基于径流侵蚀功率的长江典型流域能沙关系模型及改进

基于径流侵蚀功率概念建立流域能沙关系模型，可为长江流域泥沙变化精准模拟与水土保持规划提供技术支撑。该研究以长江典型流域及其典型小流域为研究对象，通过收集1965—2018年金沙江流域、嘉陵江流域和湘江流域等3个典型流域逐日水沙数据以及万安和李子口等2个典型小流域2014—2020年场次降水径流泥沙数据，采用径流侵蚀功率、径流量和降雨侵蚀力对比分析不同时空尺度水沙（径流量和输沙量）、雨沙（降雨侵蚀力和输沙量）和能沙（径流侵蚀功率和输沙量）关系的优劣性，解析能沙关系优越性，并识别能沙关系非一致性变化，从而改进能沙关系模型提高流域输沙量模拟精度。结果表明：1）长江流域3个典型流域及2典型小流域，在绝大部情况下能沙关系的表现总是优于水沙关系和雨沙关系，在场次、月和年尺度R²_adj最大值分别可达到0.94、0.87和0.58。2）对于不同时间尺度，其流量序列中任意2个流量乘积与输沙量的相关性较高时，第一个流量Q₁分位点总是接近1且第二个流量Q₂分位点在0.5附近或者高于0.5。基于径流侵蚀功率可以较为准确地计算不同时空尺度流域输沙量，具有明显适用性。3）随着时间升尺度，水沙、雨沙和能沙关系逐渐变差，3个典型流域径流侵蚀功率和输沙量在一些月份上均存在显著变化趋势和显著突变点（P<0.05）。特别是在年尺度上，输沙量均为显著减少趋势（P<0.05），其能沙关系均表现出非一致性变化。4）水库建设和植被增加是导致流域能沙关系变差的重要原因，其均与输沙量呈现极显著负相关（P<0.001），通过考虑水库指数和NDVI改进能沙关系模型，年R²可提高27.28%～97.62%。研究成果可支撑开发新的流域泥沙预报模型，服务长江流域生态保护与高质量发展。

Energy-sediment relationship model based on runoff erosion power and its improvement in typical Yangtze River basin

Based on the concept of runoff erosion power, to establish the basin energy-sediment relationship model can provide theoretical support for the accurate simulation of sediment load change and soil and water conservation planning in the Yangtze River Basin. In this study, the typical watersheds of the Yangtze River and its typical small watersheds were taken as the research objects. By collecting the daily precipitation, runoff and sediment data of three typical watersheds including Jinsha River Basin, Jialing River Basin and Xiangjiang River Basin from 1965 to 2018, and the rainstorm flood scale precipitation, runoff and sediment data of two typical small watersheds including Wan ''an and Lizikou from 2014 to 2020, the runoff erosion power and rainfall erosivity were used to compare and analyze the advantages and disadvantages of the relationship between runoff and sediment load, rainfall erosivity and sediment load, and runoff erosion power and sediment load at different spatial and temporal scales. The superiority of the relationship between runoff erosion power and sediment load was analyzed, and the inconsistent changes of the relationship between runoff erosion power and sediment load were identified, so as to improve the model of the relationship between runoff erosion power and sediment load and improve the simulation accuracy of sediment load in the watershed. The results showed that: (1) In most cases, the performance of runoff erosion power and sediment load relationship was always better than that of runoff and sediment load relationship, rainfall erosivity and sediment load relationship in three typical watersheds and two typical small watersheds in the Yangtze River Basin. At the event, monthly, and annual scales, the maximum R²_adj values can reach 0.94, 0.87, and 0.58, respectively. Sediment load always increased with the increase of runoff, rainfall erosivity and runoff erosion power. (2) For different time scales, the first flow Q₁ quantile was always close to 1 and the second flow Q₂ quantile was around 0.5 or higher than 0.5 when the correlation between any two flow products and sediment load in the flow sequence was high. Based on the runoff erosion power, the sediment load of watersheds at different spatial and temporal scales can be accurately calculated, which had obvious applicability. (3) With the time upscaling, the relationship between runoff and sediment load, rainfall erosivity and sediment load, and runoff erosion power and sediment load gradually deteriorated. The runoff erosion power and sediment load of the three typical watersheds had significant trends and significant change-points in some months (P<0.05). Especially on the annual scale, the sediment load showed a significant decreasing trend (P<0.05), and the runoff erosion power and sediment load relationship showed inconsistent changes. (4) Reservoir construction and vegetation increase were important reasons for the deterioration of the runoff erosion power and sediment load relationship in the basin, which were significantly negatively correlated with sediment load (P<0.001). By considering the reservoir index and NDVI to improve the runoff erosion power and sediment load relationship model, R² can be increased by 27.28%-97.62%. The research results will support the development of new watershed sediment prediction models and serve the ecological protection and high-quality development of the Yangtze River Basin.