A joined multi-metric calibration of river discharge and nitrate loads with different performance measures

A joined multi-metric calibration of river discharge and nitrate loads with different performance measures
复制标题

DOI:
10.1016/j.jhydrol.2016.03.001
复制
发表时间:
2016-05
影响因子:
6.4
通讯作者:
M. Haas;B. Guse;M. Pfannerstill;N. Fohrer
M. Haas;B. Guse;M. Pfannerstill;N. Fohrer
中科院分区:
地球科学1区
文献类型:
--
作者:
M. Haas;B. Guse;M. Pfannerstill;N. Fohrer

文献摘要

被引文献

相似文献

水文模型是研究流域水文和水质的有用工具。这些模型的校准是使模型适应集水区条件的关键步骤,可有效模拟环境过程。在模型的校正中,需要考虑不同的性能指标来代表不同的水文和水质条件,本研究提出了一个联合的多尺度校正的流量和硝酸盐负荷的生态水文模型SWAT。为此,提出了一种基于流动持续时间曲线(FDC)的校准方法,同时考虑硝酸盐持续时间曲线(NDC)。FDCs和NDC的五个区段分别进行评估,以考虑过程线和氮素线的不同阶段。为了同时考虑河流流量和硝酸盐负荷的大小和动态,Kling-Gupta效率(KGE)被额外用作统计性能指标,以实现联合多变量calibration. Results结果表明,一个单独的评估五个不同的幅度提高了校准的硝酸盐负荷。随后,适当的模型运行具有良好的性能,为不同的水文条件下,无论是流量和硝酸盐检测在一个联合的方法的基础上FDC,NDC,和KGE。通过这种方式,在同一模型运行中获得了排放和硝酸盐负荷的合理结果。使用多指标性能方法,同时进行多变量校准可以为所有量级的排放量和硝酸盐负荷产生平衡的模型结果。
Hydrological models are useful tools to investigate hydrology and water quality in catchments. The calibration of these models is a crucial step to adapt the model to the catchment conditions, allowing effective simulations of environmental processes. In the model calibration, different performance measures need to be considered to represent different hydrology and water quality conditions in combination.This study presents a joined multi-metric calibration of discharge and nitrate loads simulated with the ecohydrological model SWAT. For this purpose, a calibration approach based on flow duration curves (FDC) is advanced by also considering nitrate duration curves (NDC). Five segments of FDCs and of NDCs are evaluated separately to consider the different phases of hydrograph and nitrograph. To consider both magnitude and dynamics in river discharge and nitrate loads, the Kling–Gupta Efficiency (KGE) is used additionally as a statistical performance metric to achieve a joined multi-variable calibration.The results show that a separate assessment of five different magnitudes improves the calibrated nitrate loads. Subsequently, adequate model runs with good performance for different hydrological conditions both for discharge and nitrate are detected in a joined approach based on FDC, NDC, and KGE. In that manner, plausible results were obtained for discharge and nitrate loads in the same model run. Using a multi-metric performance approach, the simultaneous multi-variable calibration led to a balanced model result for all magnitudes of discharge and nitrate loads.