A cooperative construction strategy for multi-parameter spatial variant random fields and its application in...

The spatial variability of hydrogeological parameters is a significant source of uncertainty in groundwater numerical modeling and has a certain risk impact on the prediction of pollutant migration and transformation. Current research has focused on the effects of single-parameter spatial variant random fields or utilizing random sampling methods to randomly combine multiple-parameter spatial variant random fields while ignoring the correlation between parameters. This paper proposes an innovative concept of associated random variables to construct multi-parameter synergistic spatial variant random fields, ensuring both the spatial variability and inherent correlation of the parameters. A hypothetical case was constructed, and the Monte Carlo sampling experiment based on computer simulation was used to assess groundwater pollution risks with multiple associated parameters. The results show that hydraulic conductivity and porosity are the main sensitive parameters. The associated random variable allows for the representation of positive correlation, negative correlation, and no correlation between the hydraulic conductivity and porosity. The pollutant mass concentrations in each observation well conform to the generalized extreme value distribution, and the pollution risks of each water well as well as the concentration distribution intervals of pollutants with different probabilities can be obtained. The influence of associated parameters on the cumulative risk of contaminants in observation wells and pollution degree range is only related to their mathematical distribution and is independent of correlations between parameters. This study addresses the issues of spatial variability and inherent correlation of hydrogeological parameters, which are of great significance for groundwater pollution risk assessment and the promotion of sustainable water quality management of groundwater resources.