A Hybrid AHP-AI Framework for Assessing and Mitigating Tunneling Impacts on Confined Spring Discharge

Abstract

Tunnel excavation poses a significant threat to groundwater systems, particularly impacting confined tunnel springs and jeopardizing hydrological balance and water resource sustainability. This study tackles the challenges of tunnel-induced groundwater changes by introducing a novel hybrid framework. This framework combines expert qualitative insights with advanced quantitative modeling, leveraging the Analytic Hierarchy Process (AHP) and artificial intelligence. Building upon existing research that uses numerical and empirical models to assess tunneling impacts on aquifers and spring discharge, our approach integrates multi-criteria decision-making to identify and prioritize factors driving groundwater loss. We evaluate parameters impacting spring discharge, highlighting the importance of karst cavities, rock mass permeability, fracture aperture, and rock mass type. The proposed model categorizes excavation sites into six impact levels, from harmless to completely hazardous. Key findings underscore the influence of tunnel depth, excavation method, and geological conditions on groundwater drawdown and spring depletion. This framework offers an improved decision model for predicting hydrological consequences, bridging research gaps, and providing practical guidance for mitigating environmental impacts, ensuring sustainable groundwater management, and optimizing tunnel construction