Analysis of the impacts of Arctic sea ice decrease on future climate change

Sea ice cover is one of the key components of the Earth’s climate system and regulates global weather and climate through a number of processes. Observations show that Arctic sea ice is undergoing a dramatic reduction, enhanced by the phenomenon of polar amplification, mediated by the sea-ice-albedo feedback, with potentially important impacts on climate change in a broad range of regions also outside the polar areas. The objective of this thesis is to investigate the global climate response to projected future Arctic sea ice loss, through the analysis of the patterns of many relevant climate variables. The study is carried out on an ensemble of 32 global climate models belonging to the CMIP6 framework. Their future projections following the SSP1-2.6, SSP2-4.5 and SSP5-8.5 scenarios are studied and post-processed until the end of the century. Two reference time intervals are evaluated in the historical and future periods. For each model, sea ice cover differences between the two time frames are computed and normalised by the corresponding differences in global mean surface temperature (GMT), in order to limit the influence of GMT on the results. Models are then divided into two clusters: one including models that project smaller normalised Northern Hemisphere sea ice cover difference, and one including models that project greater normalised Northern Hemisphere sea ice cover difference. The differences between the two clusters are attributed to the impacts of sea ice decrease, and it is supported by rigorous statistical testing. The effects of the difference in the amount of sea ice decrease between the two clusters are studied at four levels of global mean temperature warming relative to the historical baseline, thus allowing us to evaluate models by global warming levels, independently from the specific SSP scenario. Arctic sea ice loss impacts are investigated at global scale and for several climate variables (e.g., temperature, precipitation, sea level pressure and geopotential height), and for the atmospheric circulation by examining future zonal wind and mass streamfunction changes. Moreover, an analysis using the Empirical Orthogonal Function (EOF) is carried out, in order to detect variations in the mean patterns of large-scale variability, in particular in the North Atlantic Oscillation (NAO). Sea ice effects are studied both on an annual and seasonal (winter and summer) time scales. It arises that at the same global warming level, sea ice affects temperature not homogeneously in the Northern Hemisphere. Models projecting the stronger decline in Arctic sea ice exhibit precipitation increase in Central-Northern Europe and a marked annual mean precipitation increase in Central Africa. Furthermore, the variation in sea ice cover brings about changes in the strength of the Hadley cell, as well as modifications in other climate variables patterns. Our work confirms results from existing literature, which focuses mainly on the winter impacts of sea ice decrease, also proving the soundness of our approach, and further expands upon them by analysing impacts on a broader range of regions. Therefore, the thesis constitutes an important contribution in the framework of the investigation of the climatic consequences of projected Arctic sea ice loss.