A Smart Near-Real-Time Urban Groundwater Digital Twin-Based Information System for Monitoring, Forecasting, and Pump Control

Abstract. In response to climate variability challenges, urban planners and water authorities increasingly rely on near-real-time urban groundwater information systems that integrate sensor data, predictive modelling, and geospatial analytics. Within smart-city frameworks, Digital Twins (DTs) provide solutions for integrating these data streams into unified operational monitoring and decision-support environments. This study proposes a DT framework for dynamic groundwater table (GWT) monitoring and control during rainfall and drought fluctuations. The contribution consists of (1) a DT-based approach enabling near-real-time GWT monitoring and operational control, (2) an integrated architecture linking sensors, weather data, machine learning-based GWT prediction from rainfall data, 3D geospatial visualisation, and actuator control for water pumps, and (3) a municipal decision support prototype demonstrated in the city of Enschede. The DT visualises historical, current, and anticipated GWT conditions and enables threshold-based actuation (activating or deactivating real-world water pumps), supporting operational groundwater management. This proof-of-concept demonstrates the potential of video game engines and DT for groundwater monitoring and control, supporting prescriptive operational decisions by detecting when simulated groundwater reaches the ground surface and triggering pump activation or deactivation. The findings highlight the potential of integrating Digital Twins for urban space maintenance and for managing climate change-related challenges, such as drought and water ingress, through prompt actions based on GWT behaviour.