Taking Anle Village mining area as the research object, this paper systematically analyzes the spatial and temporal differentiation characteristics of the chemical components of groundwater and surface water under the influence of mining activities and their driving mechanisms by integrating the hydrogeological model and artificial intelligence algorithms. A three-dimensional hydrogeological model was constructed based on the measured data of 45 groups of water samples collected in 2024. Combining geostatistics and machine learning methods to optimize the parameter estimation, the model was solved by the finite unit method to verify the accuracy. It is found that the surface water and groundwater in the mining area show significant differences in chemical components, and the mean values of the mass concentrations of major ions in the surface water are higher than those in the groundwater, except for NO3 and pH. The hydrochemical evolution is dominated by carbonate rock weathering, supplemented by silicate rock dissolution contributions, and cation exchange shows directional differences. Soil moisture has a significant positive correlation with groundwater level fluctuations, but there is a phase difference of 15-30 days. This paper confirms that the multi-source data fusion model can effectively reveal the dynamic evolution law of hydrochemical processes under the complex geological environment, and provide a scientific basis for the sustainable utilization of water resources in mining areas.
