基于三维随机水文地质结构模型的地下水流数值模拟: 以西辽河平原为例

Numerical Simulations of Groundwater Based on Three-dimensional Stochastic Hydrogeologic Structure Model: A Case Study from West Liaohe Plain

为了探究水文地质结构对地下水流数值模拟的不确定性,可以运用随机模拟建立地下水位的预测模型。根据转移概率地质统计方法模拟多孔介质岩性分布,利用非线性规划的思路计算岩性与水文地质参数之间的关系,从而建立相对精确的随机水文地质参数场。将不同的水文地质参数场运用到MODFLOW中,得到不同的随机模拟结果。通过比较随机模型和确定模型的末流场拟合情况以及水位动态拟合图,发现确定模型和随机模型具有相似趋势,都能与实测流场拟合较好,但是随机模型更能体现真实的水文地质特征。对随机模型预测10年后的地下水水位做不确定性分析,得到水位平均变幅介于-5～5 m之间,且95%置信度水位变幅的平均上限线约为0.146 m。研究结果为决策者提供科学依据。

To explore the uncertainty of the hydrogeological structure on the groundwater numerical simulation the West Liaohe Plain,the groundwater level was predicted using stochastic simulation.In this paper,lithologic distribution of porous media was simulated by the transition probability geostatistical method,and the relationship between lithology and hydrogeological parameters were determined by nonlinear programming.The many accurate stochastic hydrogeological parameters obtained were fed into MODFLOW,which generated the stochastic simulation.By comparing the fitting of the groundwater terminal flow field and the water level dynamic process of the stochastic and deterministic models,we found that the stochastic and deterministic models have similar trends,and are well-fitted to the measured data.The stochastic model can better reflect the real hydrogeological characteristics.According to the uncertainty analysis of the predicted groundwater level after 10 years,the average amplitude of the water level variation is ±5 m,and the average upper limit of the 95% confidence level is 0.146 m.This provides a scientific basis for decision makers.