1. Introduction
The Arctic sea ice has experienced dramatic reduction during the last few decades. The observed sea ice decline has been partly attributed to anthropogenic forcing and associated positive feedbacks, collectively known as “Arctic amplification” (Deser et al. 2010; Screen and Simmonds 2010; Notz and Marotzke 2012; Notz and Stroeve 2016; Screen et al. 2018; England et al. 2021). In addition, the oceanic and atmospheric internal variabilities also act as key drivers for sea ice variability (e.g., Liu et al. 2004; Shimada et al. 2006; Wu et al. 2006; Ogi and Wallace 2007; Koenigk and Brodeau 2014; Zhang 2015; Ding et al. 2017, 2019; Wernli and Papritz 2018; Huang et al. 2019; Luo et al. 2021), and the atmospheric process accounts for 30%–50% of the observed sea ice decline since the 1980s (Kay et al. 2011; Swart et al. 2015; Ding et al. 2017, 2019).
In comparison with other seasons, Arctic amplification is relatively weak in summer (e.g., Manabe and Stouffer 1980; Serreze and Francis 2006; Kumar et al. 2010; Screen and Simmonds 2010; Dai et al. 2019), suggesting that the atmospheric internal variability may play a more important role in influencing sea ice. Previous studies proposed that a barotropic low-frequency high pressure (or anticyclonic) anomaly in the Arctic significantly affects the summer sea ice melting (Deser et al. 2000; Maslanik et al. 2007; Ogi and Wallace 2007; Kay et al. 2008; Ogi et al. 2008; Lim et al. 2019; Ding et al. 2017, 2019; Topál et al. 2020; Yu et al. 2021). Both thermodynamic and dynamic processes are important for linking the Arctic high pressure anomaly and sea ice decline. The tropospheric warming and moistening associated with the high pressure anomaly result in sea ice melting through emitting more downwelling longwave radiation (Ding et al. 2017, 2019; Topál et al. 2020; Luo et al. 2021). On the other hand, the anticyclonic anomaly may also affect sea ice cover by a dynamic ice drifting effect (Wu et al. 2006; Ogi and Wallace 2007; Rampal et al. 2011; Ogi et al. 2010; Grunseich and Wang 2016; Wang et al. 2020). Both the thermodynamic and dynamic processes associated with the high pressure anomaly are suggested to exist and operate on interannual and interdecadal time scales (Ogi and Wallace 2007; Wettstein and Deser 2014; Ding et al. 2017; Topál et al. 2020).
However, summertime circulation anomalies associated with Arctic sea ice show a large diversity and many kinds of atmospheric circulation patterns have been suggested to be associated with the Arctic high pressure anomaly. For instance, some studies highlighted the role of the Greenland blocking high (GB)—characterized by positive geopotential height anomalies above Greenland and the Arctic Ocean (Hanna et al. 2016; Ding et al. 2017, 2019; Topál et al. 2020)—in affecting the summer sea ice decline. Some studies emphasized the role of the positive phase of the Arctic dipole anomaly (AD; featuring positive sea level pressure anomalies over the North America side and negative anomalies over the Siberia side of the Arctic; Wu et al. 2006; Wang et al. 2009; Overland et al. 2012; Heo et al. 2021) and the negative phase of the Arctic Oscillation (AO; featuring high pressure anomalies in the Arctic; Maslanik et al. 2007; Ogi and Wallace 2007; Ogi et al. 2008), in affecting the sea ice decline. The results of Serreze et al. (2016) even suggested that the anticyclonic anomaly differs markedly in the magnitude and location for each abnormal low sea ice extent year.
The diversity in circulation patterns associated with Arctic sea ice leads to different dominance of thermodynamic and dynamic processes. For instance, the thermodynamic process is suggested to be important for linking the GB and sea ice decline (Ding et al. 2017, 2019; Topál et al. 2020; Luo et al. 2021), while for AD and AO, the wind-driven effect is more emphasized (Wu et al. 2006; Maslanik et al. 2007; Ogi and Wallace 2007; Ogi et al. 2008; Wang et al. 2009; Overland et al. 2012). We know that the summer sea ice shows both interannual variability and long-term variation (Kay et al. 2011; Wernli and Papritz 2018; Choi et al. 2019; Liu et al. 2021); however, there is a lack of literature attributing the different processes to different time scales. In this study, we aim to separate the interannual variability and decadal change of summer Arctic sea ice and shed light on the differences in circulation–sea ice connections between the two different time scales.
The rest of this paper is arranged as follows. Section 2 describes the dataset and definitions associated with Arctic sea ice extent. Section 3 shows the large-scale circulation anomalies associated with the interannual variability and decadal change of summer sea ice. Section 4 and section 5 investigate the thermodynamic process and wind-driven sea ice drift associated with circulation–sea ice connections on interannual and decadal time scales, respectively. Section 6 provides a summary and discussion.
2. Data and definitions associated with Arctic sea ice extent
a. Data
1) Sea ice
This study uses the monthly mean Arctic sea ice observation from the National Snow and Ice Data Center (NSIDC; Cavalieri et al. 1996). The sea ice extent is the direct output from version 3 of the NSIDC. The sea ice concentration is derived from Special Sensor Microwave Imager/Sounder (SSMIS) on board the Defense Meteorological Satellite Program (DMSP) satellites using the NASA team algorithm (Comiso et al. 2017) and is provided on the Equal-Area Scalable Earth (EASE)-Grid with a spatial resolution of 25 km. The monthly sea ice motion data are from the similar 25-km Polar Pathfinder EASE-Grid archived by NSIDC, which is retrieved from Advanced Microwave Scanning Radiometer (AMSR)-E and AMSR-2 passive microwave (Tschudi et al. 2020). In addition, we also use the sea ice motion data from the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS; Zhang and Rothrock 2003) to confirm the results.
2) Reanalysis data
The monthly mean reanalysis data from the fifth-generation European Centre for Medium-Range Weather Forecast (ECMWF) reanalysis (ERA5; Hersbach et al. 2020; 0.5° × 0.5° horizontal resolution) are used. The daily ERA5 data are used for calculating horizontal advection of moisture and temperature. Although it has been shown by Ding et al. (2017) that the ERA-Interim reanalysis (previous version of ERA5) reliably represents the radiosonde measurements in and around the Arctic, we attempt to reduce the uncertainty associated with reanalysis data in the Arctic (e.g., Sedlar 2018) as follows. First, the National Centers for Environmental Prediction Reanalysis 2 (NCEP2; Kanamitsu et al. 2002; 2.5° × 2.5° horizontal resolution) and the Japanese 55-Year Reanalysis (JRA-55; Kobayashi et al. 2015; 1.25° × 1.25° horizontal resolution) datasets are used to assess the robustness of radiation, temperature, specific humidity, and circulation anomalies. The NCEP2 and JRA-55 results are basically similar to the ERA5 results and are briefly discussed in this study. Second, satellite-based radiation taken from the International Satellite Cloud Climatology Project (ISCCP; Rossow and Schiffer 1999) are used to confirm the radiation anomalies in ERA5. The analyzed time period is from 1979 to 2019, except that for ISCCP, which is available from 1984 to 2016. The June–August (JJA) average is used to represent the summer season in this study.
