基于SWAT模型的沣河流域径流模拟不确定性分析
首发时间:2011-03-14
摘要:水文模型是水资源综合管理的重要工具之一,模拟的不确定性是其应用中不可忽略的影响因素。本文以沣河流域为研究区域,借助GLUE(Generalized Likelihood Uncertainty Estimation)方法和SWAT(Soil and Water Analysis Tool)模型重点分析了流域径流模拟的不确定性,并讨论了GULE方法中似然函数取值对不确定性的影响,以及实测径流特征和模拟不确定性之间的关系。结果表明,SWAT模型中径流曲线数(CN2)、基流alpha因子(ALPHA_BF)、地下水延迟时间(GW_DELAY)、土壤有效含水量(SOL_AWC)四个参数在沣河径流模拟中较为敏感,模型可以较好的模拟沣河流域径流过程,但是模拟的置信区间不能完全覆盖实测径流过程;随似然函数的取值的增大,置信区间的区间宽度减小,但覆盖率变化不大;区间宽度与年径流均值和标准差存在较好的一致性,而区间覆盖率与年径流均值和标准差之间相关性不明显。
关键词: 不确定性分析 SWAT模型 GLUE方法 沣河流域
For information in English, please click here
Uncertainty Analyses of streamflow simulation in Fenghe watershed using SWAT model
Abstract:Hydrological model is an essencial tool for integrated water resources management, and the uncertainty of hydrological models should not be overlooked when it's applied. Using Fenghe watershed as the study area, this paper analyzed the uncertainty of streamflow simulation by the GLUE methodology and SWAT model, as the main research object, and discussed the impact of the likelihood function value of the GULE method for the uncertainty of streamflow simulation and the relationship between runoff characteristics and the uncertainty. The results show that the parameters CN2, ALPHA_BF, GW_DELAY and SOL_AWC are sensitive, and this model can simulate Fenghe watershed runoff process, but the confidence intervals of simulated runoff process does not fully cover the actual process of streamflow. Moreover, with the increasing of likelihood function value, the coverage rate of confidence interval changed little, but its width decreases significantly. Simultaneity there is a good consistency between Interval width and mean and standard deviation of annual runoff, but no significant correlation exists between the coverage rate and runoff characteristics.
Keywords: Uncertainty Analyses SWAT model GLUE methodology Fenghe Watershed
论文图表:
引用
No.4413976573408129****
同行评议
共计0人参与
勘误表
基于SWAT模型的沣河流域径流模拟不确定性分析
评论
全部评论0/1000