Operational intelligence for climate-resilient water systems: a systematic–integrative review of digital technologies, hydro-climatic constraints, and system-level resilience

Abstract

Graphical abstract showing a three-part framework. The left panel summarizes the review process: 400 records identified, 120 full texts assessed, and 76 studies included. The central panel presents Operational Intelligence as a closed loop linking monitoring, forecasting, decision-making, control, and learning. The right panel shows four water-suitability outcomes: timing, quality, reliability, and usability. A bottom panel identifies four major research gaps: governance, interoperability, cybersecurity, and explainability. Water systems increasingly face hydro-climatic non-stationarity, extreme events, and water-quality deterioration that reduce the effectiveness of static infrastructure-based management. This systematic–integrative review examines how digital technologies can support climate-resilient water management by converting real-time information into operational decisions and adaptive control. Searches of Web of Science and Scopus identified 400 records; 120 full-text articles were assessed and 76 studies were included. Artificial intelligence and machine learning formed the largest domain (20 studies; 26.3%), followed by decision-support and Operational Intelligence systems (15; 19.7%), hydrology and climate systems (12; 15.8%), Internet of Things applications (10; 13.2%), water reuse and desalination (8; 10.5%), digital twins (6; 7.9%), and governance, nexus, and system-transformation research (5; 6.6%). Evidence is concentrated on monitoring, prediction, and modelling, while closed-loop control, governance, interoperability, cybersecurity, and long-term evaluation remain limited. Operational Intelligence is defined as a closed-loop capability linking monitoring, state estimation, forecasting, decision support, control, response evaluation, and learning. The review also reframes performance from water availability to water suitability, comprising timing, quality and safety, reliability, and practical usability. Future research should prioritize trustworthy AI, secure interoperable architectures, adaptive governance, multi-scale digital twins, and longitudinal validation.