Generative AI models capture realistic sea-ice evolution from days to decades

TL;DR

This paper introduces GenSIM, a generative AI model capable of realistically predicting Arctic sea-ice evolution over decades, capturing key dynamics and long-term trends from sub-daily forecasts.

Contribution

The paper presents the first generative AI model that accurately simulates long-term sea-ice dynamics at high temporal resolution, bridging the gap between physics-based and AI models.

Findings

GenSIM produces realistic 30-year sea-ice evolution predictions.

It captures sea-ice features like leads, ridges, and long-term trends.

The model implicitly learns multi-year ice-ocean interactions.

Abstract

Sea ice plays an important role in stabilising the Earth system. Yet, representing its dynamics remains a major challenge for models, as the underlying processes are scale-invariant and highly anisotropic. This poses a dilemma: physics-based models that faithfully reproduce the observed dynamics are computationally costly, while efficient AI models sacrifice realism. Here, to resolve this dilemma, we introduce GenSIM, the first generative AI model to predict the evolution of the full Arctic sea-ice state at 12-hour increments. Trained for sub-daily forecasting on 20 years of sea-ice-ocean simulation data, GenSIM makes realistic predictions for 30 years, while reproducing the dynamical properties of sea ice with its leads and ridges and capturing long-term trends in the sea-ice volume. Notably, although solely driven by atmospheric reanalysis, GenSIM implicitly learns hidden signatures of multi-year ice-ocean interaction. Therefore, generative AI can extrapolate from sub-daily forecasts to decadal simulations, while retaining physical consistency.