3) Climate indices
The AO index is obtained from https://psl.noaa.gov/gcos_wgsp/Timeseries/AO/ and GB index is from https://psl.noaa.gov/gcos_wgsp/Timeseries/GBI_UL/. The AD index is defined as the time series of second EOF mode of JJA-mean sea level pressure over 70°–90°N, 0°–360° (Wu et al. 2006; Wang et al. 2009; Overland and Wang 2010; Overland et al. 2012). In addition, we use the Pacific decadal oscillation (PDO) and the Atlantic multidecadal oscillation (AMO) indices from https://psl.noaa.gov/gcos_wgsp/Timeseries/ to discuss the possible role of these oceanic modes in affecting the sea ice variability.
4) CESM-LE
As a supplementary discussion, the Community Earth System Model version 1 Large Ensemble (CESM-LE; Kay et al. 2015; Deser et al. 2020) is employed to analyze the contribution of internal variability and external forcing to sea ice–circulation connections. It includes 40 ensemble members for 1920–2100 differing by round-off level perturbations in their initial conditions. The model output is generated under historical forcing for the years preceding 2006 and RCP8.5 forcing after 2006. The analyzed period is also 1979–2019.
b. Definitions associated with Arctic sea ice extent
To describe the sea ice change in summer, we define a sea ice change index (SI) as Arctic sea ice extent anomalies between May and September, which highlights the sea ice melting in summer. Thus, a positive (negative) SI represents more (less) sea ice loss in summer. We have also defined SI by using the difference between June and September and obtained similar results (not shown), suggesting that the results are not sensitive to the exact start date of summer. We focus on the effect of JJA-mean horizontal wind anomalies in the lower troposphere, considering that the temperature and moisture anomalies associated with the thermodynamic process mainly appear in the lower troposphere and sea ice drift is more affected by low-level winds.
Figure 1a shows the time series of the SI. The SI shows both year-to-year variability and a regime shift, characterized by more sea ice melting in recent decades. We identify the regime shift of SI by using the sequential regime shift detection algorithm developed by Rodionov (2004, 2006), where a probability level of 0.05, a cutoff length of 10 years, and a Huber’s weight parameter of 1.0 were set. It is found that the regime shift occurred in 2006/07, consistent with a number of previous studies (e.g., Livina and Lenton 2013; Wang and Liu 2016; Francis and Wu 2020; Wu and Li 2021). The SI increased by 1.33 km2 from 1979–2006 to 2007–19, which is statistically significant at the 99% confidence level. We extract the interannual component of SI (hereafter SI-I) by subtracting the averages during 1979–2006 and 2007–19 from the original time series of SI, such as subtracting the 1979–2006 average for the year of 2006 and the 2007–19 average for the year of 2007 (Fig. 1b), and the difference in averages between 2007–19 and 1979–2006 is considered as the decadal component of SI (hereafter SI-D), which highlights the contrasts between these two periods. This method of obtaining SI-I is referred to as the subtraction approach. The SI-I accounts for 39% of the total variance of the SI: the variance of SI is 0.66 km4, and that of SI-I is 0.26 km4. Therefore, the interannual and decadal components both make a large contribution to the variability of SI.
An alternative approach to obtain the interannual component is by removing the 9-yr running mean from the original SI, which separates the variability greater and smaller than 9 years. This way of separating the interannual and decadal components is referred to as the smoothing approach. It is notable that choosing an 11- or 13-yr running mean would obtain similar results (not shown). However, the interannual variance for the subtraction approach is smaller than that for the smoothing approach during 1983–2015 (0.28 vs 0.30 km4), which suggests that the subtraction approach can remove decadal variations more effectively and the increase of SI in recent decades looks more like an abrupt change. Similarly, for all other variables, the interannual components are also obtained by the subtraction approach. In the following, the anomalies associated with SI-I are obtained by regressing the interannual component of variables against the normalized SI-I and those associated with SI-D are simply defined as the difference in averages between 2007–19 and 1979–2006. The correlations with SI-I refer to the interannual component of variables and the correlations with other indices refer to the raw data, unless otherwise stated.
It is notable that in CESM-LE, the sea ice area is calculated as the product of sea ice concentration and grid element area in each sea ice grid, and then the total sea ice extent in each month is defined as the sum of sea ice area in all Arctic grid cells where the ice concentration is equal to or larger than 15%. The SI also refers to the anomalies of May minus September sea ice extent. The interannual (SI-I) and decadal (SI-D) components are obtained by the smoothing approach, as the turning points in each member are different.
3. Large-scale circulation anomalies associated with SI-I and SI-D
Figure 2 shows the May minus September sea ice concentration, the JJA-mean 850-hPa horizontal wind and sea level pressure anomalies 1) regressed against the normalized SI-I and 2) associated with SI-D, respectively. Here, the anomalies associated with SI-D are presented by the difference in averages between 2007–19 and 1979–2006, as mentioned in section 2. The sea ice concentration anomalies associated with the SI-I and SI-D show a similar pattern (Figs. 2a,b); that is, there is significant positive sea ice melting over almost the whole Arctic, which acts as a precondition for the following comparisons. However, the circulation anomalies associated with SI-I and SI-D are distinctly different (Figs. 2c,d). For SI-I (Fig. 2c), there is an anticyclonic anomaly centered over the central Arctic, which resembles the AO (e.g., Ogi and Wallace 2007). Associated with this anticyclonic anomaly, there are significantly circumpolar easterly anomalies over the marginal seas. By contrast, for SI-D (Fig. 2d), the circulation anomalies are characterized by an anticyclonic anomaly over Greenland and a cyclonic anomaly over the coastal northern Siberia, which resemble the AD (Overland et al. 2012; Wettstein and Deser 2014; Serreze et al. 2016; Wu and Li 2021). Associated with this pair of anticyclonic and cyclonic anomalies, there are significant transpolar circulation anomalies from the eastern Siberian Sea, through the central Arctic Ocean, to the northern Atlantic. In addition, the circulation anomalies are barotropic in the troposphere for both SI-I and SI-D (i.e., the 200-hPa circulation anomalies show similar patterns to those at 850 hPa; see Fig. S1 in the online supplemental material).
