Abstract

The decline in Barents Sea ice has been implicated in forcing the “warm-Arctic cold-Siberian” (WACS) anomaly pattern via enhanced turbulent heat flux (THF). This study investigates interannual variability in winter [December–February (DJF)] Barents Sea THF and its relationship to Barents Sea ice and the large-scale atmospheric flow. ERA-Interim and observational data from 1979/80 to 2011/12 are used. The leading pattern (EOF1: 33%) of winter Barents Sea THF variability is relatively weakly correlated (r = 0.30) with Barents Sea ice and appears to be driven primarily by atmospheric variability. The sea ice–related THF variability manifests itself as EOF2 (20%, r = 0.60). THF EOF2 is robust over the entire winter season, but its link to the WACS pattern is not. However, the WACS pattern emerges consistently as the second EOF (20%) of Eurasian surface air temperature (SAT) variability in all winter months. When Eurasia is cold, there are indeed weak reductions in Barents Sea ice, but the associated THF anomalies are on average negative, which is inconsistent with the proposed direct atmospheric response to sea ice variability. Lead–lag correlation analyses on shorter time scales support this conclusion and indicate that atmospheric variability plays an important role in driving observed variability in Barents Sea THF and ice cover, as well as the WACS pattern.

1. Introduction

Arctic sea ice has decreased in all seasons during recent decades. The largest declines in areal extent have occurred during summer and early autumn (up to 10% decade−1; Serreze et al. 2007), but the thicker multiyear ice cover is shrinking rapidly in winter as well (Comiso 2012). Wintertime Arctic sea ice area declines are concentrated in the Barents Sea (Serreze et al. 2009; Screen and Simmonds 2010a; Parkinson and Cavalieri 2012). The diminishing Arctic sea ice cover has been accompanied by near-surface warming in the high latitudes, particularly during the winter season (e.g., Screen and Simmonds 2010b), but the midlatitudes have recently experienced some anomalous winters with severe cold spells and above-normal snow cover in parts of Europe, Russia, and North America (e.g., Cohen et al. 2007; Cattiaux et al. 2010; Ghatak et al. 2010; Guirguis et al. 2011; Orsolini et al. 2012; Coumou and Rahmstorf 2012). The cold Eurasian surface temperatures are part of what has been termed the “warm-Arctic cold-Siberian” (WACS) anomaly pattern (or slight variations thereof; see Inoue et al. 2012; Mori et al. 2014). The relationship between this WACS pattern and Barents Sea ice conditions is the focus of this study, which probes reanalysis data for clues about how the two are linked.

Various mechanisms have been proposed to explain how the WACS pattern is generated. Early on, Honda et al. (2009) pointed to the fact that, where sea ice retreats, we expect enhanced ocean-to-atmosphere energy loss in the form of turbulent heat fluxes (THFs). They suggested that such a THF perturbation triggers a stationary Rossby wave train that amplifies the climatological-mean wintertime high pressure over Siberia and enhances northerly cold-air advection over eastern Eurasia. Other studies focus on the role of the THF perturbation in modifying the near-surface meridional temperature gradient, leading to altered cyclone pathways and a cold anticyclonic flow anomaly north of the Eurasian continent (Inoue et al. 2012), or weakened zonal winds and reduced heat transport from the North Atlantic to the interior of the Eurasian continent (Petoukhov and Semenov 2010; Outten and Esau 2012). A more global (less regional) alternative suggests that the modified (weaker) meridional temperature gradient causes the atmospheric flow to be wavier, thus leading to more extreme midlatitude weather in general (Francis and Vavrus 2012).

A different class of studies examines the phenomenon of cold Eurasian winters independently of Arctic sea ice. These studies attribute cold winters to persistent atmospheric blocking but not necessarily in connection with declining ice cover. Wintertime blocking patterns in recent cold years (e.g., 2005/06, 2009/10) are shown to be linked to sudden stratospheric warmings and/or sea surface temperature anomalies (Scaife and Knight 2008; Croci-Maspoli and Davies 2009; Cattiaux et al. 2010), with Rossby wave trains playing a possible amplifying role over Siberia (Takaya and Nakamura 2005; Park et al. 2011; Cheung et al. 2013; Sato et al. 2014). While such atmospheric circulation changes could be forced by sea ice changes (Mori et al. 2014), they could also be an expression of natural internal variability or reflect a direct atmospheric response to external radiative forcing.

The range of proposed mechanisms, including some that do not explicitly require a role for sea ice, has led to some debate over details of the relationship between the WACS pattern and sea ice cover. Some studies suggest that the WACS pattern is a delayed wintertime atmospheric response to reduced sea ice cover in the preceding autumn (Honda et al. 2009; Francis et al. 2009; Overland and Wang 2010; Overland et al. 2011) possibly via an indirect response involving the stratosphere (Jaiser et al. 2012; Orsolini et al. 2012; Cohen et al. 2014; Kim et al. 2014; García-Serrano et al. 2015; King et al. 2015). Others suggest that it is a simultaneous response to winter Barents Sea ice anomalies (Petoukhov and Semenov 2010; Hori et al. 2011; Inoue et al. 2012; Outten and Esau 2012). The link to winter sea ice anomalies has been shown to be more robust in at least one comparison (Tang et al. 2013). Furthermore, while some studies focus on interannual covariability in sea ice and Eurasian winters (Honda et al. 2009; Petoukhov and Semenov 2010; Hori et al. 2011; Inoue et al. 2012), others examine potential linkages to strong negative trends in sea ice cover (Outten and Esau 2012; Mori et al. 2014). Most recently, a series of observational and modeling studies has questioned whether there is adequate evidence that sea ice influences atmospheric circulation, blocking, and Eurasian winter temperatures at all (Hopsch et al. 2012; Barnes 2013; Screen and Simmonds 2013; Barnes et al. 2014; Wallace et al. 2014; Woollings et al. 2014).

Despite unresolved issues about the mechanism, seasonality, and time scale, many of the existing hypotheses attempting to explain the WACS pattern and its variability share a common starting point: an enhancement of ocean-to-atmosphere THFs in late autumn to winter as a direct consequence of reduced sea ice. These enhanced THFs are hypothesized to create a large-scale atmospheric response leading to the WACS pattern. This study will explore the robustness of this causal chain in reanalysis data by examining the THF covariability associated with both Barents Sea ice conditions and cold Eurasian winters.

Studies using AGCMs provide qualified support for the causal chain summarized in the previous paragraph. Movements of the sea ice edge are found to affect the spatial distribution of the THF field, producing a distinctive dipole signature with enhanced ocean-to-atmosphere THF in regions of ice loss next to the new ice edge and reduced THF next to the old ice edge (e.g., Alexander et al. 2004; Deser et al. 2004; Magnusdottir et al. 2004; Singarayer et al. 2006; Seierstad and Bader 2009; Budikova 2009; Deser et al. 2010). However, the large-scale atmospheric response to these THF anomalies is weak compared to natural variability (Screen et al. 2014); varies considerably among AGCM studies (Bader et al. 2011); exhibits nonlinear behavior (e.g., Lau and Holopainen 1984; Peng et al. 1997; Deser et al. 2007, 2010; Liptak and Strong 2014); and depends strongly on the magnitude, location, and polarity of the sea ice changes (e.g., Honda et al. 1999; Magnusdottir et al. 2004; Deser et al. 2004; Kvamstø et al. 2004), as well as the season (e.g., Alexander et al. 2004; Bhatt et al. 2008; Seierstad and Bader 2009; Deser et al. 2010).

