Water Resources Research Act Program

Details for Project ID GU_2024_Kim2

The analysis of saltwater intrusion into a freshwater resource with deep learning tools.

Institute: Guam
Year Established: 2024 Start Date: 2024-09-01 End Date: 2025-08-31
Total Federal Funds: $20,460 Total Non-Federal Funds: Not available

Principal Investigators: Yong Sang Kim, Byoungyong Lee

Project Summary: Currently, the Guam Waterworks Authority (GWA) produces 85% of the potable water (45 MGD) from its main source, the Northern Guam Lens Aquifer (NGLA). GWA manages 100 to 120 deep vertical production wells in this aquifer to meet demand. However, many of these production wells have shown increased chloride levels over several decades. WERI’s chloride research targeting GWA production wells revealed that chloride concentrations have exhibited significantly increasing trends in freshwater production wells (70% of production wells). There are several possible reasons for the increase in chlorides: 1) increased water production rate, 2) increased sea-level rise since 1993, and 3) varied precipitation rate and intensity. Deep learning (DL) is a valuable modeling tool for applying historical datasets and estimating contaminant trends. Environmental field monitoring data alone are not easy to analyze due to their complexity, which incorporates precipitation, tide, or sea level variation, but DL application is expected to provide more accurate insights into changes in chlorides in groundwater resources. In this study, we will select production wells located in the basal zone of the Finegayan basin and/or Yigo-Tumon basin as target wells to project chloride variation. Using historical chloride data, the DL model will learn Guam’s chloride variation trends in the fresh lens of the Aquifer. After calibrating and validating the DL model, future chloride variation trends will be projected, and the impact of the results will be interpreted.