Figure 3 shows the time series of the normalized CPI and TPI. The CPI shows clear interannual variation, but no distinct decadal variation (Fig. 3a). The decadal component of CPI, which is defined as the 9-yr running mean of the original CPI, accounts for only 11% of the total variance of CPI during 1983–2015. The correlation coefficient between CPI and SI-I is 0.49 (Table 1), which is statistically significant at the 99% confidence level. This correlation coefficient is slightly greater than that between the CPI and SI (0.44), suggesting that the circumpolar easterly anomalies mainly connect to the interannual component of summer sea ice melting. By contrast, TPI shows both clear interannual and decadal variations (Fig. 3b). The TPI significantly increased since the mid-2000s, and the regime shift (Rodionov 2004, 2006) occurred at 2004/05, almost simultaneously with the increase of summer sea ice melting (Fig. 1a). The average of the original TPI is only −0.01 m s−1 during 1979–2006, but 1.40 m s−1 during 2007–19. In addition, the correlation coefficient between the TPI and SI is 0.51 during 1979–2019, but only 0.17 between the interannual component of TPI and SI-I. Here, the interannual component of TPI is obtained via the subtraction approach, namely by subtracting the averages during 1979–2006 and 2007–19 from the original time series of TPI, respectively. These results indicate that the transpolar wind anomalies mainly connect to the decadal change of sea ice, but not the interannual variability. In addition, the correlation coefficient between CPI and TPI is only 0.20, suggesting that these two indices are largely independent of each other.
Correlation coefficients between sea ice and circulation indices during 1979–2019. Boldface numbers are significant at the 95% confidence level based on Student’s t test. The circulation indices (except CPI) have extracted the interannual component when calculating their correlations with SI-I.
In the rest of this paper, we compare the anomalies associated with CPI (TPI) to those associated with SI-I (SI-D) for the following purposes: 1) to demonstrate the connection between circulation anomalies and sea ice variabilities and 2) to make it clearer to what extent the circulation could explain the SI-I and SI-D.
Figure 4 shows the May minus September sea ice concentration, JJA-mean 850-hPa circulation, and sea level pressure anomalies regressed against the normalized CPI and TPI, respectively. The CPI- and TPI-related sea ice concentration anomalies are similar to those associated with SI-I and SI-D, respectively (Figs. 2a and 4a vs Figs. 2b and 4b). There are positive anomalies of sea ice change over the marginal seas for both CPI and TPI, consistent with the above-mentioned significant correlation coefficients between these circulation and sea ice indices.
The circulation anomalies associated with CPI exhibit as a significant anticyclonic anomaly over the central Arctic (Fig. 4c). Associated with this anticyclonic anomaly, there are circumpolar easterly anomalies in the marginal seas, consistent with those associated with SI-I (Fig. 2c). For TPI, the circulation anomalies are characterized by an anticyclonic anomaly over Greenland and a cyclonic anomaly over northern Siberia, and there are significantly transpolar circulation anomalies between (Fig. 4d), consistent with the circulation anomalies associated with SI-D (Fig. 2d). These results suggest that the definitions of CPI and TPI are reasonable.
The CPI and TPI show a close correlation with AO and AD, respectively (Table 1). The correlation coefficient between CPI and the JJA-mean AO index is −0.73, and that between TPI and the AD index is −0.84 during 1979–2019. However, CPI and TPI are more closely associated with the variability of sea ice melting than the AO and AD indices, respectively. The correlation coefficient between the AO index and SI-I is −0.27, weaker than that between CPI and SI-I (0.49), and the correlation coefficient between the AD index and SI is −0.39, also weaker than that between TPI and SI (0.51). Thus, CPI and TPI can better describe the circulation anomalies associated with sea ice change in summer. In addition, CPI and TPI both show a significant correlation with the GB (Table 1), which has been considered as an important factor affecting the Arctic sea ice on both interannual and interdecadal time scales (e.g., Ding et al. 2017; Lim et al. 2019; Luo et al. 2021). By separating the interannual variability and decadal change, the features of atmospheric circulation patterns associated with sea ice melting on different time scales are much clearer.
4. Thermodynamic process
a. Role of downwelling longwave radiation in connecting sea ice and circulation anomalies
This section investigates the role of the thermodynamic process in connecting the large-scale circulation anomalies and summer sea ice change on interannual and decadal time scales. Figure 5 shows the JJA-mean downwelling longwave radiation (DLR) anomalies 1) regressed against the normalized SI-I and 2) associated with SI-D, respectively. To confirm the ERA5 results, the DLR anomalies obtained by ISCCP observation are also shown. In ERA5, the DLR anomalies contribute to sea ice melting, for both SI-I and SI-D (Figs. 5a,b). The correlation coefficient between the interannual component of the Arctic DLR, which is averaged over 70°–90°N, 0°–360°, and SI-I is 0.59 (Table 2). The difference in averages of the Arctic DLR between 2007–19 and 1979–2006 is 2.82 W m−2. The correlation coefficient and the DLR difference between the two periods are both statistically significant at the 99% confidence level. However, the spatial patterns of DLR anomalies show different features between SI-I and SI-D. For SI-I (Fig. 5a), positive DLR anomalies mainly appear over the central Arctic; for SI-D (Fig. 5b), positive anomalies tend to appear over marginal seas, especially north of Greenland and the Canadian Archipelago, and the coastal area of Siberia. The DLR anomalies in ISCCP observation are consistent with ERA5 (Figs. 5c,d), confirming the positive contribution of DLR anomalies to both SI-I and SI-D. In addition, the JRA-55 and NCEP2 reanalysis datasets also obtain similar results (Fig. S2).
Correlation coefficients between sea ice or circulation indices and the Arctic thermodynamic indices during 1979–2019. Boldface numbers are significant at the 95% confidence level based on Student’s t test. The Arctic DLR, 850-hPa temperature (T850), and specific humidity (S850) indices are defined as anomalies averaged over 70°–90°N, 0°–360°. The indices have extracted the interannual component when calculating their correlations with SI-I.
We have also examined the contribution of upwelling longwave radiation (ULR), downwelling shortwave radiation (DSR), upwelling shortwave radiation (USR), downwelling sensible heat flux (SH), and latent heat flux (LH) to SI-I and SI-D (Figs. S3–S5). It was found that there are significant positive ULR anomalies in the Arctic associated with SI-I and SI-D (Figs. S3c,d), suggesting that the ULR contributes negatively to both SI-I and SI-D. On the other hand, the net longwave radiation (i.e., the DLR minus ULR) is positive (Figs. S3e,f); the DSR also contributes negatively to both SI-I and SI-D (Figs. S4a,b), but the USR anomalies over the Arctic are strong and negative (Figs. S4c,d), suggesting that the USR anomalies contribute positively to SI-I and SI-D (Text S1). The strong and negative USR anomalies lead to positive net shortwave radiation anomalies, for both SI-I and SI-D (Figs. S4e,f). However, the USR is more likely due to the feedback of sea ice reduction (i.e., USR would be decreased as the lower albedo associated with the sea ice decline). Therefore, we mainly focus on the contribution of DLR to sea ice change. In addition, the SH and LH anomalies associated with the SI-I and SI-D are both much smaller than the longwave and shortwave radiation anomalies (Fig. S5), and thus these anomalies are not discussed.