Furthermore, AGCM experiments cannot capture the two-way interactions that link the ocean and the atmosphere in nature. The uncoupled AGCM setup allows us to investigate the atmospheric response to variability in the surface ocean (sea surface temperature and sea ice). It does not account for the other direction of coupling, known to be important on weekly to interannual time scales (Bjerknes 1964; Cayan 1992; Deser et al. 2000; Gulev et al. 2013), namely, the atmospheric driving of surface ocean variability via changes in near-surface winds, temperature gradients, and humidity gradients across the ocean–atmosphere interface (Fang and Wallace 1994; Prinsenberg et al. 1997; Deser et al. 2000; Rigor et al. 2002; Sorteberg and Kvingedal 2006; Wu and Zhang 2010; Germe et al. 2011; Wettstein and Deser 2014). As a result, uncoupled experiments often exhibit unrealistically strong and persistent ocean-to-atmosphere fluxes and reduced internal atmospheric variability (Barsugli and Battisti 1998). The THF anomalies in AGCM studies are by definition caused by the prescribed surface ocean conditions, and thus they cannot be used as evidence for the proposed causal chain.

In the current study, we explore the coupled interactions between Barents Sea ice cover, THF, and the large-scale atmospheric circulation during winter to further examine the causal chain linking Barents Sea ice conditions to the WACS pattern. We focus on THF in the Barents Sea, the region with the most pronounced winter sea ice variability, as the coupling agent between the ocean and atmosphere. Specifically, we explore whether Barents Sea THF variability is driven by sea ice or atmospheric variability in the observational record. Whereas previous observational studies investigated the WACS pattern in association with a few extremely low Barents Sea ice years (Honda et al. 2009; Hori et al. 2011; Inoue et al. 2012) or in association with trends in Barents Sea ice (Outten and Esau 2012; Outten et al. 2013), we focus on interannual variability during all years of the satellite record, excluding the linear trend. By removing the trend, we maximize the statistical degrees of freedom and isolate mechanisms that can produce the WACS pattern independent of lower-frequency climate change. This allows us to evaluate a candidate null hypothesis that the WACS pattern can be attributed at least in part to internal atmospheric variability. The corollary of our detrending decision is that the results of this study are not directly comparable to an examination of coupled interactions in the context of long-term, externally forced climate change.

The paper is organized as follows: Data and methods are described in section 2, results are reported in section 3, and the discussion is presented in section 4. A summary completes the paper.

2. Data and methods

We use data from the European Centre for Medium-Range Weather Forecasts interim reanalysis (ERA-Interim; Dee et al. 2011) from 1979/80 to 2011/12. We focus analyses on the 3-month winter season from December to February (DJF), but the results and conclusions are also valid over an extended winter season (October–February), a point we will return to in the discussion section. Monthly and daily THF, sea surface temperature (SST), surface air temperature (SAT), 10-m near-surface wind in the zonal (U) and meridional (V) directions, sea level pressure (SLP), and geopotential height at 500 hPa (Z500) are analyzed in the 1.5° × 1.5°-resolution ERA-Interim. The THF is defined as the sum of the surface latent and sensible heat fluxes. Positive THF values indicate heat transfer from the ocean to the atmosphere. Sea ice concentration (SIC) is estimated from passive microwave satellite data on a 25 km × 25 km grid (NSIDC; Cavalieri et al. 1996). This satellite-based SIC is independent from, but consistent with, SIC obtained from ERA-Interim for the Barents Sea as a whole (r = 0.98 for a raw correlation using monthly data, r = 0.97 when linearly detrended) and also in terms of the geographic distribution of the sea ice edge (Fig. 1). All results and conclusions are similar when SIC from ERA-Interim is used instead of the NSIDC data. We define two regions (both indicated in Fig. 6) for the analyses: the Barents Sea (67°–81°N, 12°– 73°E) and the Eurasian region (45°–81°N, 12°–120°E). The Eurasian region includes mid- and high-latitude continental Eurasia and the adjacent Arctic seas. Note that the Barents Sea region is wholly subsumed within the northwest quadrant of the larger Eurasian region. The results in this study are not very sensitive to the exact definition of the regions, several of which were tested and all of which produced qualitatively similar results. The Barents Sea ice index (ICEBar) is defined as the detrended area-averaged SIC over the Barents Sea region, with positive ICEBar corresponding to less Barents Sea ice.

Fig. 1.

(a) Time series of DJF Barents Sea ice area in NSIDC (black) and ERA-Interim (purple). The correlation between time series is 0.98 for the seasonal-mean time series (also 0.98 for the corresponding monthly time series). Dots indicate greater (red) or less (blue) than average Barents Sea ice area for the linearly detrended data. (b) Climatological-mean SIC (color shading) from NSIDC and the 15% isopleth of SIC from NSIDC (black) and ERA-Interim (purple). The Barents Sea domain (67°–81°N, 12°–73°E) is indicated by the yellow box.

Fig. 1.

(a) Time series of DJF Barents Sea ice area in NSIDC (black) and ERA-Interim (purple). The correlation between time series is 0.98 for the seasonal-mean time series (also 0.98 for the corresponding monthly time series). Dots indicate greater (red) or less (blue) than average Barents Sea ice area for the linearly detrended data. (b) Climatological-mean SIC (color shading) from NSIDC and the 15% isopleth of SIC from NSIDC (black) and ERA-Interim (purple). The Barents Sea domain (67°–81°N, 12°–73°E) is indicated by the yellow box.

Unless otherwise stated, results in this study are based on monthly anomalies calculated by removing the linear trend from each month individually and then removing the climatological-mean seasonal cycle from the detrended data. Standard empirical orthogonal function (EOF)/principal component (PC) analysis is performed on monthly anomalies of 1) THF over the Barents Sea and 2) SAT over the Eurasian region. Gridpoint values of the anomaly fields are weighted by the square root of the cosine of latitude (equivalent to area weighting the variance) before performing the EOF analysis. EOFs presented in this study are well separated from each other and lower-order EOFs according to the criterion described by North et al. (1982). The resulting spatial patterns are referred to as EOFs and the associated time series are referred to as PCs. Broader climate relationships are explored by regressing monthly anomalies of relevant variables onto standardized PCs and ICEBar, such that the resulting regression maps are displayed in physical units per standard deviation of the index.

Later in this study we create “synthetic” time series of the WACS pattern at daily (WACSdsyn) and seasonal (WACSssyn) time scales. In both cases, the synthetic time series is created by projecting Eurasian SAT anomaly patterns at these two time scales onto the second EOF of Eurasian DJF SAT variability (i.e., what is defined in the results section as the WACS pattern). For WACSdsyn, the linear trend and climatological daily mean are removed to create the anomaly patterns, and these are used in a variety of lead–lag correlation and regression analyses to provide additional context to our evaluation of the causal chain described in the introduction section. For WACSssyn, only the climatological seasonal mean is removed to create the anomaly patterns, and we assess the variance associated with the trend and in the linearly detrended WACSssyn time series.

