基于FCM-SCEMUA的水文模型参数不确定性估计方法

Uncertainty estimation of the hydrological model based on FCM and SCEMUA

引入模糊C-均值聚类(FCM)方法对水文过程进行分类,结合SCEMUA方法,建立了基于FCM-SCEMUA的水文模型参数不确定性分析方法。选择南水北调水源区所在的汉江上游的江口流域,以新安江模型为例进行了实例研究。结果表明,FCM-SCEMUA方法通过对不同分类的似然函数分别设置阈值,在阈值同样为70%的情况下,所得到的有效参数组比通过SCEMUA方法得到的减少了64.8%的不合理参数组。所推求的参数后验分布更能够朝着高概率密度区进化,推导出更加合理的水文模型参数的后验分布,从而得到更加合理的预测区间,有效地减少了水文模拟与预测的不确定性。

An uncertainty estimation approach for hydrological model based on the （FCM Fuzzy C-Means Clustering Algorithm） and SCEMUA （Shuffled Complex Evolution Metropolis Algorithm） methods is proposed,in which the FCM method is used to classify the hydrological processes.The Jiangkou Catchment located in the upper Hanjiang River Basin is adopted as a typical basin,and the Xinanjiang model is used as a typical model,to carryout the case study.The results show that for the number of behavioral parameter sets,unreasonable parameter sets obtained by FCM-SCEMUA method,which sets the thresholds of likelihood function for different classes,decreases 64.8% as compared with that obtained by SCEMUA method under the same threshold 70%.The parameter posterior distribution obtained by FCM-SCEMUA method is found to evolve efficiently to a higher probability density （HPD） region,so as to obtain more reasonable parameter posterior distribution of the hydrological model.The FCM-SCEMUA method can be used to derive more accurate prediction bounds and to effectively reduce the uncertainty in hydrological modeling and forecasting.