Figure 6 shows the DLR anomalies regressed against the normalized CPI and TPI, respectively, in ERA5 and ISCCP. The results in ERA5 and ISCCP are quite similar for both SI-I and SI-D (Figs. 6a,c; Figs. 6b,d). For CPI, there are significantly positive anomalies over the central Arctic (Figs. 6a,c), which resemble those associated with SI-I (Figs. 5a,c). The correlation coefficient between CPI and the Arctic DLR is 0.45 in ERA5 (Table 2), statistically significant at the 99% confidence level. By contrast, for TPI (Figs. 6b,d), there are also positive anomalies in the Arctic, but large values tend to appear over north of Greenland and Canadian Archipelago and parts of the coastal area of Siberia. These anomalies resemble those associated with SI-D (Figs. 5b,d). The correlation coefficient between TPI and the Arctic DLR is 0.30 in ERA5, which is marginally significant (Table 2). These results suggest that the DLR has a role in linking large-scale circulation and sea ice anomalies on the decadal time scale, but the role is not as significant as that on the interannual time scale.
b. Contribution of temperature and moisture to DLR anomalies
Figure 7 shows the 850-hPa temperature and specific humidity anomalies 1) regressed against the normalized SI-I and 2) associated with SI-D, respectively. For SI-I, the temperature and moisture both show positive anomalies over the central Arctic (Figs. 7a,c), which are favorable for the positive DLR anomalies there (Fig. 5a). The correlation coefficient is 0.83 and 0.90, respectively, between the interannual component of DLR and 850-hPa Arctic temperature anomalies, and between the interannual component of DLR and specific humidity anomalies. In addition, there is a close relationship between the interannual component of moisture and temperature anomalies, with a correlation coefficient of being 0.81. This high correlation coefficient can be explained as follows: the moisture, as a kind of greenhouse gas, favors higher temperature when it is higher than normal. And the increased temperature also favors increased moisture according to the Clausius–Clapeyron relation. We also analyzed the vertically integrated (1000–300 hPa) moisture and temperature anomalies and obtained similar results (Fig. S6).
For SI-D, there are increased temperature and moisture in the Arctic (Figs. 7b,d), suggesting that the temperature and moisture anomalies both favor the increased DLR (Fig. 5b). However, the anomalies show different features from those associated with SI-I. First, largest positive temperature anomalies do not locate in the central Arctic, but over the Canadian Arctic and Baffin Bay and in eastern Siberia (Fig. 7b). Second, large moisture anomalies are also located over Baffin Bay and eastern Siberia rather than the central Arctic, but barely significant (Fig. 7d). These results confirm that the temperature and moisture and the resultant DLR anomalies more significantly affect SI-I than SI-D.
Figure 8 shows the 850-hPa temperature and specific humidity anomalies regressed against the normalized CPI and TPI, respectively. For CPI, there are significant positive temperature and moisture anomalies over the central Arctic (Figs. 8a,c), which contribute to the enhanced DLR there (Fig. 6a). The correlation coefficient between CPI and 850-hPa Arctic temperature anomalies is 0.62 (Table 2), which is significant at the 99% confidence level. The correlation coefficient between CPI and the Arctic specific humidity is also significant (0.31), although not as strong as that between CPI and temperature. In addition, the temperature and moisture anomalies associated with CPI resemble those associated with SI-I (left-hand panels of Figs. 7 and 8). These results confirm that the thermodynamic process is important for connecting the circumpolar circulation anomalies and sea ice change on interannual time scale.
For TPI, significantly positive temperature and moisture anomalies appear over Canadian Arctic (Figs. 8b,d), consistent with those associated with SI-D (Figs. 7b,d). The correlation coefficients between the Arctic temperature or moisture anomalies and TPI are barely significant at the 95% confidence level (Table 2), consistent with that between DLR and TPI. There results confirm that the thermodynamic process on a decadal time scale is not as significant as that on an interannual time scale. In addition, it is also notable that the temperature or moisture anomalies associated with TPI may not only act as a result of circulation anomalies, but could also be affected by external forcing, such as human activities.
In addition to temperature and moisture, anomalous clouds also affect the DLR anomalies. It is suggested that more clouds tend to contribute to increased DLR, since the infrared atmospheric window is “closed” (e.g., Takara and Ellingson 2000), and vice versa. However, there is a large uncertainty of cloud amount in both reanalysis and observation datasets (see Text S2 and Fig. S9 in the online supplemental material), which makes it hard to access to the exact role of clouds in linking the sea ice and circulation anomalies.
c. Origin of temperature and moisture anomalies
We explored the origin of temperature and moisture anomalies associated with SI-I and CPI. The contributions of horizontal and vertical advection are mainly considered. The horizontal and vertical advections in the temperature budget equation can be expressed as −V ⋅ ∇T and
Figure 9 shows the correlation coefficients between the horizontal and vertical advection of 850-hPa temperature and moisture, averaged over 70°–90°N, 0°–360°, and SI-I and CPI. For temperature anomalies (Fig. 9a), the anomalous descending motion associated with the central Arctic anticyclonic anomaly contributes positively to low-level warming. The correlation coefficient between vertical advection and CPI and SI-I are 0.71 and 0.48, respectively, both of which are statistically significant at the 99% confidence level. The horizontal advection tends to contribute negatively to temperature anomalies, and the correlation coefficients between the Arctic horizontal advection and CPI and SI-I are −0.25 and −0.27, respectively.
For moisture anomalies, the descending motion associated with the anticyclonic anomaly makes a negative contribution (Fig. 9b). The correlation coefficients between vertical advection of moisture in the Arctic and CPI and SI-I are −0.75 and −0.46, respectively. On the other hand, the horizontal advection tends to contribute positively to enhanced moisture, but in marginal significance. The correlation coefficients between horizontal advection in the Arctic and CPI and SI-I are 0.33 and 0.17, respectively. The vertically integrated (1000–300 hPa) results suggest that the horizontal advection and vertical advection contribute negatively to moisture, for both CPI and SI-I (Fig. S7). Therefore, it can be inferred that the enhanced moisture comes from enhanced evaporation or reduced liquid phase of water (i.e., clouds or precipitation). The evaporation could be enhanced by the easterly anomalies, as in climatological state, there are weak surface easterlies in the Arctic marginal seas and the easterly anomalies associated with SI-I or CPI enhance the wind speed there (Fig. S8). There could also be reduced liquid phase of water, as the descending adiabatic warming associated with the anticyclonic anomaly may induced less clouds or precipitations. However, reanalysis datasets and observations show a considerable uncertainty in depicting evaporation and clouds/precipitation (Figs. S9 and S10), resulting in a challenge of accessing the exact origin of moisture.
