Ash Gilbert (University of Colorado-Boulder, USA) will give this seminar, organized by the ELI-C department.  

Abstract 

Since the 1980s, observations show the Arctic surface has warmed four times more than the global mean. Over the Arctic Ocean, this recent large warming is connected to sea ice loss. 

While earth system models are useful tools for prediction, exact replication of observed Arctic warming and sea ice loss is not expected in freely-evolving models because of internal climate variability. 

Previous studies have shown that historical hind casts with model winds nudged to reanalysis can reproduce recent Arctic warming and sea ice loss. However, the influence of observed winds on these recent Arctic changes in absence of anthropogenic forcing has not been assessed. 

I will show that nudging to recent (1980–2023) observed winds alone in a pre-industrial model experiment does not reproduce the magnitude of observed warming and sea ice extent loss. This means that the large-scale winds are not the primary driver of recently observed large Arctic trends. 

Yet, the winds do partially reproduce the interannual, seasonal, and spatial variability, especially in spring. 

I will also show that in a pre-industrial climate simulation, these results are largely independent of mean state sea ice thickness. 

Lastly, I will briefly discuss a new attribution of Arctic temperature and sea ice changes using wind nudging.