Abstract

Large-scale hydrological extremes — such as droughts and floods — disrupt food, water, energy, and ecological systems, amplifying vulnerabilities across both natural and human systems. These risks intensify when events occur as spatially and temporally synchronized phenomena, whether as simultaneous, in-phase occurrences locally or as opposite-signed dipoles, or "seesaws," across regions. Such interconnected extremes can trigger cascading effects throughout socio-economic systems, from global food price spikes to supply chain disruptions, with severe consequences for vulnerable populations. Although droughts and floods represent opposing sides of the same hydrological cycle, they present a profound conundrum for hazard governance: how can we effectively manage these dual threats when strategies addressing them have traditionally been studied in isolation?

This lecture synthesizes my past and ongoing research to advance our understanding of these hydrological extremes through three interlinked themes: Catalogue, Connection, and Complementarity. The Catalogue theme introduces the Global Drought and Flood Catalogue (1950–2016), which establishes a consistent risk assessment framework by integrating in-situ observations, remote sensing, and advanced hydrological modelling. The Connection theme uncovers critical linkages between droughts and floods, leveraging recent advances in event coincidence analysis and complex network approaches. These techniques overcome computational barriers and allow us to reveal spatially and temporally compound events, including drought-pluvial seesaw phenomena, across scales ranging from global to local. The Complementarity theme explores how innovative solutions, such as utilizing flood-managed aquifer recharge to enhance drought resilience and leveraging renewable-hydropower complementarity through adaptive reservoir operations, can strengthen energy integration and water security.

Through case studies in California, the Mekong River Basin, and China’s Southern Power Grid, I demonstrate how climate-informed, multi-sectoral modeling frameworks — powered by latest deep reinforcement learning techniques — can inform adaptive infrastructure investments and guide climate-aware policy decisions despite profound uncertainties across the water-energy nexus. This integrated perspective, developed through the Pathways for REsilient Planning of water-energy-food Nexus Transformation (PREP-NexT) lab, empowers society to PREP for the NexT frontier of environmental challenges, transforming climate risks into opportunities for building resilient and sustainable systems.