论文部分内容阅读
模型参数的不确定性分析是模型不确定性研究的重要内容之一。本文以示踪剂Br和反应性溶质Cu在石英砂中的运移为例,采用GLUE方法探讨了多孔介质中溶质运移模型参数的不确定性。研究结果表明,仅对水力学参数θs和λ进行识别时,θs和λ的可识别性较强。对耦合Freundlich等温吸附的模型参数进行识别时,由于参数间的相互作用,θs和λ的可识别性降低;吸附特性参数kF的后验分布基本呈均匀分布,可识别性较差,吸附特性参数β、ω、f的取值区间则相对收敛,可识别性较强。K-S检验结果表明,参数区域敏感度由高到底的排序为f、ω、β、kF、λ、θs,主要是因为石英砂对Cu的吸附以动力学反应为主,而f和ω是与动力学吸附反应相关的两个参数。上述结论有助于加深对溶质运移模型参数的理解和提高模型预测的可靠性。
Uncertainty analysis of model parameters is one of the important contents of model uncertainty research. In this paper, the transport of tracer Br and reactive solute Cu in quartz sand was taken as an example. The GLUE method was used to investigate the uncertainty of solute transport model parameters in porous media. The results show that θs and λ are more recognizable only when the hydraulic parameters θs and λ are identified. Recognition of model parameters coupled with Freundlich isotherm decreases the identifiability of θs and λ due to the interaction between the parameters. The posterior distribution of adsorption parameter kF is basically uniform and has poor recognizability. The adsorption parameters The interval of β, ω, f is relatively convergent, so it is more recognizable. The results of KS test show that the order of the sensitivity of the parameters is f, ω, β, kF, λ, θs. The main reason is that the adsorption of Cu on quartz sand is dominated by the kinetic reaction, while f and ω are related to the dynamic Learn two parameters related to the adsorption reaction. These conclusions help to deepen the understanding of solute transport model parameters and improve the reliability of model predictions.