The evolution of digital twins in hydrology and environmental science: From physical models to AI-assisted autonomous systems

Abstract

Digital twin (DT) technologies have emerged as a transformative framework in hydrology, enabling adaptive, real-time modeling of water systems through data-driven intelligence. This position article proposes a five-level technological evolution model for hydrological DTs, tracing the field’s progression over the last three decades (1995–2025) from physical models toward autonomous and interconnected systems. Each level is anchored in technological milestones such as web systems, geographical information systems, Internet of Things, artificial intelligence (AI) and immersive environments that collectively enhance the interactivity, intelligence and integration capacity of DT systems. The study introduces a layered implementation framework that links enabling technologies to functional capabilities across the DT life cycle. Drawing from this 30-year synthesis and real-world applications, we illustrate how DTs are being used for flood prediction, watershed management, infrastructure resilience and stakeholder engagement. A cross-level capability matrix is presented to analyze DT levels in terms of data requirements, visualization methods, computational demand, expertise and cost. The study also identifies critical research challenges in data interoperability, AI ethics, cybersecurity and institutional coordination. Recommendations are provided to guide future research, emphasizing open standards, modular architectures, participatory design and public trust. This study presents a vision for hydrological DTs as scalable, intelligent and ethically grounded systems that support climate resilience, disaster mitigation and sustainable water governance.