5. Wind-driven sea ice drift
In this section, we compare the wind-driven sea ice drift associated with SI-I and SI-D. Figure 10 shows the anomalous sea ice motion vectors 1) regressed against the normalized SI-I and 2) associated with SI-D, respectively, in NSIDC and PIOMAS. To facilitate description, significant 10-m horizontal wind anomalies are also shown. The results in NSIDC and PIOMAS are quite similar (Figs. 10a,c vs Figs. 10b,d). The anomalous sea ice motion vectors closely coincide with surface wind anomalies, for both SI-I and SI-D. For SI-I (Figs. 10a,c), sea ice is characterized by anomalous circumpolar movement. Compared with the surface easterly anomalies in the central Arctic, sea ice shows a more northward movement, suggesting that the surface easterlies could contribute to the reduction of sea ice in marginal seas by way of the Ekman drift. Moreover, the surface easterly anomalies promote transport of ice away from the Canadian Archipelago, which also favors reduction of sea ice there. In addition, the anomalous easterly winds enhance the background easterlies and thus may increase the deformation rate of Arctic sea ice. On the other hand, these sea ice drift anomalies induced by surface wind are in marginal significance.
By contrast, for SI-D (Figs. 10b,d), the sea ice drift is strong and more meridionally distributed. First, strong and significant offshore drift of sea ice occurs in the coasts of eastern Siberia and Alaska toward the North Pole, resulting in sea ice retreat in these marginal seas. Second, the transpolar wind anomalies also enhance transport of sea ice out of the Arctic Ocean and into the North Atlantic through the Fram Strait, which is an important gate of sea ice outflow (e.g., Kwok et al. 2004). The transpolar wind anomalies could also result in a decline in sea ice thickness (Fig. S11; Labe et al. 2018). As thinner sea ice has much higher growth rate than thicker ice under freezing conditions, the ice growth shows a relatively strong increase after 2007 (Zhang 2021), which may act as negative feedback to sea ice concentration decline. In addition, there is also offshore drift of ice in Canadian Archipelago, consistent with the sea ice retreat there (Fig. 2b). These results indicate that the wind-driven movement of sea ice dominates the decadal change of summer sea ice retreat.
Figure 11 shows the anomalous sea ice motion vectors regressed against the normalized CPI and TPI, respectively, in NSIDC and PIOMAS. The significant 10-m horizontal wind anomalies are also shown to facilitate description. Results in NSIDC and PIOMAS are consistent with each other (Figs. 11a,c vs Figs. 11b,d). The sea ice movement associated with CPI agrees well with that associated with SI-I (e.g., Figs. 10a and 11a), except that the CPI-related circumpolar movement is more significant. The sea ice motion vectors for TPI also agree well with those for SI-D, but tend to be more statistically significant (e.g., Figs. 10b and 11b). In addition, the sea ice movement is consistent with the surface wind anomalies, with the ice motion vectors turning over the right-hand side of the wind vectors. These results further demonstrate that the wind-driven advection of sea ice more dominantly affects sea ice retreat on a decadal time scale than that on an interannual time scale.
6. Discussion and conclusions
a. Discussion
This study focuses on the connection between atmospheric circulation anomalies and sea ice variability. In addition to the atmospheric internal variability, oceanic internal drivers and external forcing also exert important influences on sea ice variability. The CESM-LE simulation indicates that the internal variability can induce the summer sea ice decline on both interannual and decadal time scales (Figs. 12a,b) and the external forcing exerts much more significant influences on SI-D than SI-I (Figs. 12c,d). The circulation anomalies associated with SI-I (i.e., the circumpolar circulation anomalies) can be well reproduced by internal variability (Fig. 13a). The thermodynamic anomalies on interannual time scale are also well simulated (Figs. S12–S14), consistent with early studies (Ding et al. 2019; Topál et al. 2020). However, the internal variability and external forcing, as well as their combination, cannot reproduce the circulation anomalies associated with SI-D (i.e., the transpolar circulation anomalies; right-hand panels of Fig. 13).
Previous studies have suggested that the oceanic internal processes, such as the AMO, PDO, or Atlantic/Pacific heat transport into the Arctic, can drive low-frequency Arctic sea ice variability (e.g., Day et al. 2012; Ding et al. 2014; Zhang 2015; Screen and Francis 2016; Tokinaga et al. 2017; Castruccio et al. 2019). Therefore, it is possible that the oceanic drivers modulate the atmospheric circulation and sea ice relationship on decadal time scale. As shown in Table 3, the SI and TPI both show a significant relationship with AMO and PDO. Unfortunately, these sea surface temperature modes cannot be reproduced in CESM-LE simulations (Baxter et al. 2019; Ding et al. 2019), which probably results in the absence of observed transpolar circulation anomalies in CESM-LE. These results suggest that more effort should be put to model improvement to correctly reproduce the observed circulation–sea ice connection on a decadal time scale.
Correlation coefficients between SI/TPI and 9-yr running mean of PDO/AMO indices during 1979–2019. Boldface numbers are significant at the 95% confidence level based on Student’s t test.
b. Conclusions
Large-scale circulation anomalies exert an important influence on both the interannual variability and long-term trend of Arctic sea ice extent in summer. In this study, we separate the interannual variability and decadal change of summer sea ice, characterized by the difference in sea ice extent between May and September, and focus on the contrasts in circulation–sea ice couplings between the two different time scales.
It is found that the circulation anomalies associated with the interannual component of summer sea ice decline are characterized by a barotropic anticyclonic anomaly in the central Arctic, with circumpolar easterly anomalies in the Arctic. The thermodynamic process is crucial for the circulation–sea ice coupling: on one hand, the descending motion associated with the central Arctic anticyclonic anomaly induces adiabatic warming in low levels, and favors sea ice decline by enhancing the downwelling longwave radiation. On the other hand, the anticyclonic anomaly favors increased moisture in low levels, possibly by enhancing evaporation or reducing liquid phase of water, while the exact process cannot be determined due to data uncertainties. The moisture and temperature enhance each other, and they both further contribute to sea ice melting by emitting more downwelling longwave radiation over sea ice. In addition, the circumpolar easterly anomalies could also cause sea ice extent reduction by way of Ekman drift in marginal seas.
By contrast, the circulation anomalies associated with the decadal change of summer sea ice decline are distinctly different and the circulation–sea ice connection involves in different dominant processes. The circulation anomalies are characterized by an anticyclonic anomaly over Greenland and a cyclonic anomaly over northern Siberia, which resemble the AD (Wu et al. 2006; Wang et al. 2009; Overland et al. 2012; Wettstein and Deser 2014). The wind-driven transport of sea ice significantly contributes to the circulation–sea ice connection: the transpolar circulation anomalies between the anticyclonic and cyclonic anomalies force ice drift away from the coasts of Siberia and Alaska toward the North Pole, and also transport sea ice out of the Arctic Ocean to the North Atlantic through the Fram Strait. In addition, these circulation anomalies also link to sea ice melting through enhancing the downwelling longwave radiation, while the corresponding thermodynamic process is not as significant as that on an interannual time scale.
The present results suggest that the dominant processes in circulation–sea ice connections on the interannual and decadal time scales are distinctly different, and it is necessary to distinguish these two kinds of variabilities in future studies. On the other hand, it is also notable that the sea ice decline could also, in turn, affect circulation anomalies (Screen et al. 2013; Liu et al. 2016), but the response of circulation patterns to sea ice decline shows large discrepancies in different models (Screen et al. 2018). The feedback of sea ice decline to circulation anomalies deserves further investigation.
