Large-ensemble Numerical Study On The Impact Of Arctic And Antarctic Sea Ice On The East Asian Summer Monsoon Under Global Warming

In recent decades,the global temperature has been increasing,the Arctic sea ice has warmed more significantly,and the Arctic sea ice has rapidly melted.The Antarctic continent is located at high latitudes in the southern hemisphere,and the trend of Antarctic sea ice shows regional variability.In recent years,the Antarctic temperature has also been increasing and the trend of Antarctic sea ice has been melting.In recent decades,precious studies have partially revealed the possible impact of the reduction of the Arctic and Antarctic sea ice on climate change in East Asia,but there is still uncertainty in the understanding how the sea surface temperature(SST)and the sea ice affect East Asian summer monsoon in the context of global warming.On the one hand,the conclusions are still controversial due to the inconsistency between the observational analyses and numerical simulations.On the other hand,the model response to the Arctic sea ice forcing is generally weak and not easily discerned from the noise.The large ensemble can effectively reduce the model noise and make the results more robust.In this paper,we first evaluate the simulation skill of the large ensemble for the Asian monsoon and the southern hemisphere circulation.Due to the demand of conducting numerical experiments,we also evaluated the East Asian summer monsoon for the Atmospheric Model Intercomparison Project(AMIP)and historical runs of the Chinese Academy of Sciences(CAS)Flexible Global Ocean–Atmosphere–Land System(FGOALS-f3-L)model.Then,the large ensemble simulations of the CMIP6 Polar Amplification Model Intercomparison Project(PAMIP)were used to investigate the effects of the Arctic sea ice and Antarctic sea ice reduction and the global SST increase on the East Asian summer monsoon and their differences.The physical mechanisms were further explored.On this basis,further numerical experiments are conducted using the atmospheric model FAMIL of the FGOALS-f3-L climate system model to analyze the synergistic effects of the simultaneous reduction of bipolar sea ice on East Asian summer precipitation,and to give quantitative responses of East Asian summer monsoon to the simultaneous reduction of bipolar sea ice and the response to the reduction of unipolar sea ice.The main conclusions are as follows:(1)The simulation capability for the Asian summer monsoon(ASM)and the Southern Hemisphere Annular Mode(SAM)in the large ensemble and the simulation skill for the interannual variability of the East Asian summer monsoon in the FGOALS-f3-L model are revealed.For Asian summer precipitation,the mean of the 20-member perturbed parameter ensemble(PPE-20M)could well capture the precipitation and the wind patterns of ASM.The spatial patterns of precipitation are hardly affected by the parameter perturbations,while the sensitivity of the perturbation parameters shows mainly the effect on the magnitude of precipitation biases.Four parameters,such as deep entrainment amplitude,are sensitive to the bias of the ASM precipitation simulation.We also evaluated the simulation capability of SAM.The multi-model mean(MME)for both AMIP experiment and historical experiment could capture the basic spatial patterns of the SAM for four seasons.The simulation skill varies for different seasons: the skill for boreal spring(March-May,MAM)is the lowest,while it is highest in boreal winter(December-February,DJF).Further analysis shows that the simulation of the asymmetric part of the SAM is the key to the SAM pattern simulation.In addition,the FGOALS-f3-L historical experiment can basically reproduce the multivariate empirical orthogonal function(MV-EOF)lead mode.The historical experiment could also better simulate the time frequency of the EASM variability than the AMIP experiment.Further analysis shows that air-sea interactions in the tropical Indian Ocean and western Pacific appear to be important for simulation of EASM,while El Ni?o-Southern Oscillation(ENSO)simulations have a large impact on EASM interannual variability simulations.