The conclusions of the current study have been found to be robust when other satellite-based [OAFlux (Yu et al. 2008) and HOAPS (Andersson et al. 2011)] and reanalysis-based [NCEP1 (Kalnay et al. 1996)] THF products are used instead of ERA-Interim data. The OAFlux and HOAPS products are thought to give improved heat flux estimates compared to reanalysis products, but a disadvantage is that they mask out data in areas with sea ice concentration above 50%, which makes it impossible to examine comprehensively the associations between THF and SIC variability.

3. Results

Because ocean-to-atmosphere THF has been widely invoked as the intermediary for communicating sea ice changes to the atmosphere (cf. references in the introduction section), we first examine the leading patterns of winter (DJF) THF variability and their association with sea ice cover in the Barents Sea. Figure 2 shows the leading Barents Sea THF PCs along with the ICEBar index defined in section 2. The first PC (PC1) is associated with 33% of the total DJF variability in Barents Sea THF, but it is relatively weakly correlated with the ICEBar time series (r = 0.30). The second PC (PC2) is associated with less of the variability (20%), but it is more strongly correlated with the ICEBar (r = 0.60). The relationships between Barents Sea THF PCs and ICEBar are robust not only for DJF but also throughout the extended [October–February (ONDJF)] winter season (Table 1).

Fig. 2.

Standardized time series of DJF ICEBar (black), THF PC1 (blue), and THF PC2 (red) based on seasonal-mean data with the seasonal cycle removed. Positive ICEBar corresponds to less Barents Sea ice area. Correlation values are r(PC1, ICEBar) = 0.46 (0.30 for the corresponding monthly time series) and r(PC2, ICEBar) = 0.63 (0.60 for the corresponding monthly time series).

Fig. 2.

Standardized time series of DJF ICEBar (black), THF PC1 (blue), and THF PC2 (red) based on seasonal-mean data with the seasonal cycle removed. Positive ICEBar corresponds to less Barents Sea ice area. Correlation values are r(PC1, ICEBar) = 0.46 (0.30 for the corresponding monthly time series) and r(PC2, ICEBar) = 0.63 (0.60 for the corresponding monthly time series).

Table 1.

Correlations between Barents Sea THF PCs and the ICEBar for various months from October to February. THF PCs and the ICEBar time series are based on detrended monthly data with the seasonal cycle removed.

Correlations between Barents Sea THF PCs and the ICEBar for various months from October to February. THF PCs and the ICEBar time series are based on detrended monthly data with the seasonal cycle removed.
Correlations between Barents Sea THF PCs and the ICEBar for various months from October to February. THF PCs and the ICEBar time series are based on detrended monthly data with the seasonal cycle removed.

The spatial patterns associated with ICEBar and the THF PCs support the idea that THF PC2 is more connected to sea ice cover than THF PC1. The THF signature of ICEBar is familiar (Fig. 3a)—strongly enhanced ocean-to-atmosphere THF over areas of reduced SIC accompanied by moderate reductions in THF directly to the south, indicating a northward shift in intense turbulent heat loss from the ocean associated with the sea ice edge (cf. Deser et al. 2004; Magnusdottir et al. 2004; Alexander et al. 2004; Deser et al. 2010). Recall that increases in ICEBar correspond to reductions in Barents-averaged SIC. Similar THF and SIC signatures appear in association with the second pattern of Barents Sea THF variability (THF EOF2; Fig. 3c), suggesting that THF EOF2 captures at least some aspects of the atmospheric response to sea ice reductions. The leading pattern (THF EOF1), however, tells a different story (Fig. 3b). THF EOF1 describes a broad, Barents-wide reduction in THF (reduced ocean-to-atmosphere heat loss in the yellow box) that is associated with relatively weak sea ice reductions. The THF–sea ice relationship (reduced THF with reduced sea ice) in Fig. 3b is incompatible with the expected relationship associated with an atmospheric response to sea ice reduction (enhanced THF with reduced sea ice).

Fig. 3.

DJF THF (color shading) and SIC regressions (blue contours indicate reduced sea ice, red contours indicate increased sea ice; contour interval 5%) onto (a) ICEBar (b) THF PC1, and (c) THF PC2. White dots indicate regression values that are not significant using a two-sided t test and a 0.05 significance threshold. Only significant values are plotted for SIC. The Barents Sea domain is indicated by the yellow box. Baffin Bay is the only area with positive (red) SIC regressions.

Fig. 3.

DJF THF (color shading) and SIC regressions (blue contours indicate reduced sea ice, red contours indicate increased sea ice; contour interval 5%) onto (a) ICEBar (b) THF PC1, and (c) THF PC2. White dots indicate regression values that are not significant using a two-sided t test and a 0.05 significance threshold. Only significant values are plotted for SIC. The Barents Sea domain is indicated by the yellow box. Baffin Bay is the only area with positive (red) SIC regressions.

The strength of winter THF depends on near-surface wind speeds and on the temperature difference between the (warm) ocean and (cold) atmosphere (Maykut 1982; Schröder et al. 2003; Alexander et al. 2004). If THF is enhanced as a result of a reduction in sea ice cover (as represented by ICEBar and THF PC2), then the ocean provides more heat to the atmosphere and we expect warm SAT anomalies. This relationship is shown in Figs. 3a and 4d and 3c and 4f for ICEBar and THF PC2, respectively, meaning that the variability described by the two indices is thus far consistent with an atmospheric response to Barents Sea ice perturbations. In contrast, both THF and sea ice cover are reduced with THF PC1 increases (Fig. 3b), while SAT anomalies are also warm (Fig. 4e). This suggests that the THF reductions associated with THF PC1 are likely driven by atmospheric changes, in particular a warming of the overlying atmosphere that reduces the ocean–atmosphere temperature difference. Oceanic (SST) correlations with THF PC1 are generally much weaker than the corresponding atmospheric (SAT) correlations (cf. Figs. 4b and 4e), which is to be expected if the ocean is responding to atmospheric changes. In comparison, THF PC2 exhibits spatially coherent oceanic and atmospheric patterns and slightly stronger correlations with SST than with SAT (cf. Figs. 4c and 4f), in concert with an atmospheric response to ocean changes, such as Barents Sea ice variability. Together, the analyses in Figs. 24 suggest an active role for the atmosphere in driving a substantial portion of THF variability over the Barents Sea.

Fig. 4.

(top) DJF correlations (color shading) between SST and (a) ICEBar, (b) THF PC1, and (c) THF PC2 with SIC regressions (blue contours indicate reduced sea ice, red contours indicate increased sea ice; interval 5%) onto the same indices. (bottom) DJF correlations (color shading) between SAT and (d) ICEBar, (e) THF PC1, and (f) THF PC2 with 10-m wind regressions (vectors) onto the same indices. White dots indicate values of the shaded fields that are not significant using a two-sided t test and a 0.05 significance threshold. Only significant values are plotted for SIC and 10-m winds. Vectors representing wind speeds below 0.5 m s−1 are omitted for clarity.

Fig. 4.

