Identifying equivalent migration pathways for water–sand inrush in coal mines by integrating multisource data and sensitivity analysis

The migration of water–sand mixtures through subsurface pathways can trigger severe geological disasters, particularly when inrush occurs into active underground spaces. Accurate identification of such pathways is therefore critical for effective hazard prevention and early warning. In this study, an equivalent pathway inversion method that integrates multisource data with numerical modelling to infer subsurface water–sand mixture migration pathways is proposed. The approach is applicable to both postdisaster investigations and predisaster risk assessments and is validated using a real case from the Wanfu Coal Mine (China). Multisource evidence, including sediment characteristics, groundwater observations, and spatial modelling, was used to constrain the source and pathway configuration. The results indicated that the primary sand source originated from a confined Neogene sand aquifer located at depths of approximately 707–724 m. Two types of pathways, single and combined, were identified, with combined pathways being more likely. Global sensitivity analysis further revealed that pathway geometry and hydraulic properties jointly control system behaviour. The inferred inlet zones are located 310–390 m from the inrush site. This study provides a reproducible approach for pathway identification in complex geological settings where direct imaging is infeasible, offering practical implications for risk assessment and disaster mitigation in underground engineering.