| Most of previous researches on the relationship between sea ice and climate were based on the sea ice data after year 1979 when satellite observations began.Due to the limited timespan,the relationship revealed by these observational dataset may be disturbed by atmospheric internal natural variability.As a result,there is uncertainty as for the relationship between the atmosphere and the sea ice.One way to reduce this uncertainty is to use longer sea ice data.In recent years,sea ice data has been extended to the period prior to 1979,the start era of satellite observations.But the record of sea ice before 1979 was discrete and not consecutive,the quality of the data is not yet known.In this study,two sets of sea ice data were evaluated,one from the UK Met Office's Hadley Centre(Hadley)and the other from National Sea Ice Data Center(NSIDC),NOAA.The Barents-Kara Seas(BKS)were focused on,based on their importance to the climate discovered by previous study.After determining the quality of sea ice data before 1979,the subsequent analysis was performed on using the data with a better quality.We studied the decadal variation of the relationship between the sea ice and climate.To sum up,the main contents of this study are as follows:(1)First,make a comparison of the two sets of sea ice data.For the sea ice changes in the winter BKS from 1958 to 2013,the two sets of data are generally consistent.When the timespan was divided into two sections,comparisons were made from1958 to 1978 and from 1979 to 2013,and it was found that there was still a good agreement after 1979.However,there are significant differences between the two sets of sea ice data before 1979.The correlation coefficient between the two sets of sea ice data was only 0.64.(2)To establish a sea ice proxy and evaluate the quality of the two sets of sea ice data based on it.Observation analysis shows the closely relationship between air temperature on the two islands near BKS(Novaya Zemlya and Franz Josef Land)and BKS sea ice change.Further analysis shows that this connection still exists in the historical experiments of the Coupled Model Intercomparison Project Phase 5(CMIP5).Therefore,this connection is physically reasonable.Thus,using the method of linear regression,based on the temperature observations of the two islands,a sea ice proxy can be constructed.Since the temperature has a longer observation sequence than sea ice,the constructed sea ice proxy has a longer sequence length than the actual sea ice observation.The evaluation results based on the proxy show that pre-satellite observations of NSIDC sea ice concentration data are more realistic and reliable than Hadley sea ice concentration data.(3)Based on NSIDC sea ice data,he interannual relationship between the autumn BKS sea ice and the late winter East Asia climate was analyzed in two periods after 1958,from 1958 to 1978 and from 1979 to 2013.It was found that there was an decadal change in the relationship between them.After 1979,when the BKS sea ice decreased in September,the temperature in some parts of East Asia decreased in winter.However,before 1979,when the BKS sea ice decreased in September,there was no significant temperature anomaly in East Asia during the winter.In addition,the circulation fields corresponding to the two time periods are also inconsistent.The interannual relationship between the sea ice and climate have the characteristics of decadal variation.(4)The differences between the winterss BKS sea ice and the winter East Asian climate in the interannual time scale in the two time periods were also compared and discussed.After 1979,when the BKS sea ice decreased in winter,the temperature in East Asia decreased during the winter.But before 1979,when the BKS sea ice decreased in winter,the temperatures in the middle and high latitudes of Eurasia increased during the winter.The conclusion is similar to(3),that is,the interannual relationship between them is characterized by a decadal shift. |
PDF Full Text Request