(top) DJF correlations (color shading) between SST and (a) ICEBar, (b) THF PC1, and (c) THF PC2 with SIC regressions (blue contours indicate reduced sea ice, red contours indicate increased sea ice; interval 5%) onto the same indices. (bottom) DJF correlations (color shading) between SAT and (d) ICEBar, (e) THF PC1, and (f) THF PC2 with 10-m wind regressions (vectors) onto the same indices. White dots indicate values of the shaded fields that are not significant using a two-sided t test and a 0.05 significance threshold. Only significant values are plotted for SIC and 10-m winds. Vectors representing wind speeds below 0.5 m s−1 are omitted for clarity.

Before exploring the atmosphere’s role further, we check the robustness of the sea ice–related THF variability (i.e., THF PC2) in individual winter months (Fig. 5). The localized dipole-like structure in THF appears in all months, and is consistently associated with enhanced ocean-to-atmosphere THF over areas of reduced SIC (Figs. 5a–c) and warmer SATs over the Barents Sea (Figs. 5d–f). On a hemispheric scale, the SAT regressions in all months show a generally cooler Eurasian continent similar to the WACS pattern described by Inoue et al. (2012), but the cooling is only statistically significant in December (Figs. 5d–f). In contrast, there is substantial variation in the large-scale atmospheric circulation patterns associated with THF PC2 (Fig. 5g–i). There are few robust features in the SLP regressions across all months, and we do not observe the expected linear atmospheric response to enhanced local heating—a negative SLP anomaly directly downstream of a heat source (e.g., Hoskins and Karoly 1981). We see some evidence of northerly cold-air advection over Siberia, in agreement with the circulation pattern described by Honda et al. (2009) and other studies listed in the introduction section. Similar results are obtained if the ICEBar index is used instead of THF PC2, particularly during midwinter.

Fig. 5.

Monthly regressions onto standardized and contemporary monthly THF PC2. (top) THF (color shading) and SIC (blue contours indicate reduced sea ice, red contours indicate increased sea ice; contour interval 5%), (middle) SAT (color shading), and (bottom) SLP (color shading) and 10-m winds (vectors). White dots indicate values of the shaded fields that are not significant using a two-sided t test and a 0.05 significance threshold. Only significant values are plotted for SIC and 10-m winds. Vectors representing wind speeds below 0.5 m s−1 are omitted for clarity.

Fig. 5.

Monthly regressions onto standardized and contemporary monthly THF PC2. (top) THF (color shading) and SIC (blue contours indicate reduced sea ice, red contours indicate increased sea ice; contour interval 5%), (middle) SAT (color shading), and (bottom) SLP (color shading) and 10-m winds (vectors). White dots indicate values of the shaded fields that are not significant using a two-sided t test and a 0.05 significance threshold. Only significant values are plotted for SIC and 10-m winds. Vectors representing wind speeds below 0.5 m s−1 are omitted for clarity.

The regression analyses in Fig. 5 show that sea ice–related THF variability is present in all winter months and is consistently associated with warming over the Barents Sea, but the large-scale circulation patterns vary substantially. As established in the introduction section, the variation from month to month could be due to nonlinearities in the dynamical response of the atmosphere or to internal variability, which is known to play a role even in the presence of strong forcing (Deser et al. 2004, 2012; Wettstein and Deser 2014). Internal variability is especially important on larger spatial scales; so even if the regional signature of THF PC2 can be understood as a simple atmospheric response to sea ice variability, hemispheric signatures like the WACS pattern seem to be more complicated.

Following this idea, we consider variability associated with the large-scale WACS pattern (Inoue et al. 2012; Mori et al. 2014) identified independently of Barents Sea ice variability. The WACS pattern emerges consistently as the second EOF (20%) of monthly Eurasian SAT variability in all winter months (Figs. 6d–f).1 The SAT and SLP regressions onto the WACS index are substantially stronger and generally more robust across all winter months than the comparable regressions onto THF PC2 (Figs. 5d–i) or ICEBar (not shown). In its positive phase, the WACS pattern exhibits robust out-of-phase SAT anomalies between warmer temperatures over the Barents–Kara Seas and colder temperatures over Eurasia. It is consistently associated with a prominent SLP high over the Eurasian continent and southerly near-surface wind anomalies over the Barents Sea from December to January (Figs. 6g–i). When Eurasia is cold, Barents Sea ice concentration is somewhat lower than average, but the associated Barents Sea THF anomalies are, on average, negative (Figs. 6a–c) for each winter month. This is the opposite of what would be expected if the WACS pattern were a direct atmospheric response to reduced Barents Sea ice. Thus, atmospheric variability appears not only to shape the WACS pattern itself, but also to contribute substantially to observed THF variability over the Barents Sea and large-scale SAT variability.

Fig. 6.

As in Fig. 5, but all regressions are onto the WACS index (PC2 of Eurasian SAT) instead of THF PC2. The domain of the Eurasian region (45°–81°N, 12°–120°E) is indicated in (f).

Fig. 6.

As in Fig. 5, but all regressions are onto the WACS index (PC2 of Eurasian SAT) instead of THF PC2. The domain of the Eurasian region (45°–81°N, 12°–120°E) is indicated in (f).

Causality is better explored by examining lead–lag relationships between the WACS pattern and atmospheric or oceanic indicators on shorter time scales. Figure 7 shows lagged correlations between a variety of indicators and a daily index of the WACS pattern (WACSdsyn as defined in the data and methods section). On daily time scales, the Barents-averaged THF (THFdBar; blue curve) leads the WACS pattern, with reduced ocean-to-atmosphere heat fluxes peaking 2 days before the WACS pattern. This echoes the message from the monthly analysis (Fig. 6), particularly that the WACS pattern does not appear to be a direct atmospheric response to reduced Barents Sea ice (and increased THF) on short (daily) or long (monthly) time scales. Note that because THF PC1 represents a basinwide pattern of THF variability in the Barents Sea, it is very similar to THFdBar except for a change in sign (see Fig. 3b). The lagged correlation curves associated with the daily version of these two indices are also very similar except for a change in sign (r = −0.96; see also Fig. 3b). On the other hand, the daily version of the THF PC2 index, which represents sea ice–related THF variability, exhibits low correlations with WACSdsyn at all lags (red curve). The lagged correlation between ICEdBar and WACSdsyn (black curve) peaks at zero lag, but it is stronger with ICEdBar lagging rather than leading, casting further doubt on the idea that the WACS pattern is a direct consequence of reduced Barents Sea ice.

Fig. 7.

Lagged correlations between the synthetic daily PC2 of Eurasian SAT (WACSdsyn) and a number of other daily indices averaged over the Barents Sea region for DJF 1979/80 to 2011/12: THF (THFdBar, blue; positive indicates ocean-to-atmosphere heat flux), THF PC2dsyn (red; positive index associated with lower sea ice and a THF dipole similar to that shown in Fig. 3c), Barents Sea ice area (ICEdBar, black; positive index indicates less ice area), and 10-m winds (VdBar, green; positive index indicates southerly winds). The autocorrelation of WACSdysn is shown in orange. The shading shows the 95% confidence interval on the correlations calculated using Fisher’s z transform.

Fig. 7.

