Numerical Simulation Study on Prediction of Dynamic Water Inflow Process in the Bailongdong Section of the Yinjiang-to-Hanjiang Water Diversion Project

【Objective】To address the frequent water inrush hazards in Southwest China's karst tunnels and the inability of traditional static methods to characterize complex hydrological behaviors, 【Methods】this study proposes a coupled simulation method integrating a surface rainfall-runoff model (TOPMODEL) with a groundwater conduit-fissure dual-porosity model (MODFLOW-CFP). For the Bailongdong section of the Yinjiangbuhan Project, a 3D numerical model was developed to represent the concentrated point recharge and diffuse areal infiltration of the underground river system. Calibrated against dynamic outlet discharge data, the model predicted the magnitude and transient process of tunnel water inrush. 【Results】Results indicate that the coupled approach accurately captures the "flashy" impulsive response of the karst system to precipitation, achieving a Nash-Sutcliffe Efficiency (NSE) of 0.88. The simulation reveals a profound risk of conduit-type water inrush when the construction adit excavates through the discharge zone, exposing the main karst conduit. Highly sensitive to rainfall with rapid flow concentration and sharp recession, the peak inrush discharge surges to 24.68 m³/s under a 10-year extreme rainstorm, severely threatening construction safety. Consequently, an alignment optimization scheme is proposed, shifting the tunnel route northward to the upstream recharge area. This adjustment effectively bypasses strongly developed karst zones, reduces the catchment area, and significantly mitigates disaster risks. 【Conclusion】Ultimately, by delivering both the peak discharge and the time-variant hydrograph, this coupled model offers a robust scientific basis for construction organization and water hazard mitigation in complex karst environments.