Simulation study of groundwater flow and solute transport processes in karst underground rivers based on GMS

XIE Zixuan; JIANG Feng; WANG Ruofan; JIQIN Kebuzi; SHI Zheming; ZHAO Liangjie

doi:10.11932/karst20250104

Abstract：As the demand for the development and protection of groundwater resources in karst regions increases,accurately simulating the flow and solute transport characteristics of these waters becomes crucial.This study focuses on the Longdong underground river system in Zunyi City to construct a numercial model,utilizing the Groundwater Modeling System(GMS).By conceptualizing the water flow characteristics of the karst underground river with a high hydraulic conductivity coefficient(K),this study is aimed to achieve high simulation efficiency with fewer parameters.This approach is particularly advantageous in karst areas,where traditional modeling techniques may struggle to capture the complex interactions between water flow and geological features.Given the unique geological structures,karst conduits,and fractures present in these regions,a sophisticated model is necessary for proper conceptualization.The model considers the distinct heterogeneity and anisotropy of karst aquifers,with a particular focus on the complex flow patterns characteristic of conduit-dominated flow.Karst aquifers are known for their irregular and often unpredictable flow paths,which can significantly influence the movement of both water and solutes.This study underscores the importance of understanding these flow patterns,as they are critical for effective water resource management and pollution control in karst environments. To ensure the reliability of the simulation results,model identification and validation were conducted with the use of discharge data from the underground river's outlet from 2022 to 2023.This validation process is essential,as it not only confirms the model's accuracy but also enhances the credibility of its predictive capabilities for future scenarios.By utilizing measured data,the study improves the model's reliability,making it a valuable tool for researchers and pollution analysts.Subsequently,tracer tests were conducted to compare simulated and observed data,revealing temporal and spatial scale errors present in current solute transport simulations,particularly highlighting the limitations of using high hydraulic conductivity Darcy flow to characterize karst underground rivers.Tracer tests are vital in hydrological studies,as they provide insights into the movements of solutes within aquifers,enabling researchers to effectively assess of their models.The discrepancies observed in this study highlight the challenges faced in accurately modeling solute transport in environments with high hydraulic conductivity,where traditional assumptions may no longer hold true.This underscores the need for continuous refinement of modeling techniques to better align with the dynamic nature of karst systems. The findings indicate that GMS demonstrates good consistency in simulating karst groundwater flow;however,there is a need to enhance the accuracy of solute transport simulations,especially under high hydraulic conductivity(9 000 m/d),where deviations between predicted results and actual observations were noted.These discrepancies emphasize the challenges of accurately modeling solute transport under high hydraulic conductivity conditions,suggesting that reliance on high hydraulic conductivity values may lead to oversimplifications that inadequately represent the complexities of solute movement in karst systems.By adjusting the hydraulic conductivity within the model and implementing buffer zones,simulation accuracy was improved,highlighting the significant impact on the range and velocity of solute transport.This indicates that parameter optimization is key to enhancing the predictive accuracy of the model.The introduction of buffer zones also underscores their potential to mitigate scale-dependent errors,providing a novel approach for managing uncertainty in karst system modeling.This innovative method not only enhances the reliability of the model but also serves as a transitional area that helps to smooth discrepancies between modeled and observed data,thereby facilitating model convergence. Finally,this study proposes several improvements for karst groundwater simulations,including optimizing model parameter settings,incorporating more complex hydrodynamic models(such as EPM,DC,and CDC models),and increasing the model's applicability under non-Darcy flow conditions.Furthermore,the study advocates the integration of other models to better represent the interactions between karst conduits and the surrounding matrix,which could lead to more accurate predictions of both flow and solute transport in complex karst terrains.Future research should prioritize the continued optimization of model parameters and enhance the model's predictive accuracy and applicability through extensive field validation.This will provide robust scientific support for decision-making in the management and conservation of water resources in karst regions.In summary,the development of more refined models to improve the resolution of both macroscopic conduit networks and microscopic fracture systems is essential for advancing karst hydrogeological research.