Lagged correlations between the synthetic daily PC2 of Eurasian SAT (WACSdsyn) and a number of other daily indices averaged over the Barents Sea region for DJF 1979/80 to 2011/12: THF (THFdBar, blue; positive indicates ocean-to-atmosphere heat flux), THF PC2dsyn (red; positive index associated with lower sea ice and a THF dipole similar to that shown in Fig. 3c), Barents Sea ice area (ICEdBar, black; positive index indicates less ice area), and 10-m winds (VdBar, green; positive index indicates southerly winds). The autocorrelation of WACSdysn is shown in orange. The shading shows the 95% confidence interval on the correlations calculated using Fisher’s z transform.

Finally, southerly wind anomalies over the Barents region (VdBar; green curve) also peak 2 days before WACSdsyn. Supposing that these wind anomalies are associated with an anomalous atmospheric circulation pattern, this could suggest 1) advection of warm air from the south into the Barents Sea region, 2) a basinwide reduction of THF as a result of a correspondingly weakened temperature contrast between the warmer ocean and colder atmosphere, and 3) a more slowly evolving decrease in Barents Sea ice cover driven mechanically by southerly winds pushing the sea ice edge northward. This chain of events is consistent with the lead–lag relationships in Fig. 7 and suggests, along with previous results, that the WACS pattern does not originate directly from enhanced THF caused by Barents Sea ice reduction.

The corresponding spatial patterns at various lags (Fig. 8) provide some final clues about how the WACS pattern, THF, and atmospheric circulation over the Barents Sea are linked. The patterns are consistent with the results shown in Fig. 7: the WACS pattern is associated with negative Barents Sea THF anomalies (reduced ocean-to-atmosphere THF) and a relatively weak reduction in the Barents Sea ice cover (Figs. 8a–c), which together are inconsistent with the notion that the WACS pattern is a direct atmospheric response to reduced Barents Sea ice. The spatial distribution and diminishing intensity of THF anomalies over time (Figs. 8a–c) are instead consistent with atmospheric changes driving ocean-to-atmosphere THF variability (e.g., Bjerknes 1964; Cayan 1992; Gulev et al. 2013). The associated near-surface signature exhibits southerly and southwesterly winds (Figs. 8d–f), consistent with warm temperature advection over the Barents Sea. The large-scale surface (SLP) and midtropospheric (Z500) anomalies (Figs. 8g–i) resemble patterns associated with wave propagation and winter atmospheric blocking (e.g., Park et al. 2011; Cheung et al. 2013) and, to some degree, with the east Atlantic pattern (Wallace and Gutzler 1981; Smoliak 2009; Wettstein and Wallace 2010). It is also noteworthy that the strongest Z500 anomalies occur prior to the strongest (lag 0) correlation between the WACS pattern and Barents Sea ice reductions (cf. Figs. 7 and 8g–i), further supporting the interpretation that the large-scale atmospheric wave exists prior to any atmospheric changes initiated by Barents Sea ice (and the associated THF) variability. Our interpretation of the collective results is that the WACS pattern reflects a large-scale pattern of intrinsic coupled climate variability that is related to both the atmospheric forcing of and the atmospheric response to Barents Sea ice variability.

Fig. 8.

Lagged regressions onto the daily WACSdsyn index (lags expressed in days, similar to Fig. 7; positive lag indicates WACSdsyn leading). (a)–(c) THF (color shading with red indicating ocean-to-atmosphere heat flux) and SIC (blue contours indicated reduced sea ice, red contours indicate increased sea ice; contour interval 5%); (d)–(f) SAT (color shading) and 10-m winds (vectors); and (g)–(i) SLP (color shading) and Z500 (solid/dashed contours indicate positive/negative values; contour interval = 20 m). Vectors representing wind speeds below 0.5 m s−1 are omitted for clarity.

Fig. 8.

Lagged regressions onto the daily WACSdsyn index (lags expressed in days, similar to Fig. 7; positive lag indicates WACSdsyn leading). (a)–(c) THF (color shading with red indicating ocean-to-atmosphere heat flux) and SIC (blue contours indicated reduced sea ice, red contours indicate increased sea ice; contour interval 5%); (d)–(f) SAT (color shading) and 10-m winds (vectors); and (g)–(i) SLP (color shading) and Z500 (solid/dashed contours indicate positive/negative values; contour interval = 20 m). Vectors representing wind speeds below 0.5 m s−1 are omitted for clarity.

4. Discussion

In the previous section, we investigated interactions between the observed Barents Sea ice cover and large-scale atmospheric circulation with a focus on ocean-to-atmosphere THF as the coupling agent between the underlying ocean and the overlying atmosphere. The goal was to evaluate a candidate null hypothesis that the observed interannual variability in the winter WACS pattern might primarily be an expression of atmospheric variability rather than an atmospheric response to sea ice variability. Our results suggest that we cannot reject this null hypothesis, and that atmospheric circulation variability does in fact play a critical role in determining the SLP, SAT, THF, and even the sea ice distribution both over the Barents Sea and farther afield.

A number of studies mentioned in the introduction section suggest that the Barents Sea ice variability in October and November (ON) causes a delayed response in the winter atmospheric circulation (Honda et al. 2009; Francis et al. 2009; Overland and Wang 2010; Overland et al. 2011; Jaiser et al. 2012; Orsolini et al. 2012; Cohen et al. 2014; Kim et al. 2014; García-Serrano et al. 2015; King et al. 2015). Because the previous section focuses on synchronous analyses during the DJF season, it does not address the possibility of this autumn-to-winter linkage. To explore this issue, Fig. 9 shows lagged regressions between the winter (DJF)-mean WACS index with THF, SIC, SST, and SLP anomalies in the months leading up to winter. The DJF WACS pattern is indeed associated with negative SIC anomalies (in the Barents, Kara, Laptev, and east Siberian Seas) and positive THF anomalies in the vicinity of the SIC anomalies during October and November, but many of these anomalies are insignificant.

Fig. 9.

Lagged regressions onto the seasonal-mean DJF WACS pattern (SAT PC2) of the following fields, for both the contemporary DJF season and during the preceding October and November. (top) THF (color shading) and SIC (blue contours indicate reduced sea ice, red contours indicate increases sea ice; contour interval 5%); (middle) SAT (color shading); and (bottom) SLP (color shading) and 10-m winds (vectors). White dots indicate values of the shaded fields that are not significant using a two-sided t test and a 0.05 significance threshold. Only significant values are plotted for 10-m wind regressions. All values are plotted for SIC anomalies since many of them are not significant. Vectors representing wind speeds below 0.5 m s−1 are omitted for clarity. The domain of the Eurasian region (45°–81°N, 12°–120°E) is indicated in (f).

Fig. 9.

Lagged regressions onto the seasonal-mean DJF WACS pattern (SAT PC2) of the following fields, for both the contemporary DJF season and during the preceding October and November. (top) THF (color shading) and SIC (blue contours indicate reduced sea ice, red contours indicate increases sea ice; contour interval 5%); (middle) SAT (color shading); and (bottom) SLP (color shading) and 10-m winds (vectors). White dots indicate values of the shaded fields that are not significant using a two-sided t test and a 0.05 significance threshold. Only significant values are plotted for 10-m wind regressions. All values are plotted for SIC anomalies since many of them are not significant. Vectors representing wind speeds below 0.5 m s−1 are omitted for clarity. The domain of the Eurasian region (45°–81°N, 12°–120°E) is indicated in (f).

