Water is ubiquitous in the physical world that we experience. From being a dominant part of the natural environment on earth in the form of oceans, lakes, rivers, snow, clouds, and vapour, to being the medium in which life exists, water both literally and figuratively permeates every aspect of our existence. It plays a central role in a vast range of natural and technological processes. Despite its molecular simplicity, water exhibits complex behavior across multiple spatial, temporal, and energetic scales. Accurately modeling water involves  bridging these diverse scales, from the electronic structure of a hydrogen bond, to the long-range dielectric screening in the bulk, to collective behavior under extreme conditions, and large scale phenomena in atmospheric and planetary contexts. This multiscale complexity is both a challenge and a unique opportunity for molecular simulation.

Recent years have witnessed major advances in this direction. First-principles-based interaction potentials, such as the many-body MB-pol family, have set a new benchmark for accuracy, reproducing experimental observables across the gas, liquid, and solid phases. At the same time, machine learning (ML) approaches, including neural network potentials, are enabling first-principles accuracy with orders-of-magnitude improved efficiency. These developments allow simulations to reach new regimes of scale, both in time and system size, while retaining quantum-level fidelity.

At short scales, nuclear quantum effects (NQEs) play an essential role in the structure and dynamics of water. These effects are now accessible in ML-enhanced path-integral molecular dynamics, which combine accurate potential energy surfaces with efficient sampling of quantum fluctuations.

At larger scales, accurate treatment of long-range electrostatics, polarizability, and dielectric response remains essential, especially when studying water in heterogeneous environments such as interfaces, biological systems, and confined geometries.

These methodological advances are transforming our ability to address fundamental questions across a spectrum of fields:

Nanoconfined water, such as in graphene slits or zeolites, exhibits structural and dynamical behaviors distinct from the bulk, including directional hydrogen-bond networks and altered dielectric response.

Biological water mediates the function of proteins, membranes, and nucleic acids, influencing recognition, dynamics, and energetics. Accurate models must capture both the local hydration environment and collective fluctuations.

Extreme water, found in planetary interiors, exhibits exotic phases, such as plastic ice, recently observed at high pressures, and is crucial to understanding planetary formation, magnetic fields, and thermal transport.

Atmospheric water, such as the water in clouds involving coalescence of droplets, and phase change thermodynamics, evaporation from surfaces of oceans or soil, and interactions with electric fields such as those from lightning, offer examples of phenomena which may require modeling beyond conventional fluid dynamics.

The emergence of flexible, scalable models that can adapt across these diverse regimes marks a conceptual shift in the field. Rather than targeting isolated observables or phases, the frontier is now to develop unified water models that perform robustly across multiple scales and contexts, a challenge requiring careful attention to transferability, interpretability, and validation.

This workshop will bring together experts from diverse subfields, first-principles simulations, ML-driven modeling, quantum dynamics, mesoscale methods, and applications in biology, planetary and planetary science, to assess the state of the art and chart the way forward in modelling water across scales.