Acknowledgments.
We thank the anonymous reviewers for their constructive comments, which greatly helped to improve this paper. This work was supported by the Natural Science Foundation of Jiangsu Province (Grant BK20190500), the National Key R&D Program of China (2018YFA0605901), the Fundamental Research Funds for the Central Universities (B220202049), and the National Natural Science Foundation of China (Grants 41905055 and 41721004).
Data availability statement.
All the data analyzed in this study are openly available. The NSIDC sea ice extent is obtained from https://nsidc.org/data/G02135/versions/3, sea ice concentration from https://nsidc.org/data/nsidc-0051, and sea ice motion data from https://daacdata.apps.nsidc.org/pub/DATASETS/nsidc0116_icemotion_vectors_v4/. The ERA5 reanalysis data are retrieved from https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-single-levels-monthly-means?tab=overview and https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-pressure-levels-monthly-means?tab=overview, respectively. The NCEP2 data are retrieved from the data portal at https://psl.noaa.gov/data/gridded/data.ncep.reanalysis2.html and the JRA-55 at https://jra.kishou.go.jp/.JRA-55/index_en.html. The PIOMAS sea ice motion data are from http://psc.apl.washington.edu/zhang/IDAO/data_piomas.html. The ISCCP radiation data are downloaded from https://isccp.giss.nasa.gov/pub/flux-fh/. The AO, AMO, and PDO indices are from https://psl.noaa.gov/gcos_wgsp/Timeseries/. The GB index is from https://psl.noaa.gov/gcos_wgsp/Timeseries/GBI_UL/. The CESM-LE simulations are obtained from https://www.cesm.ucar.edu/projects/community-projects/LENS/data-sets.html.
REFERENCES
Baxter, I., and Coauthors, 2019: How tropical Pacific surface cooling contributed to accelerated sea ice melt from 2007 to 2012 as ice is thinned by anthropogenic forcing. J. Climate, 32, 8583–8602, https://doi.org/10.1175/JCLI-D-18-0783.1.
Castruccio, F. S., Y. Ruprich-Robert, S. G. Yeager, G. Danabasoglu, R. Msadek, and T. L. Delworth, 2019: Modulation of Arctic sea ice loss by atmospheric teleconnections from Atlantic multidecadal variability. J. Climate, 32, 1419–1441, https://doi.org/10.1175/JCLI-D-18-0307.1.
Cavalieri, D. J., C. L. Parkinson, P. Gloersen, and H. J. Zwally, 1996: Sea ice concentrations from Nimbus-7 SMMR and DMSP SSM/I-SSMIS passive microwave data, version 1 (subset used: V3-NH). National Snow and Ice Data Center Distributed Active Archive Center, https://doi.org/10.5067/8GQ8LZQVL0VL.
Choi, N., K.-M. Kim, Y.-K. Lim, and M.-I. Lee, 2019: Decadal changes in the leading patterns of sea level pressure in the Arctic and their impacts on the sea ice variability in boreal summer. Cryosphere, 13, 3007–3021, https://doi.org/10.5194/tc-13-3007-2019.
Comiso, J. C., W. N. Meier, and R. Gersten, 2017: Variability and trends in the Arctic sea ice cover: Results from different techniques. J. Geophys. Res. Oceans, 122, 6883–6900, https://doi.org/10.1002/2017JC012768.
Dai, A., D. Luo, M. Song, and J. Liu, 2019: Arctic amplification is caused by sea-ice loss under increasing CO2. Nat. Commun., 10, 121, https://doi.org/10.1038/s41467-018-07954-9.
Day, J. J., J. C. Hargreaves, J. D. Annan, and A. Abe-Ouchi, 2012: Sources of multi-decadal variability in Arctic sea ice extent. Environ. Res. Lett., 7, 034011, https://doi.org/10.1088/1748-9326/7/3/034011.
Deser, C., J. Walsh, and M. Timlin, 2000: Arctic sea ice variability in the context of recent atmospheric circulation trends. J. Climate, 13, 617–633, https://doi.org/10.1175/1520-0442(2000)013<0617:ASIVIT>2.0.CO;2.
Deser, C., R. Tomas, M. Alexander, and D. Lawrence, 2010: The seasonal atmospheric response to projected Arctic sea ice loss in the late twenty-first century. J. Climate, 23, 333–351, https://doi.org/10.1175/2009JCLI3053.1.
Deser, C., and Coauthors, 2020: Insights from Earth system model initial-condition large ensembles and future prospects. Nat. Climate Change, 10, 277–286, https://doi.org/10.1038/s41558-020-0731-2.
Ding, Q., J. M. Wallace, D. S. Battisti, E. J. Steig, A. J. Gallant, H. J. Kim, and L. Geng, 2014: Tropical forcing of the recent rapid Arctic warming in northeastern Canada and Greenland. Nature, 509, 209–212, https://doi.org/10.1038/nature13260.
Ding, Q., and Coauthors, 2017: Influence of high-latitude atmospheric circulation changes on summertime Arctic sea ice. Nat. Climate Change, 7, 289–295, https://doi.org/10.1038/nclimate3241.
Ding, Q., and Coauthors, 2019: Fingerprints of internal drivers of Arctic sea ice loss in observations and model simulations. Nat. Geosci., 12, 28–33, https://doi.org/10.1038/s41561-018-0256-8.
England, M. R., I. Eisenman, N. J. Lutsko, T. J. W. Wagner, 2021: The recent emergence of Arctic amplification. Geophys. Res. Lett., 48, e2021GL094086, https://doi.org/10.1029/2021GL094086.
Francis, J. A., and B. Wu, 2020: Why has no new record-minimum Arctic sea-ice extent occurred since September 2012? Environ. Res. Lett., 15, 114034, https://doi.org/10.1088/1748-9326/abc047.
Grunseich, G., and B. Wang, 2016: Arctic sea ice patterns driven by the Asian summer monsoon. J. Climate, 29, 9097–9112, https://doi.org/10.1175/JCLI-D-16-0207.1.
Hanna, E., T. E. Cropper, R. J. Hall, and J. Cappelen, 2016: Greenland blocking index 1851–2015: A regional climate change signal. Int. J. Climatol., 36, 4847–4861, https://doi.org/10.1002/joc.4673.
Heo, E. S., M.-K. Sung, S.-I. An, and Y.-M. Yang, 2021: Decadal phase shift of summertime Arctic dipole pattern and its nonlinear effect on sea ice extent. Int. J. Climatol., 41, 4732–4742, https://doi.org/10.1002/joc.7097.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Huang, Y., G. Chou, Y. Xie, and N. Soulard, 2019: Radiative control of the interannual variability of Arctic sea ice. Geophys. Res. Lett., 46, 9899–9908, https://doi.org/10.1029/2019GL084204.
