on the interrelationship, discussed the correlation characteristics at the multiple scales between parameter in single-parameter model and soil physical properties in different soil layers; The pedo-transfer functions at the regional scale of the parameters in single-parameter models were built on joint multi-fractal theory in 0-20 cm and20-40 cm soil layers. The main conclusions were as follows:( 1) Based on the classical statistical methods, multifractal method and joint multifractal method showed that parameter F in single parameter model of soil water characteristic curve in 0-20 cm and 20-40 cm soil layers had moderate variability in Yangling and its surrounding area. Multifractal characteristics of the parameter F in different soil layers were obvious. Respectively, the spatial variability was mainly caused by theirs low value and high value distribution. Parameter F in 0-20 cm and 20-40 cm soil layers had a high degree of correlation or some of the same spatial distribution.(2)Based on the single logarithmic model, a single parameter model of unsaturated soil water diffusivity in 0-20 cm and 20-40 cm soil layers was constructed in Yangling and its surrounding area, and the parameter S in single parameter model were all belong to the medium degree of variation. Multifractal analysis showed the multifractal characteristics of spatial distribution pattern of parameter S in different soil layers were obvious, and spatial variability is mainly caused by high value distribution. Joint multifractal analysis showed that interrelationship between spatial variability of parameter S in 0-20 cm soil layer and spatial variability of parameter S in 20-40 cm soil layer was significant in multi-scale.(3)Spatial correlation analysis of single or multiple scales comparison showed that spatial correlation characteristics between the parameter in 0-20 cm or 20-40 cm soil layers and influencing factors were not identical. Description on a single scale analysis may not be able to fully disclose in spatial correlation characteristics between the parameter in single parameter model and influencing factors. Therefore, in order to further reveal the factors contributing to spatial variability of the parameter in single parameter model should be strengthened analysis spatial correlation on multi-scale.(4)According to the joint multifractal analysis conclusions, the pedo-transfer functions at the regional scale of the parameter F in single parameter model of soil water characteristic curve or the parameter S in the soil water diffusivity single parameter model were built in0-20 cm and 20-40 cm soil layers. RMSE of parameter F in 0-20 cm and 20-40 cm soil layers predicted with the pedo-transfer functions were respectively 0.9684 and 0.6230; RMSE of parameter S in 0-20 cm and 20-40 cm soil layers predicted with the pedo-transfer functions were respectively 0.3569 and 0.4925. Predicted values of the pedo-transfer functions and measured values were close. That provided a reference for estimating soil water characteristic curve and unsaturated soil water diffusivity in the larger region. |