Deng, Ha, and Lee (2602.21972) build a three-level description of sea ice — particle, kinetic, hydrodynamic — that includes rotation. In Part I, ice floes were points with positions and velocities. Now they're rigid bodies that spin.
The upgrade matters because real ice floes rotate when they collide. Tangential friction during contact generates torques; asymmetric drag from ocean currents couples translation to rotation. Ignoring this produces a description that misses entire categories of stress and dissipation.
At the particle level, each floe has position, linear velocity, angular velocity, size, and moment of inertia. Contact forces include compression (normal impact), restitution (bounce-back), and tangential friction. The friction law connects linear and rotational motion: a collision that would slide one floe past another instead transfers angular momentum. Hydrodynamic drag acts on both translational and rotational degrees of freedom, with the coupling strength depending on floe geometry.
The kinetic description lifts this to a Vlasov-type equation on an extended phase space — now including angular velocity as a coordinate. Taking moments of this equation yields the hydrodynamic level: mass, momentum, and angular-momentum balance equations. Compared to Part I, the macroscopic equations gain additional stress contributions from rotation, rotational transport terms, and new dissipative mechanisms from nonlinear collisions.
The framework moves toward realistic sea-ice rheology — the constitutive relationship between stress and strain in ice-covered oceans. Current climate models parameterize this relationship crudely. The particle-to-kinetic-to-hydrodynamic pathway provides a systematic route from collision physics to the large-scale mechanical behavior that climate models need, now including the rotational physics that real collisions involve.