Kanamitsu, M., W. Ebisuzaki, J. Woollen, S.-K. Yang, J. J. Hnilo, M. Fiorino, and G. L. Potter, 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 1631–1644, https://doi.org/10.1175/BAMS-83-11-1631.
Kay, J. E., T. L’Ecuyer, A. Gettelman, G. Stephens, and C. O’Dell, 2008: The contribution of cloud and radiation anomalies to the 2007 Arctic sea ice extent minimum. Geophys. Res. Lett., 35, L08503, https://doi.org/10.1029/2008GL033451.
Kay, J. E., M. M. Holland, and A. Jahn, 2011: Inter-annual to multi-decadal Arctic sea ice extent trends in a warming world. Geophys. Res. Lett., 38, L15708, https://doi.org/10.1029/2011GL048008.
Kay, J. E., and Coauthors, 2015: The Community Earth System Model (CESM) large ensemble project: A community resource for studying climate change in the presence of internal climate variability. Bull. Amer. Meteor. Soc., 96, 1333–1349, https://doi.org/10.1175/BAMS-D-13-00255.1.
Kobayashi, S., and Coauthors, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 5–48, https://doi.org/10.2151/jmsj.2015-001.
Koenigk, T., and L. Brodeau, 2014: Ocean heat transport into the Arctic in the twentieth and twenty-first century in EC-Earth. Climate Dyn., 42, 3101–3120, https://doi.org/10.1007/s00382-013-1821-x.
Kumar, A., and Coauthors, 2010: Contribution of sea ice loss to Arctic amplification. Geophys. Res. Lett., 37, L21701, https://doi.org/10.1029/2010GL045022.
Kwok, R., G. F. Cunningham, and S. S. Pang, 2004: Fram Strait sea ice outflow. J. Geophys. Res., 109, C01009, https://doi.org/10.1029/2003JC001785.
Labe, Z., G. Magnusdottir, and H. Stern, 2018: Variability of Arctic sea ice thickness using PIOMAS and the CESM Large Ensemble. J. Climate, 31, 3233–3247, https://doi.org/10.1175/JCLI-D-17-0436.1.
Lim, Y. K., R. I. Cullather, S. M. J. Nowochi, and K. M. Kim, 2019: Inter-relationship between subtropical Pacific sea surface temperature, Arctic sea ice concentration, and North Atlantic Oscillation in recent summers. Sci. Rep., 9, 3481, https://doi.org/10.1038/s41598-019-39896-7.
Liu, J., J. A. Curry, and Y. Hu, 2004: Recent Arctic sea ice variability: Connections to the Arctic Oscillation and the ENSO. Geophys. Res. Lett., 31, L09211, https://doi.org/10.1029/2004GL019858.
Liu, J., Z. Chen, J. Francis, M. Song, T. Mote, and Y. Hu, 2016: Has Arctic sea ice loss contributed to increased surface melting of the Greenland Ice Sheet? J. Climate, 29, 3373–3386, https://doi.org/10.1175/JCLI-D-15-0391.1.
Liu, Z., and Coauthors, 2021: Acceleration of western Arctic sea ice loss linked to the Pacific North American pattern. Nat. Commun., 12, 1519, https://doi.org/10.1038/s41467-021-21830-z.
Livina, V. N., and T. M. Lenton, 2013: A recent tipping point in the Arctic sea-ice cover: Abrupt and persistent increase in the seasonal cycle since 2007. Cryosphere, 7, 275–286, https://doi.org/10.5194/tc-7-275-2013.
Luo, R., Q. Ding, Z. Wu, I. Baxter, M. Bushuk, Y. Huang, and X. Dong, 2021: Summertime atmosphere–sea ice coupling in the Arctic simulated by CMIP5/6 models: Importance of large-scale circulation. Climate Dyn., 56, 1467–1485, https://doi.org/10.1007/s00382-020-05543-5.
Manabe, S., and R. J. Stouffer, 1980: Sensitivity of a global climate model to an increase of CO2 concentration in the atmosphere. J. Geophys. Res., 85, 5529–5554, https://doi.org/10.1029/JC085iC10p05529.
Maslanik, J., S. Drobot, C. Fowler, W. Emery, and R. Barry, 2007: On the Arctic climate paradox and the continuing role of atmospheric circulation in affecting sea ice conditions. Geophys. Res. Lett., 34, L03711, https://doi.org/10.1029/2006GL028269.
Notz, D., and J. Marotzke, 2012: Observations reveal external driver for Arctic sea-ice retreat. Geophys. Res. Lett., 39, L08502, https://doi.org/10.1029/2012GL051094.
Notz, D., and J. Stroeve, 2016: Observed Arctic sea-ice loss directly follows anthropogenic CO2 emission. Science, 354, 747–750, https://doi.org/10.1126/science.aag2345.
Ogi, M., and J. M. Wallace, 2007: Summer minimum Arctic sea ice extent and the associated summer atmospheric circulation. Geophys. Res. Lett., 34, L12705, https://doi.org/10.1029/2007GL029897.
Ogi, M., I. G. Rigor, M. G. McPhee, and J. M. Wallace, 2008: Summer retreat of Arctic sea ice: Role of summer winds. Geophys. Res. Lett., 35, L24701, https://doi.org/10.1029/2008GL035672.
Ogi, M., K. Yamazaki, and J. M. Wallace, 2010: Influence of winter and summer surface wind anomalies on summer Arctic sea ice extent. Geophys. Res. Lett., 37, L07701, https://doi.org/10.1029/2009GL042356.
Overland, J. E., and M. Wang, 2010: Large-scale atmospheric circulation changes are associated with the recent loss of Arctic sea ice. Tellus, 62A (1), 1–9, https://doi.org/10.1111/j.1600-0870.2009.00421.x.
Overland, J. E., J. A. Francis, H. Edward, and M. Wang, 2012: The recent shift in early summer Arctic atmospheric circulation. Geophys. Res. Lett., 39, L19804, https://doi.org/10.1029/2012GL053268.
Rampal, P., J. Weiss, C. Dubois, and J.-M. Campin, 2011: IPCC climate models do not capture Arctic sea ice drift acceleration: Consequences in terms of projected sea ice thinning and decline. J. Geophys. Res., 116, C00D07, https://doi.org/10.1029/2011JC007110.
Rodionov, S. N., 2004: A sequential algorithm for testing climate regime shifts. Geophys. Res. Lett., 31, L09204, https://doi.org/10.1029/2004GL019448.
Rodionov, S. N., 2006: Use of prewhitening in climate regime shift detection. Geophys. Res. Lett., 33, L12707, https://doi.org/10.1029/2006GL025904.
Rossow, W., and R. Schiffer, 1999: Advances in understanding clouds from ISCCP. Bull. Amer. Meteor. Soc., 80, 2261–2288, https://doi.org/10.1175/1520-0477(1999)080<2261:AIUCFI>2.0.CO;2.
Screen, J. A., and I. Simmonds, 2010: The central role of diminishing sea ice in recent Arctic temperature amplification. Nature, 464, 1334–1337, https://doi.org/10.1038/nature09051.
