Wang Chenfeng, Fu Xudong, Wang Bin, Gong Zheng, Zhang Ga, Wang Xiaoping
College of Natural Resources and Environment, Northwest A&F University, Yangling, Shaanxi 712100, China; State Key Laboratory of Hydroscience and Engineering, Tsinghua University, Beijing 100084, China.
State Key Laboratory of Hydroscience and Engineering, Tsinghua University, Beijing 100084, China.
Sci Total Environ. 2023 Nov 25;901:166032. doi: 10.1016/j.scitotenv.2023.166032. Epub 2023 Aug 2.
A dynamic model of soil erosion along hillslopes considering soil detachment and sediment transport is still a major challenge in terms of its applicability to field conditions. Data availability for model calibration and validation is very limited for physically process models. An improved hillslope erosion dynamic model (HED) with a simple structure and strong application on field plots was established based on the sediment feedback mechanism in this study. Observed runoff and sediment data from field plots with slope gradients of 14.1 %-62.5 % and slope lengths of 7.9-64.7 m within runoff events for the Chagagou catchment on the Loess Plateau of China were used to evaluate the HED. We confirmed that the power function can reproduce the soil detachment capacity (Φ) and sediment transport capacity (T) under varying field slope conditions (gradients and lengths). The two parameters associated with the power function of Φ or T are consistent across the variable conditions. When the HED model simulates the process and event sediment discharge, the unified model parameters could be obtained. The simulation precision of above results ranged from 0.44 to 0.95 for Nash-Sutcliffe simulation efficiency (NSE), from 0.65 to 0.96 for R and from -32.37 % to 31.61 % for relative error (RE). The feedback term of HED was close to one as the slope length approached zero. Decreasing of sediment yield as slope length increased was quite consistent with the measured data due to the reasonable sediment feedback term. The critical slope lengths were more easily reached, and the feedback term value along the slope decreased more rapidly at steeper slopes and higher runoff events. HED can be further integrated into distributed watershed models for predicting sediment discharge within runoff events.
考虑土壤分离和泥沙输移的坡面土壤侵蚀动态模型在实际应用中仍面临重大挑战。对于物理过程模型而言，用于模型校准和验证的数据非常有限。本研究基于泥沙反馈机制，建立了一个结构简单且在田间小区具有较强适用性的改进坡面侵蚀动态模型（HED）。利用中国黄土高原岔巴沟流域径流事件中坡度为14.1%-62.5%、坡长为7.9-64.7m的田间小区实测径流和泥沙数据对HED进行评估。我们证实，幂函数能够再现不同田间坡度条件（坡度和坡长）下的土壤分离能力（Φ）和泥沙输移能力（T）。与Φ或T的幂函数相关的两个参数在不同条件下是一致的。当HED模型模拟过程和事件泥沙排放量时，可以获得统一的模型参数。上述结果的模拟精度，纳什-萨特克利夫模拟效率（NSE）为0.44至0.95，相关系数（R）为0.65至0.96，相对误差（RE）为-32.37%至31.61%。随着坡长接近零，HED的反馈项接近1。由于合理的泥沙反馈项，随着坡长增加产沙量减少与实测数据相当一致。在更陡的坡度和更高的径流事件中，临界坡长更容易达到，沿坡面的反馈项值下降得更快。HED可进一步集成到分布式流域模型中，以预测径流事件中的泥沙排放量。