The October–November regression results (Figs. 9a,b) are consistent with the suggestion that negative SIC/positive THF anomalies during autumn can influence the wintertime circulation but still do not provide adequate grounds for a direct causal interpretation. The geographical distribution of the fields (SIC, THF, SLP, and 10-m winds) is transitory over the late fall and early winter and the THF anomalies are displaced from the strongest SIC anomalies, suggesting an important role for nonlinear ice-circulation effects, internal atmospheric variability, or both. Many studies also point to more complicated, indirect pathways linking SIC and circulation via autumn stratospheric anomalies that propagate downward to create the tropospheric signals later in winter (Jaiser et al. 2012; Orsolini et al. 2012; Cohen et al. 2014; Kim et al. 2014; García-Serrano et al. 2015; King et al. 2015). While the downward propagation of stratospheric anomalies is clear in observations, the question of whether they are caused by sea ice anomalies is not yet settled (see discussion in King et al. 2015). Given the apparent complexity of the mechanisms and the relatively few degrees of freedom, a clear diagnosis of the causal linkage between autumn ice anomalies and midwinter WACS anomalies is perhaps unobtainable using reanalysis.

Sea ice loss is prominent in the observed record and is also consistently projected for the future within coupled GCM experiments forced by anthropogenic increases in greenhouse gases. The associated atmospheric response is reflected in various aspects of climate, both locally and remotely (e.g., Serreze et al. 2009; Simmonds and Keay 2009; Honda et al. 2009; Francis et al. 2009; Screen and Simmonds 2010a,b; Overland and Wang 2010; Overland et al. 2011; Blüthgen et al. 2012; Jaiser et al. 2012; Orsolini et al. 2012; Screen et al. 2013; Mori et al. 2014). To more completely evaluate the relationship between the WACS pattern, Barents Sea ice, and Barents Sea THF, we performed a full set of parallel analyses on raw (not detrended) variables.

The main conclusions of the current study remain when trends are retained (not shown). Consistent with Mori et al. (2014), we find trends toward a positive DJF WACS pattern and increased Barents Sea ice loss, both of which are particularly evident from 2004 onward. Attributing causality to this relationship is difficult, however, because of few degrees of freedom and because the correlation between the raw seasonal time series, r = 0.57 (0.47 for the monthly time series), represents substantially more shared variance than in the corresponding detrended analysis, r = 0.39 (0.33 for the monthly time series). More importantly, both synchronous and lagged regressions onto the WACS pattern are largely similar between raw and detrended analyses (not shown). In addition, an intraseasonal analysis like in Fig. 9 but performed on raw (not detrended) data returns very similar results to those shown in Fig. 9, suggesting the trend is perhaps not critically important in understanding the DJF WACS pattern.

An analysis of the total seasonally averaged variance associated with the linear trend and in linearly detrended data suggests that at least a third of the observed 1979/80 to 2011/12 variance in the winter Barents Sea ice is associated with the linear trend (Table 2). The trend accounts for a small fraction of the variance in the DJF WACS index (8%) and hardly any of the variance in winter Barents Sea THF (1%). The mismatch in the relative importance of the trend further suggests it is unlikely that the WACS pattern results from a direct linear atmospheric response to sea ice loss or variability. In short, our decision to explore the interannual variability allows for a focus on the mechanistic relationships between the atmospheric circulation anomalies, sea ice variability, and the connection of both to the upper ocean using the maximum number of degrees of freedom.

Table 2.

Fraction of total monthly variance associated with the linear trend (σ2trend) and with the linearly detrended data (σ2detrend). ICEBar, THFBar, and SATBar are the negative sea ice concentration, turbulent heat flux, and surface air temperature, respectively, averaged over the Barents Sea region (67–81°N, 12–73 °E). WACSssyn is a measure of the strength of the WACS pattern (see section 2).

Fraction of total monthly variance associated with the linear trend (σ2trend) and with the linearly detrended data (σ2detrend). ICEBar, THFBar, and SATBar are the negative sea ice concentration, turbulent heat flux, and surface air temperature, respectively, averaged over the Barents Sea region (67–81°N, 12–73 °E). WACSssyn is a measure of the strength of the WACS pattern (see section 2).
Fraction of total monthly variance associated with the linear trend (σ2trend) and with the linearly detrended data (σ2detrend). ICEBar, THFBar, and SATBar are the negative sea ice concentration, turbulent heat flux, and surface air temperature, respectively, averaged over the Barents Sea region (67–81°N, 12–73 °E). WACSssyn is a measure of the strength of the WACS pattern (see section 2).

The goal of identifying the atmospheric response to Barents Sea ice variability in the fully coupled system motivated our study, and it is critical for understanding our results in the context of previous work. Studies based on observations find compelling correlations between sea ice and atmospheric circulation variability (e.g., Honda et al. 2009; Hori et al. 2011; Inoue et al. 2012; Outten and Esau 2012; Mori et al. 2014), but causality is challenging to determine in nature because of the many complicated forcing/response interactions and the relatively few years of data. Studies based on uncoupled modeling experiments do isolate an atmospheric response by design—they prescribe an ice anomaly under an atmospheric model and examine the resulting changes to the atmospheric circulation. While there is undoubtedly a WACS-like response to prescribed Barents Sea ice reductions, the particular result is somewhat model dependent (Honda et al. 2009; Petoukhov and Semenov 2010; Orsolini et al. 2012; Mori et al. 2014). It is also unclear how much of the atmospheric response is due to the prescribed sea ice loss as compared with other surface (SST, snow cover) changes, and how much is directly attributable to changes in radiative forcing. Further complicating the task is the presence of strong natural variability and rich two-way ocean–atmosphere interactions in nature (Deser et al. 2004, 2012; Wettstein and Deser 2014; Screen et al. 2014). Our focus on the surface turbulent heat fluxes allows us to use the reanalysis-based data, which represent the fully coupled system, without having to assume that observed atmospheric circulation changes are purely a response to sea ice forcing. Our results are generally more consistent with studies attributing cold Eurasian winters to atmospheric circulation variability (e.g., Hurrell 1995; Jeong and Ho 2005; Zhang et al. 2008; Cattiaux et al. 2010; Park et al. 2011; Sato et al. 2014) than with those attributing them to Arctic sea ice variability. Relationships between ocean circulation, SST, Barents Sea ice, and Barents Sea THF (Årthun et al. 2012; Sato et al. 2014) are certainly intriguing, but cleanly assessing their relative importance in a coupled WACS response is inherently difficult with low-frequency variability and so few years of data.

The implications of this observational study must be viewed strictly in the context of the fully coupled climate system on interannual time scales. In this context, the ocean, ice, and atmosphere are continuously interacting and none of the three components is obviously dominant. Atmospheric circulation variability is driven by many factors, sea ice being just one of these (see also Deser et al. 2012; Wettstein and Deser 2014). The results collected here demonstrate that the proposed direct link between variability in Barents Sea ice and the WACS pattern is at best an incomplete characterization of nature. Our study suggests that a broader and more rigorous diagnosis of the causes of Eurasian continental temperature extremes is warranted.

5. Summary

Our study examines relationships between Barents Sea ice, turbulent heat flux (THF), and the “warm-Arctic cold-Siberian” (WACS) pattern. Our main findings are as follows:

  • The leading pattern (EOF1) of winter Barents Sea THF variability appears to be driven primarily by atmospheric rather than sea ice variability. The sea ice–related THF variability appears as EOF2.