Screen, J. A., and J. A. Francis, 2016: Contribution of sea ice loss to Arctic amplification is regulated by Pacific Ocean decadal variability. Nat. Climate Change, 6, 856–860, https://doi.org/10.1038/nclimate3011.
Screen, J. A., and C. Deser, 2019: Pacific Ocean variability influences the time of emergence of a seasonally ice-free Arctic Ocean. Geophys. Res. Lett., 46, 2222–2231, https://doi.org/10.1029/2018GL081393.
Screen, J. A., I. Simmonds, C. Deser, and R. Tomas, 2013: The atmospheric response to three decades of observed Arctic sea ice loss. J. Climate, 26, 1230–1248, https://doi.org/10.1175/JCLI-D-12-00063.1.
Screen, J. A., and Coauthors, 2018: Consistency and discrepancy in the atmospheric response to Arctic sea-ice loss across climate models. Nat. Geosci., 11, 155–163, https://doi.org/10.1038/s41561-018-0059-y.
Sedlar, J., 2018: Spring Arctic atmospheric preconditioning: Do not rule out shortwave radiation just yet. J. Climate, 31, 4225–4240, https://doi.org/10.1175/JCLI-D-17-0710.1.
Serreze, M. C., and J. A. Francis, 2006: The Arctic amplification debate. Climatic Change, 76, 241–264, https://doi.org/10.1007/s10584-005-9017-y.
Serreze, M. C., J. Stroeve, A. P. Barrett, and L. N. Boisvert, 2016: Summer atmospheric circulation anomalies over the Arctic Ocean and their influences on September sea ice extent: A cautionary tale. J. Geophys. Res. Atmos., 121, 11 463–11 485, https://doi.org/10.1002/2016JD025161.
Shimada, K., T. Kamoshida, M. Itoh, S. Nishino, E. Carmack, F. McLaughlin, S. Zimmermann, and A. Proshutinsky, 2006: Pacific Ocean inflow: Influence on catastrophic reduction of sea ice cover in the Arctic Ocean. Geophys. Res. Lett., 33, L08605, https://doi.org/10.1029/2005GL025624.
Spielhagen, R. F., and Coauthors, 2011: Enhanced modern heat transfer to the Arctic by warm Atlantic water. Science, 331, 450–453, https://doi.org/10.1126/science.1197397.
Svendsen, L., N. Keenlyside, I. Bethke, Y. Gao, and N.-E. Omrani, 2018: Pacific contribution to the early twentieth-century warming in the Arctic. Nat. Climate Change, 8, 793–797, https://doi.org/10.1038/s41558-018-0247-1.
Swart, N. C., J. C. Fyfe, E. Hawkins, J. E. Kay, and A. Jahn, 2015: Influence of internal variability on Arctic sea-ice trends. Nat. Climate Change, 5, 86–89, https://doi.org/10.1038/nclimate2483.
Takara, E. E., and R. G. Ellingson, 2000: Broken cloud field longwave-scattering effects. J. Atmos. Sci., 57, 1298–1310, https://doi.org/10.1175/1520-0469(2000)057<1298:BCFLSE>2.0.CO;2.
Tokinaga, H., S. P. Xie, and H. Mukougawa, 2017: Early 20th century Arctic warming intensified by Pacific and Atlantic multidecadal variability. Proc. Natl. Acad. Sci. USA, 114, 6227–6232, https://doi.org/10.1073/pnas.1615880114.
Topál, D., Q. Ding, J. Mitchell, I. Baxter, M. Herein, T. Haszpra, R. Luo, and Q. Li, 2020: An internal atmospheric process determining summertime Arctic sea ice melting in the next three decades: Lessons learned from five large ensembles and multiple CMIP5 climate simulations. J. Climate, 33, 7431–7454, https://doi.org/10.1175/JCLI-D-19-0803.1.
Tschudi, M. A., W. N. Meier, and J. S. Stewart, 2020: An enhancement to sea ice motion and age products at the National Snow and Ice Data Center (NSIDC). Cryosphere, 14, 1519–1536, https://doi.org/10.5194/tc-14-1519-2020.
Wang, J., J. Zhang, E. Watanabe, M. Ikeda, K. Mizobata, J. E. Walsh, X. Bai, and B. Wu, 2009: Is the dipole anomaly a major driver to record lows in Arctic summer sea ice extent? Geophys. Res. Lett., 36, L05706, https://doi.org/10.1029/2008GL036706.
Wang, L., L. Zhang, and W. Yang, 2020: The impact of concurrent variation of atmospheric meridional heat transport in western Baffin Bay and eastern Greenland on summer Arctic sea ice. Acta Oceanol. Sin., 39, 14–23, https://doi.org/10.1007/s13131-020-1614-0.
Wang, S.-Y., and J. Liu, 2016: Delving into the relationship between autumn Arctic sea ice and central–eastern Eurasian winter climate. Atmos. Oceanic Sci. Lett., 9, 366–374, https://doi.org/10.1080/16742834.2016.1207482.
Wernli, H., and L. Papritz, 2018: Role of polar anticyclones and mid-latitude cyclones for Arctic summertime sea-ice melting. Nat. Geosci., 11, 108–113, https://doi.org/10.1038/s41561-017-0041-0.
Wettstein, J. J., and C. Deser, 2014: Internal variability in projections of twenty-first-century Arctic sea ice loss: Role of the large-scale atmospheric circulation. J. Climate, 27, 527–550, https://doi.org/10.1175/JCLI-D-12-00839.1.
Wu, B., and Z. Li, 2021: Possible impacts of anomalous Arctic sea ice melting on summer atmosphere. Int. J. Climatol., 42, 1818–1827, https://doi.org/10.1002/joc.7337.
Wu, B., J. Wang, and J. E. Walsh, 2006: Dipole anomaly in the winter Arctic atmosphere and its association with sea ice motion. J. Climate, 19, 210–225, https://doi.org/10.1175/JCLI3619.1.
Yu, L., S. Zhong, T. Vihma, and B. Sun, 2021: Attribution of late summer early autumn Arctic sea ice decline in recent decades. npj Climate Atmos. Sci., 4, 3, https://doi.org/10.1038/s41612-020-00157-4.
Zhang, J., 2021: Recent slowdown in the decline of Arctic sea ice volume under increasingly warm atmospheric and oceanic conditions. Geophys. Res. Lett., 48, e2021GL094780, https://doi.org/10.1029/2021GL094780.
Zhang, J., and D. A. Rothrock, 2003: Modeling global sea ice with a thickness and enthalpy distribution model in generalized curvilinear coordinates. Mon. Wea. Rev., 131, 845–861, https://doi.org/10.1175/1520-0493(2003)131<0845:MGSIWA>2.0.CO;2.
Zhang, R., 2015: Mechanisms for low-frequency variability of summer Arctic sea ice extent. Proc. Natl. Acad. Sci. USA, 112, 4570–4575, https://doi.org/10.1073/pnas.1422296112.