Snow plays a crucial role in how sea ice forms and evolves, yet it is still poorly represented in many computer models. Snow controls how heat and sunlight move between the atmosphere and the ocean in ice-covered seas. In winter, it can slow sea ice growth, while in summer it helps ice survive longer by reflecting sunlight back into the atmosphere. Because snow processes on sea ice are not fully understood, current models struggle to accurately simulate sea-ice thickness and its response to climate change. This limits our ability to predict future changes in the climate system. The Snow-Emu project addresses this gap by combining advanced physical modelling with machine-learning techniques to improve how large-scale climate models represent snow on sea ice. By better capturing these processes, Snow-Emu aims to deliver more reliable predictions of sea ice and climate change.

2026

2028