Simulations And Projections Of Arctic Winter Sea-ice-air Interactions Using CMIP6 Climate Model

Climate change has been becoming one of the major challenges to the sustainable development of human society,and the Arctic is one of the most sensitive regions to global climate change.The rapid climate change in the Arctic will lead to faster melting of sea ice,which has significant impacts on Arctic ecosystems,fisheries resources,and Arctic shipping.In addition,the significant reduction of sea ice in the Barents and Kara Seas during autumn and winter has potential implications for the weather and climate systems of the mid and high latitude continents of the Northern Hemisphere.The latest climate projections show that the Arctic climate will continue to change rapidly in the future,but climate models have greater uncertainties in Arctic climate projections than at mid and low latitudes.Diagnosing the main sources of uncertainty in Arctic climate projections and conducting research on the changing patterns of Arctic sea-ice-air interactions in the context of climate warming are the frontier scientific issues in Arctic climate change research.Based on the latest CMIP6(Coupled Model Intercomparison Project phase 6)climate model,this research systematically investigates the Arctic sea-ice-air interactions in the context of climate warming.The characteristics of the changes in the mixed layer depth,sea surface heat flux,sea ice concentration,surface air temperature,ocean heat transport and the interrelationships between these quantities in the Arctic Ocean under the background of climate warming are analyzed and the change regulation of the Arctic sea-ice-air interaction is given;using observed and reanalyzed data,the ability of the CMIP6 climate model to simulate the Arctic Ocean mixed layer depth,sea ice coverage,surface air temperature,and ocean heat transport are evaluated,and important sources of uncertainty in climate model simulations and projections of Arctic sea-ice-air interactions are diagnosed;finally,for the simulation of sea ice in the Barents Sea,the CMIP6 climate model is systematically evaluated to give the coincident deviations and causes of the model,to project the changes of sea ice in the Barents Sea in winter,and to explore the sensitivity of sea ice in the Barents Sea to global warming in the CMIP6 climate model.The results show that the Arctic winter will have stronger sea-ice-air interactions in the context of climate warming,especially in the Eurasian basin,mainly manifested by the deepening of the upper ocean mixed layer,increased heat release from the sea surface,dramatic reduction of sea ice and rapid increase in surface air temperature,which is mainly caused by the increase in the polar heat transport from the Atlantic Ocean(Atlantification).However,there are large differences(uncertainties)in the CMIP6 climate model projections of Arctic climate change described above,and further research indicates that the main source of these uncertainties is the ocean component of the climate model,rather than the atmospheric or sea ice component models.Climate models with different ocean component models show significantly different projections of Arctic climate.Specifically,the CMIP6 climate model with the NEMO(Nucleus European Modelling of Ocean)ocean component model(NEMO family climate model)with relatively smaller errors in simulations of mixed layer depth,Barents Sea sea ice,and ocean heat transport during the historical period,and the NEMO family climate model shows a greater increase in future polar heat transport through the Barents Sea,more significant Atlantification,stronger Arctic sea-ice-air interactions,and more intense Arctic climate change.Therefore,future research on the differences between the NEMO and other ocean models could help to significantly reduce the uncertainty in climate model projections of Arctic climate change.The research on sea ice in the Barents Sea shows that the CMIP6 climate model has the common problem of overestimating the sea ice coverage in the Barents Sea and underestimating the trend of sea ice decline,which is mainly related to the under-simulation of the polar oceanic heat transport through the Barents Sea in the climate model,and improving the horizontal resolution of the ocean model can reduce the above common bias.The projection of sea ice changes in the Barents Sea in future winters is carried out by three methods: model classification,model preference,and emergence constraint.The projection results indicate that the sea ice in the Barents Sea will decrease rapidly in future winters under the background of climate warming.Under the SSP585(Shared Socioeconomic Pathway 585)scenario,the Barents Sea will be ice-free in winter for the first time around 2060 s.It is also suggested that the sensitivity of winter Barents Sea sea ice to global warming in the CMIP6 climate model is weaker than observed,and therefore the first winter ice free time in the Barents Sea may be earlier than the above projection.This study presents the first pattern of sea-ice-air interactions in the Arctic Ocean basin under a warming context and finds that the main source of uncertainty in Arctic climate projection is the ocean component model rather than the atmospheric or sea ice component model.Moreover,it is diagnosed that the overestimation of sea ice simulation in the Barents Sea by the climate model is partly due to the underestimation of the ocean heat transport through the Barents Sea by the low resolution.The above results in this paper provide an important scientific basis for deepening the understanding of rapid Arctic climate change,reducing uncertainties in Arctic climate,and improving simulation and projection capabilities.