The Link Between Extreme Low Temperature In Northern China And Arctic Warming And The Prediction Experiment Of Extreme Low Temperature

Under the background of global warming,the Arctic warming rate is at least twice the global warming rate.While warming has stagnated and extremely low temperature events have occurred frequently in northern China.The relationship between abnormal warming in the Arctic and stagnant warming/frequent extremely low temperature events in mid-latitude continents has attracted much attention.In this paper,using the grid data set(CN05.1)constructed by Chinese site observations and multiple reanalysis data,the relationship between the abnormal warming of the Barents-Kara Sea in winter and the frequency of extremely low temperatures in northern China and the associated mechanism were analyzed.Based on the pattern of a warm Arctic and cold Eurasia(WACE),two predictors of early autumn were searched,physics-based empirical models were built,and the frequency of extremely low temperatures in Northern China were predicted.The main conclusions are as follows:(1)The close relationship between the frequency of extremely low temperatures in northern China and the abnormal warming over the Barents and Kara Seas in winter was found.Through comparing the difference of probability density distribution of regional average in time scale greater than 90 days and less than 90 days temperature disturbance anomalies in Northern China between the warm years and the cold years over the Barents and Kara Seas,it is found that the warm years over the Barents and Kara Seas correspond to the greater amplitude of winter seasonal cycle in northern China,which is conducive to the occurrence of extremely low temperatures.(2)Physical mechanism between the frequency of extremely low temperatures in northern China and the abnormal warming over the Barents and Kara Seas in winter was revealed.Based on the analysis of composite differences between the warm years and the cold years over the Barents and Kara Seas,it is found that the positive geopotential height anomaly appears over Novaya Zemlya and Urals with the abnormal warming over the Barents and Kara Seas.Significant cold advection and sinking motion anomalies appear on the east and south sides of the positive geopotential height anomaly.Cold advection anomalies correspond to an increase in the conversion of average effective potential energy to low-frequency disturbance effective energy and sinking motion anomalies correspond to an increase in the conversion of low-frequency disturbance effective potential energy to low-frequency disturbance kinetic energy.Abnormal divergent flow appears near the ground,which makes Siberian high strengthen and extend to Southeast.The diagnosis of thermodynamic equation,further indicates that cold advection anomaly caused by northeast wind on the south side of the Siberian high pressure anomaly,ascending motion anomaly caused by easterly wind anomalies passing the terrain and abnormal diabatic cooling caused by radiation,sensible heat,and latent heat result in an increase in the amplitude of the near-surface temperature's seasonal cycle in mid-latitude Asia,which conducive to frequent occurrence of extremely low temperatures.(3)Based on the pattern of abnormal warming over the Barents and Kara seas and frequent extremely low temperature events in Northern China,predictors in the preceding autumn were searched and the frequency of extremely low temperatures in Northern China were predicted.The Eurasian Arctic sea ice Concentration in September and October(ARC-SIC)is selected as the first predictor,and the tropical eastern Pacific trend in autumn(TRO-DSST)is selected as the second predictor.Both predictors can influence the pattern of a warm Arctic and cold Eurasia in varying degrees.Based on two predictors and linear trend(TREND),multiple linear regression was used to establish prediction models.The prediction models were established by using data during the epochs of 1978-2007 and the effects of the models were verified by using data during the epochs of 2008-2017.Two factors of TREND,ARC-SIC and three factors of TREND,ARC-SIC and TRO-DSST were used to establish the prediction model,and the effect of independent sample test was evaluated.The results show that the two factors and three factors models can predict 32% and 36% of the total variance of the numbers of extremely cold days anomalies of regional average in Northern China(NECD),respectively.The prediction skills of the two models are good in North China and Xinjiang.Compared with the two factors model,the prediction skills of the three factors model is significantly improved in the west of Northeast China.