Aquifer heterogeneity characterization was recognized as crucial for enhancing the accuracy of groundwater models and the reliability of resource assessments and pollution control strategies. The Nugusitai Water Source, located in the Horqin District of Tongliao City, Inner Mongolia, was selected as a case study. Spatial distributions of hydraulic conductivity (K) and storativity (S_s) in a confined aquifer were characterized using three methods: Parameter Zoning, Ordinary Kriging (OK), and Sequential Gaussian Simulation (SGS). Heterogeneous isotropic models for flow and transport were constructed to evaluate the predictive impacts of each method.Results indicated that K fields generated through OK and SGS were associated with improved fits to observed hydraulic heads compared to Zoning, as evidenced by higher R² and Nash-Sutcliffe efficiency (NSE) values and lower root mean square error (RMSE). In pumping scenarios, the Zoning approach was found to cause abrupt changes in hydraulic gradients along sub-zone boundaries due to discontinuities in K values, resulting in distorted equipotential lines. The oversimplification inherent in Zoning also led to high sensitivity of predicted contaminant plume extent to the estimated K values within sub-zones. By contrast, SGS, which preserved spatial variance, produced more conservative and spatially constrained plume predictions. Storativity characterization methods were determined to have a negligible effect on model outcomes. Under common constraints such as limited borehole density, it was concluded that method selection should be aligned with specific objectives of the numerical simulation: OK was more suitable for delinerating the maximum Pollution risk Control zone, while SGS was more suitable for balancing spatial variability with local uncertainty, such as in precision remediation applications. This study provided empirical support for the selection of appropriate modeling methods in groundwater management and pollution risk mitigation efforts.