含水介质非均质概化对地下水数值模拟的影响分析

Analysis on influence of aquifer heterogeneous simplification to groundwater numerical modeling

实际地下水系统的复杂程度远远超出目前人们可以用数学方法准确刻画的程度,地下水模型建立通常依赖于一系列对野外真实情况的假设和近似。由于观测资料的稀疏性以及观测误差等人为因素,导致含水介质非均质性参数分区不能准确地反映实际情况,存在不同程度的概化,便会引起数值模拟结果的不确定性。本文着重探讨了不同程度的含水介质非均质概化对地下水数值模拟中参数识别、降深预报以及风险评估等问题的影响。通过将贝叶斯理论耦合地下水流数值模拟软件MODFLOW,结合单分量自适应Meteropolis采样的马尔可夫链蒙特卡罗方法(SCAM-MCMC),可以用来获取模型参数和降深的贝叶斯后验分布。算例研究的结果表明,该方法对于含水介质非均质概化引起的不确定性能进行系统的量化分析。且借助贝叶斯后验分布对模型预报量能给出全面有效的风险分析,结果可为地下水资源利用和管理提供科学决策依据。

Groundwater modeling commonly relies on a serial of hypothesis and approximations for field reality.Since the real hydrologic systems are far more complex than that we could mathematically characterize,the real conditions could be reflected owing to the lack of the observation data and observation errors and the divisions of the parameters of the heterogeneous aquifer,thus the different kinds of simplification may contribute to the model uncertainty.The impacts of different kinds of simplification to the heterogeneous aquifer are studied for the parameter identification,the drawdown prediction and the risk analysis in the groundwater numerical simulation.The Bayesian algorithm is proposed based on the SCAM-MCMC sampler which is coupled with the groundwater modeling software MODFLOW,and the posterior distribution of the parameters and the drawdown could be obtained.An ideal example is presented to illustrate that this algorithm could quantify the uncertainties which are caused by different kinds of the simplification of the heterogeneous aquifer.The comprehensive and effective risk analysis is also derived from the Bayesian results to draw trade-off curves for the decision making about the exploitation and management of groundwater resources.