Applications of AI for water management

The emergence of increasingly capable Artificial Intelligence (AI) technologies is changing the technical landscape in many scientific and technical disciplines, and hydrology is no exception. The numerous applications include surface water supply, groundwater modelling, hydropower generation, agriculture and irrigation, climate change risk and flood risk management, water-energy-food nexus, water governance, among others. Although AI technologies have the potential to unlock new capabilities in the context of water management, these developments are also subjected to several limitations. For instance, data-driven methods often require access to large-scale (space and time) and high-quality measurements, which sometimes are not representative of extreme values. Additionally, the lack of interpretability or explainability of these models is highly relevant since they support decisions and control over critical processes and structures, while also having ethical implications. This publication reviews the current state-of-the-art of AI and Machine Learning (ML) applications within water management, introducing some of the main concepts and providing the reader with a general understanding of different technologies and concepts. Further, it features examples of the most influential applications of AI within water management and highlights the ethical challenges when streamlining AI for water resources management.