Water Resource Diversification through Integrated Management of Groundwater Dams and Bedrock Wells

The importance of securing stable water resources has become increasingly evident due to growing spatiotemporal uncertainty in precipitation and the more frequent occurrence of prolonged droughts caused by intensifying climate change. In this context, groundwater dams have gained prominence as an alternative water resource with low evaporation losses and high climate resilience. However, the independent operation of groundwater dams may have limitations in ensuring mid- to long-term water supply stability. Therefore, the development of an integrated water resource management system that considers conjunctive use with other water sources―such as surface water, bedrock wells, and artificial recharge facilities―is required.

The objective of this study is to demonstrate core technologies for a networking system linking  groundwater dams and bedrock wells. The Ssangcheon watershed in Sokcho City, Republic of Korea―where two groundwater dams are currently operated for domestic water supply―was selected as the study area. Comprehensive analyses of geological structures, aquifer distributions, and hydrological conditions were carried out. The hydraulic and geological properties along with hydrochemical types of both alluvial and bedrock groundwater were characterized. Water balance analysis was performed to quantify available water resources within the watershed and to assess surface water–groundwater interactions.

To assess water supply stability, we conducted scenario-based simulations of integrated groundwater dam and bedrock well operations, focusing on drought and high-demand conditions. In addition, deep learning models based on Long Short-Term Memory (LSTM) networks for one-step prediction and an encoder–decoder LSTM architecture for multi-step prediction were developed to predict groundwater levels in support of the integrated operation of the Ssangcheon Dam.

The results indicate that integrating groundwater dams with bedrock wells substantially improves both water supply reliability and water quality protection. Moreover, AI-based groundwater level prediction techniques proved to be effective tools for proactive water resource management and the development of smart water management systems. This study offers a practical framework for water resource diversification, providing a foundation for developing sustainable and climate-resilient management strategies.

This work was supported by the Management Technology for Groundwater Dams in Water Supply Vulnerable Areas Program of the Korea Environmental Industry & Technology Institute (KEITI), funded by the Ministry of Environment (MOE) (RS-2025-01842973).