Abstract
Cut slopes are frequently generated by construction work in hilly areas, and artificial soil is often sprayed onto them to promote ecological rehabilitation. The artificial soil properties are very important for effective management of the slopes. This paper uses fractal and moment methods to characterize soil particle size distribution (PSD) and aggregates composition. The fractal dimension (D) showed linear relationships between clay, silt, and sand contents, with coefficients of determination from 0.843 to 0.875, suggesting that using of D to evaluate the PSD of artificial soils is reasonable. The bias (CS) and peak convex (CE) coefficients showed significant correlations with structure failure rate, moisture content, and total porosity, which validated the moment method to quantitatively describe soil structure. Railway slope (RS) soil has lower organic carbon and soil moisture, and higher pH than natural slope soil. Overall, RS exhibited poor soil structure and physicochemical properties, increasing the risk of soil erosion. Hence, more effective management measures should be adopted to promote the restoration of cut slopes.