Water quality modeling under hydrologic variability and parameter uncertainty using erosion-scaled export coefficients
Water quality modeling under hydrologic variability and parameter uncertainty using erosion-scaled export coefficients
复制标题
DOI：
10.1016/j.jhydrol.2006.03.033
复制
发表时间：
2006-10
影响因子：
6.4
通讯作者：
I. Khadam;J. Kaluarachchi
中科院分区：
文献类型：
--
作者：
I. Khadam;J. Kaluarachchi
水质模型对于评估分水岭的健康状况和作出必要的管理决策以控制接收水体的现有和未来污染非常重要。由于数据要求最低，现有的输出系数法很有吸引力；然而，这种方法没有考虑到水文可变性。本文提出了一种侵蚀尺度输出系数方法，该方法可以模拟和解释水文变化在预测年入河磷(P)负荷中的作用。这里，输沙量被引入到输出系数模型中，作为水文变异性的替代。将该方法应用于华盛顿州FISTRAIL CREK的模型P，表明了该方法相对于传统的出口系数法的优越性，同时保持了其简单性和数据要求低的特点。此外，还提出了一个贝叶斯框架来评估输出系数法的参数不确定性，而不是主观赋值的不确定性。这项工作还通过变异性-不确定性联合分析显示了分开考虑水文变异性和参数不确定性的重要性，因为它们代表了总体模型不确定性的两个独立而重要的特征。文章还建议使用纵向数据收集方案，以减少出口系数的不确定性。
Water quality modeling is important to assess the health of a watershed and to make necessary management decisions to control existing and future pollution of receiving water bodies. The existing export coefficient approach is attractive due to minimum data requirements; however, this method does not account for hydrologic variability. In this paper, an erosion-scaled export coefficient approach is proposed that can model and explain the hydrologic variability in predicting the annual phosphorus (P) loading to the receiving stream. Here sediment discharge was introduced into the export coefficient model as a surrogate for hydrologic variability. Application of this approach to model P in the Fishtrap Creek of Washington State showed the superiority of this approach compared to the traditional export coefficient approach, while maintaining its simplicity and low data requirement characteristics. In addition, a Bayesian framework is proposed to assess the parameter uncertainty of the export coefficient method instead of subjective assignment of uncertainty. This work also showed through a joint variability-uncertainty analysis the importance of separate consideration of hydrologic variability and parameter uncertainty, as these represent two independent and important characteristics of the overall model uncertainty. The paper also recommends the use of a longitudinal data collection scheme to reduce the uncertainty in export coefficients.