Wave–Ice Coupling Meeting Minutes

COSIMA Waves
1

Date: 29-01-2026
Attendes: @ezhilsabareesh8, @NoahDay, @aekiss, @cbull, @sofarrell, Josh Kousal (ECMWF)
Chair & Minutes: @ezhilsabareesh8

1. Overview

This meeting focused on current developments in wave–ice interaction modelling, with an emphasis on verification approaches, model coupling strategies, computational considerations, and improvements to wave‑induced breakup and attenuation physics. Josh Kousal (ECMWF) presented recent progress in wave–ice coupling, including attenuation modelling, verification methodologies, and results from sensitivity experiments. Noah presented the ACCESS‑OM3 wave–ice coupling work developed jointly with Ezhil, highlighting advances in the wave‑induced breakup scheme, refinements to attenuation parameterisations, and ongoing model verification and evaluation.

2. Wave–Ice Interaction Model Development (Noah and Ezhil)

2.1 Current Modelling Approach

  • A three‑way coupled framework (CICE6–MOM6–WaveWatch3) is being advanced, operating globally at ~100 km resolution with hourly coupling.
  • Scientific objectives include evaluating how best to represent wave‑ice interactions in climate models, comparing coupled and standalone configurations, and improving representation of wave‑driven breakup, compaction, and deformation.
  • Work is underway to reduce computational cost by:
    • Transitioning from a stochastic breakup scheme to a parametric, strain‑based criterion.
    • Exploring options to downscale from 100 km → 25 km to better capture marginal ice zone (MIZ) dynamics.

2.2 Computational Challenges

  • Sea‑ice breakup modelling remains expensive due to fracture‑length and sea‑surface‑profile calculations.
  • Testing parametric breakup criteria expected to reduce CICE6 computation time.
  • Biases in mean wave period and significant wave height continue to be investigated.

3. Wave Attenuation in Sea Ice (Josh)

3.1 Solid vs Broken Ice Representation

  • Solid ice strongly attenuates waves; broken ice allows much greater propagation.
  • A continuous (rather than binary) representation of ice state is being used, with broken ice characterised by reduced strength (P* lowered by ~10×), resulting in higher mobility.
  • Josh showcased sensitivity experiments illustrating how wave radiative stress compacts ice at the MIZ edge.
  • Sensitivity plots demonstrate reduced ice cover (blue) or increased ice (red) consistent with established observational studies.

4. Verification and Observational Datasets (Josh)

4.1 Verification Toolkit

Josh shared his dedicated wave–ice verification package:

The toolkit:

  • Performs colocation and statistics of model data vs observational products.
  • Supports comparisons between experiments.
  • Includes example configs (e.g., config.hindcast.example.yaml).
  • Requires adaptation of:
    • Model data ingestion and staging
    • File/path conventions
    • Altimeter preprocessing logic (stageAltiData)

4.2 Recommended Observational Products

These datasets provide broad spatial sampling and are highly suitable for wave/ice attenuation validation.

5. Additional Technical Resources (Josh)

Josh also highlighted references relevant to the ACCESS wave coupling effort:

5.1 Wave‑Induced Mixing in the Ocean

5.2 Wave–Current Interaction Implementation

5.3 ECMWF Wave Model (ecWAM)

6. Discussions

6.1 Mixing Schemes

  • Problems noted with Rogers (2014) mixing implementation in NEMO when combined with wave‑forcing modifications.
  • The Met Office recently reverted from the Osmosis scheme due to tropical biases.
  • Exploring latitude‑dependent mixing but noted risks to non‑polar regions.

6.2 Machine Learning Enhancements (ECMWF IFS)

  • ECMWF’s AI‑enhanced IFS includes wave fields and improved atmospheric–wave performance (~10% improvement).
  • Trained via hindcast with updated wave datasets to emulate coupled wave–sea‑ice interactions.
1 Like