  • The sea ice–related THF variability (EOF2) is robust over the entire winter season, but its link to cold anomalies over Siberia is not.

  • The WACS pattern is a robust pattern of temperature variability (EOF2 of winter Eurasian SAT) and is associated with weak reductions in Barents Sea ice cover but also reduced ocean-to-atmosphere THF. The sense of these relationships is inconsistent with the proposed pathways of the WACS pattern being a direct atmospheric response to Barents Sea ice reductions.

The role of Barents Sea ice variability in forcing the WACS pattern appears to be minor in the detrended analyses performed here. One consistent interpretation of the results is that large-scale atmospheric circulation variability produces southerly wind anomalies that push the Barents Sea ice edge northward, advect warm air into the region, and reduce Barents Sea THF, and that are ultimately associated with cold Siberian temperatures downstream. The analyses leave open the possibility that Barents Sea ice variability drives an atmospheric response that feeds back on preexisting atmospheric circulation anomalies.

Acknowledgments

Although CL and JJW are listed alphabetically, they contributed equally to the manuscript. We wish to thank J. Kristiansen, P. Kushner, three anonymous reviewers, and the editor for their helpful comments. We acknowledge the European Centre for Medium-Range Weather Forecasts for providing the ERA-Interim data and the National Snow and Ice Data Center for providing the sea ice data. This work was supported by the NordForsk TRI program under project GREENICE 61841 (CL, JJW), the Research Council of Norway under projects CLIMARC 225032 and Bjerknes Compensation 227137 (SAS), and the Centre for Climate Dynamics (SKD) at the Bjerknes Centre (CL). JJW acknowledges support from the G. Unger Vetlesen Foundation.