(2)Multi-model large ensemble simulations of the PAMIP experiment were used to reveal the effects and possible mechanisms of Arctic sea ice decrease and sea surface temperature(SST)increase on the interdecadal trends of EASM precipitation in the context of global warming.The effect of Arctic sea ice reduction on interdecadal EASM precipitation trends in East Asia is comparable to that of sea surface temperature(SST).In June,a tripole pattern with a positive precipitation anomaly centered in central China was observed,and in July,a “southern China flood and northern China drought” pattern was observed.These patterns are closely connected with the regional differences of Arctic sea ice loss from June to July.In June,the reduction of sea ice in the Barents-Kara Sea(BKS)region triggered the propagation of upper-level circulation anomalies and stationary wave fluctuations to the eastern part of the Eurasian continent.Descending motion was formed over the South China Sea and the western Pacific Ocean.It also enhances the low-level anticyclonic circulation anomaly,which transports more water vapor from the ocean to the Yangtze River basin and facilitates the formation of meridional tripole precipitation pattern.In July,except the decrease of sea ice in BKS,the sea ice also decreases along the north coast of Russia in the East Siberian Sea.The decrease of sea ice produces a strong meridional triple wave pattern over the Eastern Eurasian continent and weakens the atmospheric baroclinicity over middle China,thus the monsoon circulation is weakened and produce a classical “southern China flood and northern China drought” pattern.(3)Multi-model large ensemble simulations from the PAMIP experiment are used to reveal the effects of Antarctic sea ice reduction on East Asian summer monsoon(EASM)precipitation variability and its possible physical processes.The decrease in Antarctic sea ice also plays an important role in the EASM precipitation anomaly,and its impact is even larger than the response caused by the decrease in Arctic sea ice.The precipitation in East Asia is a meridional dipole pattern in June when Antarctic sea ice decreases.The response of precipitation to the decrease of Antarctic sea ice in July shows the classic pattern of “southern China flood and northern China drought”.The precipitation pattern is also closely related to the regional differences in the decrease of Antarctic sea ice in June and July.In June,the Antarctic sea ice is reduced in the western hemisphere around the Antarctic continent(the South Atlantic and the eastern part of South Pacific),and the upper-level stationary waves are propagated from west to east over the western South Pacific and the Atlantic Ocean to the southeastern side of the African continent,then to South China by cross-equatorial flow.Descending motion is formed over the South China Sea and the western Pacific Ocean.It also enhances the low-level anticyclonic circulation anomaly,which transports more water vapor from the ocean to the Yangtze River basin and facilitates the formation of meridional dipole precipitation pattern.In July,in addition to the decrease in sea ice in the western hemisphere of the circum-Antarctic continent,sea ice in the eastern hemisphere of the circum-Antarctic continent is also decreasing.The decrease in sea ice produces a strong propagation of meridional fluctuations in the southern Indian Ocean,part of which spreads to the African continent and reaches eastern Asia via crossequatorial flow,resulting in precipitation anomalies and the classical “southern China flood and northern China drought” pattern.In addition,the reduction of the thermal gradient between the equator and the polar regions,which leads to the weakening of three-cell meridional circulation,provides a favorable background for precipitation anomalies.(4)The FGOALS-f3-L model was used to conduct a simultaneous bipolar sea ice reduction experiment to quantify the relative contributions of simultaneous bipolar sea ice reduction and unipolar sea ice reduction to EASM precipitation variability.The spatial pattern of the EASM precipitation anomalies caused by the simultaneous reduction of bipolar sea ice is similar to that of unipolar sea ice reduction,but it also differs in intensity and local area.The response of June precipitation to the reduction of bipolar sea ice is a meridional tripolar pattern.The response of July precipitation to the reduction of bipolar sea ice shows a southern flooding and northern drought pattern.For the central part of China(Yangtze River basin)in June,there seems to be an offsetting effect of the response of simultaneous reduction of bipolar sea ice relative to the response of unipolar sea ice reduction.For the northern region with less precipitation in July,the response of simultaneous reduction of bipolar sea ice is comparable to the sum of the response of unipolar sea ice reduction.There is some uncertainty in FGOALS-f3-L compared to the multi-model large ensemble average.The circulation response is strong overall,and the response of the circulation at mid-latitudes is generally consistent with the results of the multi-model ensemble average,but there are some uncertainties at low and high latitudes.Therefore,the response of precipitation to sea ice reduction simulated by the FGOALS f3-L large-sample ensemble is more robust in northern and central China(Yangtze River basin),while there are some biases and uncertainties in southern China.