Abstract: Mining activities significantly affect the ion concentration in groundwater, and changes in mining strategies exhibit notable heterogeneity in their impact on different aquifer lithologies. Traditional methods based on empirical groundwater chemical characteristics for identifying water sources lack reliability. This study employs causal inference models to describe the evolution and heterogeneity of water chemical characteristics, and proposes a groundwater traceability inference model based on Random Forest (RF) and Generalized Random Forest (GRF). Using nearly 20 years of groundwater chemical data from the Pingshuo mining area, combined with the RF model and data augmentation techniques, we achieved intelligent traceability of aquifer lithologies with an accuracy exceeding 97%. The results indicate that adjustments in mining strategies have a significant impact on aquifer lithologies, particularly on water from mining voids and sandstone, which exhibit strong heterogeneity in ion concentrations. This heterogeneity further affects the traceability model's classification ability. The study reveals the mechanisms of how mining strategy interventions influence the variation of water chemical characteristics in different aquifer lithologies and provides scientific guidance for optimizing groundwater resource management in mining areas.