Research progress on the prediction technology system of acidic mine drainage water quality

Acid mine drainage (AMD) prediction is an important basis for mine environmental risk identification, classification management of mine wastes, and pollution source prevention and control. In recent years, AMD prediction research has gradually moved beyond the limitations of single testing methods and has formed a comprehensive research framework jointly supported by static tests, kinetic experiments, mineralogical characterization, and machine learning. Static tests remain widely used for preliminary screening in the early stages of mine development because they are rapid, simple, and low-cost, but they cannot adequately reflect acid and metal release rates or long-term evolution. Kinetic experiments, represented by humidity cells and column leaching tests, can more realistically reveal the weathering–leaching process and therefore provide an important basis for AMD prediction; however, they are constrained by long experimental cycles, high costs, and limited scalability for extrapolation. Mineralogical methods can strengthen the mechanistic interpretation of acid-generation and neutralization behavior from the perspectives of mineral composition, occurrence state, and microstructure, but their quantification, standardization, and engineering applicability still need further improvement. In recent years, machine learning and remote sensing coupling methods have provided new approaches for predicting key AMD water-quality indicators, regional identification, and dynamic monitoring, but they still face challenges such as insufficient samples, strong regional heterogeneity, scarce labels, limited interpretability, and weak cross-mine generalization ability. In the future, AMD prediction should further promote the integrated coupling of mineralogical parameter extraction, kinetic process characterization, field monitoring, and data-driven models, so as to build a comprehensive prediction technology system oriented toward mine waste classification management, pollution source prevention and control, and remediation decision-making.