Water moves continuously between the atmosphere, the land surface and the subsurface, yet models are usually built for one compartment at a time. Atmospheric models simplify land and groundwater processes, hydrological models treat precipitation as fixed input, and groundwater models often reduce recharge to a boundary condition. Human interventions such as reservoirs, irrigation, pumping, land-use change and grey–green infrastructure further alter these connections. The gaps matter most during extremes: floods, droughts, landslides, subsidence and seawater intrusion often arise from processes that cascade across compartments, and models that miss these links can underestimate hazard and misjudge risk.

This session welcomes process-based, data-driven and hybrid modelling studies, from plot to global scale, on topics including: (1) atmospheric water, such as precipitation downscaling, extreme rainfall, land–atmosphere coupling and moisture recycling; (2) land-surface and subsurface water, such as runoff, soil moisture, evapotranspiration, snow and glacier melt, recharge, groundwater flow and surface water–groundwater exchange; (3) coupled frameworks linking compartments across scales; (4) human interventions and their feedbacks on the water cycle; (5) hydrological hazards and risk, including compound and cascading events, hazard, exposure and vulnerability assessment, and impact-based early warning; (6) machine learning, physics-informed and explainable AI, data assimilation and digital twins; and (7) uncertainty, scenario analysis and decision support.

We particularly welcome studies from data-scarce, Mediterranean and monsoon-dominated regions, work that tests models against observed events, approaches that link hazard modelling to risk reduction and water management, and model intercomparisons or benchmark datasets that help the community move forward. Early-career scientists are strongly encouraged to submit.