Abstract:
Objective The sand-loess binary structure is a special landform formed by multiple dynamic processes in the wind-water compound erosion region of the Loess Plateau, with a unique erosion pattern. Investigating the influence of sand cover thickness on the infiltration characteristics of loess slopes helps deepen the understanding of the erosion mechanisms of the sand-loess binary structure. However, most existing studies focus on runoff and sediment yield processes, and whether the existing soil infiltration models are applicable to the simulation of the sand-loess binary structure still needs further investigation.
Methods Laboratory simulated rainfall experiments were conducted to investigate the effects of different sand cover thicknesses (1 cm, 2.5 cm, and 5 cm) on the infiltration characteristics of loess slopes under a slope gradient of 12° and a rainfall intensity of 1.5 mm/min. The applicability of three commonly used soil infiltration models—Kostiakov, JIANG Dingsheng, and Philip models—was also evaluated in this region.
Results 1) The wetting front transport on sand-covered slopes exhibited two stages: a rapid increase before runoff and relative stabilization after runoff. Sand cover significantly increased the transport distance and rate of the wetting front. The initial runoff time was prolonged with increasing sand cover thickness. The initial runoff times for slopes with 1 cm, 2.5 cm, and 5 cm of sand cover were 1.4, 1.6, and 2.2 times longer than that of the loess slopes, respectively. 2) The presence of sand cover layer increased the soil infiltration rate, thereby shortening the time for slopes to reach stable infiltration. In addition, continued rainfall caused loamy interflow within the sand layer, resulting in surface collapse, fissure development, and rill erosion. Consequently, the infiltration rate on sand-covered slopes showed greater fluctuations compared to that on loess slopes. 3) All three models demonstrated certain predictive ability for the infiltration rate after runoff initiation on sand-covered slopes. The Kostiakov model was the most suitable for simulating the infiltration process under the sand-loess binary structure, with R2 values ranging from 0.831 to 0.931 and RMSE values from 0.055 to 0.146 mm/min. This was followed by the Philip model and the Jiang Dingsheng model, with R2 values ranging from 0.841 to 0.871 and from 0.778 to 0.905, and RMSE values ranging from 0.076 to 0.142 mm/min and from 0.065 to 0.167 mm/min, respectively.
Conclusions In summary, the presence of sand cover significantly affects soil infiltration capacity, and exhibits different patterns with increasing sand cover thickness. This study reveals the influence of sand cover thickness on the infiltration process in the sand-loess binary structure and evaluates the simulation performance of selected soil infiltration models for this unique landform. The findings provide scientific references for the comprehensive management and model construction in the wind-water compound erosion region of the Loess Plateau.