| 黄洁钰,庞树江,王晓燕.基于浓度与流量突变的河流总磷通量估算[J].农业环境科学学报,2024,43(3):644-653. |
| 基于浓度与流量突变的河流总磷通量估算 |
| Estimation of total phosphorus flux in river based on the change-point of concentration and flow |
| 投稿时间：2023-08-22 |
| DOI：10.11654/jaes.2023-0679 |
| 中文关键词: 突变分析 浓度与流量关系 总磷 LOADEST模型 潮河 |
| 英文关键词: change-point analysis concentration-flow relationship total phosphorus LOADEST model Chaohe River |
| 基金项目:北京市自然科学基金委员会-北京市教育委员会联合资助项目（KZ201810028047） |
|
| 摘要点击次数: 372 |
| 全文下载次数: 212 |
| 中文摘要: |
| 为进一步加深对流量与水质浓度突变特征的理解，探究浓度与流量突变点对通量模拟的影响，本研究基于Mann-Kendall突变检验法以及贝叶斯突变点模型，对潮河流域20多年来（1992—2014年）的径流量、总磷（TP）浓度以及二者之间关系进行突变分析，并结合LOADEST模型估算TP通量。结果表明：潮河流域径流、TP浓度整体均呈下降趋势，径流突变点发生于1998年，突变前后的平均流量分别为7.92 m3·s-1和2.86 m3·s-1；TP浓度突变点发生于1993年和1996年，1992—1993年、1994—1996年和1997—2014年平均浓度分别为0.08、0.06 mg·L-1和0.03 mg·L-1。TP浓度-流量关系在2004年12月前后发生突变，前后两个阶段的流量阈值分别为2.36 m3·s-1和9.08 m3·s-1。突变点前，TP浓度与流量的关系是典型的流量驱动模式；突变点后二者关系会在高流量情况下转变为稀释主导模式。基于突变点识别的分段建模有助于改善LOADEST模型的模拟效果，而不同类型突变点各有优势。基于TP浓度突变点的分段模型的整体模拟效果最佳，使纳什系数从0.50提高到0.96；基于浓度-径流关系突变点的模型对关系突变后TP通量的模拟效果最佳，使纳什系数从-0.31提高到0.89。此外本研究讨论了降水及流域管理措施的可能影响。研究表明，进行水质建模及负荷模拟时，考虑水质浓度与流量及其关系的突变可在一定程度上提高模型的适用性。 |
| 英文摘要: |
| It is necessary to gain more insights into the characteristics of flow and water-quality concentration change points and explore the role of concentration and flow change points in flux simulation. Mann-Kendall trend analysis and a Bayesian-based change point recognition model were used to evaluate the material flux of rivers with the LOADEST model. The Chaohe River Basin was selected as the research area, and the long-term(1992-2014) variation and change points of flow, TP concentration, and the relationship between the two were explored. The results indicated that the overall flow trend and TP concentration were decreasing. The flow changed in 1998, with an average flow rate of 7.92 m3·s-1 and 2.86 m3·s-1 before and after it happened. The changes in TP concentration occurred in 1993 and 1996, with average concentrations of 0.08, 0.06 mg·L-1, and 0.03 mg·L-1 from 1992 to 1993, 1994 to 1996, and 1997 to 2014, respectively.Around December 2004, there was a change point in the relationship between the TP concentration and flow. Before and after the relationship change-point, the flow thresholds were 2.36 m3·s-1 and 9.08 m3·s-1, respectively. The TP concentration-flow relationship was typical of flow-driven regimes before the change point but changed to dilution-dominant regimes under high flow conditions. The simulation performance of the LOADEST model could be enhanced by segmental modeling with change points, and different types of change points had their advantages. The segmented model based on the TP concentration change-point showed the best overall simulation results, which increased the Nash-Sutcliffe efficiency coefficient(NSE) from 0.50 to 0.96. The model based on concentration-flow relationship change-points performed best after the relationship change-point, increasing the NSE of that period from -0.31 to 0.89. In addition, the possible impacts of precipitation and watershed management measures were discussed. In summary, when simulating water quality and flux, analyzing the change points in water quality concentration, flow, and their relationship may improve the model's applicability to a certain extent. Furthermore, the potential effects of precipitation and watershed management measures were discussed. To sum up, when simulating water quality and flux, evaluating the change points in water quality concentration, flow, and their relationship may enhance the model's applicability to a specific extent. |
| HTML
查看全文
查看/发表评论 下载PDF阅读器 |
|
|
|