REFERENCES

REFERENCES
Alexander
,
M. A.
,
U. S.
Bhatt
,
J. E.
Walsh
,
M. S.
Timlin
,
J. S.
Miller
, and
J. D.
Scott
,
2004
:
The atmospheric response to realistic Arctic sea ice anomalies in an AGCM during winter
.
J. Climate
,
17
,
890
905
, doi:.
Andersson
,
A.
,
C.
Klepp
,
K.
Fennig
,
S.
Bakan
,
H.
Grassl
, and
J.
Schulz
,
2011
:
Evaluation of HOAPS-3 ocean surface freshwater flux components
.
J. Appl. Meteor. Climatol.
,
50
,
379
398
, doi:.
Årthun
,
M.
,
T.
Eldevik
,
L. H.
Smedsrud
,
Ø.
Skagseth
, and
R. B.
Ingvaldsen
,
2012
:
Quantifying the influence of Atlantic heat on the Barents Sea ice variability and retreat
.
J. Climate
,
25
,
4736
4743
, doi:.
Bader
,
J.
,
M. D. S.
Mesquita
,
K. I.
Hodges
,
N.
Keenlyside
,
S.
Østerhus
, and
M.
Miles
,
2011
:
A review on Northern Hemisphere sea-ice, storminess and the North Atlantic Oscillation: Observations and projected changes
.
Atmos. Res.
,
101
,
809
834
, doi:.
Barnes
,
E. A.
,
2013
:
Revisiting the evidence linking Arctic amplification to extreme weather in midlatitudes
.
Geophys. Res. Lett.
,
40
,
4734
4739
, doi:.
Barnes
,
E. A.
,
E.
Dunn-Sigouin
,
G.
Masato
, and
T.
Woollings
,
2014
:
Exploring recent trends in Northern Hemisphere blocking
.
Geophys. Res. Lett.
,
41
,
638
644
, doi:.
Barsugli
,
J. J.
, and
D. S.
Battisti
,
1998
:
The basic effects of atmosphere–ocean thermal coupling on midlatitude variability
.
J. Atmos. Sci.
,
55
,
477
493
, doi:.
Bhatt
,
U. S.
,
M. A.
Alexander
,
C.
Deser
,
J. E.
Walsh
,
J. S.
Miller
,
M. S.
Timlin
,
J.
Scott
, and
R. A.
Tomas
,
2008
: The atmospheric response to realistic reduced summer Arctic sea ice anomalies. Arctic Sea Ice Decline: Observations, Projections, Mechanisms, and Implications, Geophys. Monogr., Vol. 180, Amer. Geophys. Union, 91–110, doi:.
Bjerknes
,
J.
,
1964
: Atlantic air-sea interaction. Advances in Geophysics, Vol. 10, Academic Press, 1–82, doi:.
Blüthgen
,
J.
,
R.
Gerdes
, and
M.
Werner
,
2012
: Atmospheric response to the extreme Arctic sea ice conditions in 2007. Geophys. Res. Lett., 39, L02707, doi:.
Budikova
,
D.
,
2009
:
Role of Arctic sea ice in global atmospheric circulation: A review
.
Global Planet. Change
,
68
,
149
163
, doi:.
Cattiaux
,
J.
,
R.
Vautard
,
C.
Cassou
,
P.
Yiou
,
V.
Masson-Delmotte
, and
F.
Codron
,
2010
:
Winter 2010 in Europe: A cold extreme in a warming climate
.
Geophys. Res. Lett.
,
37
,
L20704
, doi:.
Cavalieri
,
D. J.
,
C. L.
Parkinson
,
P.
Gloersen
, and
H. J.
Zwally
,
1996
: Sea ice concentrations from Nimbus-7 SMMR and DMSP SSM/I passive microwave data, version 1. National Snow and Ice Data Center, Boulder, CO, digital media. [Available online at http://nsidc.org/data/nsidc-0051.]
Cayan
,
D. R.
,
1992
:
Latent and sensible heat flux anomalies over the northern oceans: The connection to monthly atmospheric circulation
.
J. Climate
,
5
,
354
369
, doi:.
Cheung
,
H. N.
,
W.
Zhou
,
Y.
Shao
,
W.
Chen
,
H. Y.
Mok
, and
M. C.
Wu
,
2013
:
Observational climatology and characteristics of wintertime atmospheric blocking over Ural–Siberia
.
Climate Dyn.
,
41
,
63
79
, doi:.
Cohen
,
J.
,
M.
Barlow
,
P. J.
Kushner
, and
K.
Saito
,
2007
:
Stratosphere–troposphere coupling and links with Eurasian land surface variability
.
J. Climate
,
20
,
5335
5343
, doi:.
Cohen
,
J.
, and Coauthors
,
2014
:
Recent Arctic amplification and extreme mid-latitude weather
.
Nat. Geosci.
,
7
,
627
637
, doi:.
Comiso
,
J. C.
,
2012
:
Large decadal decline of the Arctic multiyear ice cover
.
J. Climate
,
25
,
1176
1193
, doi:.
Coumou
,
D.
, and
S.
Rahmstorf
,
2012
:
A decade of weather extremes
.
Nat. Climate Change
,
2
,
491
496
, doi:.
Croci-Maspoli
,
M.
, and
H. C.
Davies
,
2009
:
Key dynamical features of the 2005/06 European winter
.
Mon. Wea. Rev.
,
137
,
664
678
, doi:.
Dee
, and Coauthors
,
2011
:
The ERA-Interim reanalysis: Configuration and performance of the data assimilation system
.
Quart. J. Roy. Meteor. Soc.
,
137
,
553
597
, doi:.
Deser
,
C.
,
J. E.
Walsh
, and
M. S.
Timlin
,
2000
:
Arctic sea ice variability in the context of recent atmospheric circulation trends
.
J. Climate
,
13
,
617
633
, doi:.
Deser
,
C.
,
G.
Magnusdottir
,
R.
Saravanan
, and
A.
Phillips
,
2004
:
The effects of North Atlantic SST and sea ice anomalies on the winter circulation in CCM3. Part II: Direct and indirect components of the response
.
J. Climate
,
17
,
877
889
, doi:.
Deser
,
C.
,
R. A.
Tomas
, and
S.
Peng
,
2007
:
The transient atmospheric circulation response to North Atlantic SST and sea ice anomalies
.
J. Climate
,
20
,
4751
4767
, doi:.
Deser
,
C.
,
R. A.
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
, doi:.
Deser
,
C.
,
A. S.
Phillips
,
V.
Bourdette
, and
H.
Teng
,
2012
:
Uncertainty in climate change projections: The role of internal variability
.
Climate Dyn.
,
38
,
527
546
, doi:.
Fang
,
Z.
, and
J. M.
Wallace
,
1994
:
Arctic sea ice variability on a timescale of weeks and its relation to atmospheric forcing
.
J. Climate
,
7
,
1897
1914
, doi:.
Francis
,
J. A.
, and
S. J.
Vavrus
,
2012
:
Evidence linking Arctic amplification to extreme weather in mid-latitudes
.
Geophys. Res. Lett.
,
39
,
L06801
, doi:.
Francis
,
J. A.
,
W.
Chan
,
D. J.
Leathers
,
J. R.
Miller
, and
D. E.
Veron
,
2009
:
Winter Northern Hemisphere weather patterns remember summer Arctic sea-ice extent
.
Geophys. Res. Lett.
,
36
,
L07503
, doi:.
García-Serrano
,
J.
,
C.
Frankignoul
,
G.
Gastineau
, and
A.
de la Cámara
,
2015
:
On the predictability of the winter Euro-Atlantic climate: Lagged influence of autumn Arctic sea ice
.
J. Climate
,
28
,
5195
5216
, doi:.
Germe
,
A.
,
M.-N.
Houssais
,
C.
Herbaut
, and
C.
Cassou
,
2011
:
Greenland Sea sea ice variability over 1979–2007 and its link to the surface atmosphere
.
J. Geophys. Res.
,
116
,
C10034
, doi:.
Ghatak
,
D.
,
A.
Frei
,
G.
Gong
,
J.
Stroeve
, and
D.
Robinson
,
2010
:
On the emergence of an Arctic amplification signal in terrestrial Arctic snow extent
.
J. Geophys. Res.
,
115
,
D24105
, doi:.
Guirguis
,
K.
,
A.
Gershunov
,
R.
Schwartz
, and
S.
Bennett
,
2011
:
Recent warm and cold daily winter temperature extremes in the Northern Hemisphere
.
Geophys. Res. Lett.
,
38
,
L17701
, doi:.
Gulev
,
S. K.
,
M.
Latif
,
N.
Keenlyside
,
W.
Park
, and
K. P.
Koltermann
,
2013
:
North Atlantic Ocean control on surface heat flux on multidecadal timescales
.
Nature
,
499
,
464
467
, doi:.
Honda
,
M.
,
K.
Yamazaki
,
H.
Nakamura
, and
K.
Takeuchi
,
1999
:
Dynamic and thermodynamic characteristics of atmospheric response to anomalous sea-ice extent in the Sea of Okhotsk
.
J. Climate
,
12
,
3347
3358
, doi:.
Honda
,
M.
,
J.
Inoue
, and
S.
Yamane
,
2009
:
Influence of low Arctic sea-ice minima on anomalously cold Eurasian winters
.
Geophys. Res. Lett.
,
36
,
L08707
, doi:.
Hopsch
,
S.
,
J.
Cohen
, and
K.
Dethloff
,
2012
:
Analysis of a link between fall Arctic sea ice concentration and atmospheric patterns in the following winter
.
Tellus
,
64A
,
18624
, doi:.
Hori
,
M. E.
,
J.
Inoue
,
T.
Kikuchi
,
M.
Honda
, and
Y.
Tachibana
,
2011
:
Recurrence of intraseasonal cold air outbreak during the 2009/2010 winter in Japan and its ties to the atmospheric condition over the Barents-Kara Sea
.
SOLA
,
7
,
25
28
, doi:.
Hoskins
,
B. J.
, and
D. J.
Karoly
,
1981
:
The steady linear response of a spherical atmosphere to thermal and orographic forcing
.
J. Atmos. Sci.
,
38
,
1179
1196
, doi:.
Hurrell
,
J. W.
,
1995
:
Decadal trends in the North Atlantic Oscillation: Regional temperatures and precipitation
.
Science
,
269
,
676
679
, doi:.
Inoue
,
J.
,
M. E.
Hori
, and
K.
Takaya
,
2012
:
The role of Barents Sea ice in the wintertime cyclone track and emergence of a warm-Arctic cold-Siberian anomaly
.
J. Climate
,
25
,
2561
2568
, doi:.
Jaiser
,
R.
,
K.
Dethloff
,
D.
Handorf
,
A.
Rinke
, and
J.
Cohen
,
2012
:
Impact of sea ice cover changes on the Northern Hemisphere atmospheric winter circulation
.
Tellus
,
64A
,
11595
, doi:.
Jeong
,
J.-H.
, and
C.-H.
Ho
,
2005
:
Changes in occurrence of cold surges over East Asia in associated with Arctic Oscillation
.
Geophys. Res. Lett.
,
32
,
L14704
, doi:.
Kalnay
,
E.
, and Coauthors
,
1996
:
The NCEP/NCAR 40-Year Reanalysis Project
.
Bull. Amer. Meteor. Soc.
,
77
,
437
471
, doi:.
Kim
,
B.-M.
,
S.-W.
Son
,
S.-K.
Min
,
J.-H.
Jeong
,
S.-J.
Kim
,
X.
Zhang
,
T.
Shim
, and
J.-H.
Yoon
,
2014
:
Weakening of the stratospheric polar vortex by Arctic sea-ice loss
.
Nat. Commun.
,
5
,
4646
, doi:.
King
,
M. P.
,
M.
Hell
, and
N.
Keenlyside
,
2015
:
Investigation of the atmospheric mechanisms related to the autumn sea ice and winter circulation link in the Northern Hemisphere
.
Climate Dyn.
, doi:, in press.
Kvamstø
,
N. G.
,
P.
Skeie
, and
D. B.
Stephenson
,
2004
:
Impact of Labrador sea-ice extent on the North Atlantic oscillation
.
Int. J. Climatol.
,
24
,
603
612
, doi:.
Lau
,
N.-C.
, and
E. O.
Holopainen
,
1984
:
Transient eddy forcing of the time‐mean flow as identified by geopotential tendencies
.
J. Atmos. Sci.
,
41
,
313
328
, doi:.
Liptak
,
J.
, and
C.
Strong
,
2014
:
The winter atmospheric response to sea ice anomalies in the Barents Sea
.
J. Climate
,
27
,
914
924
, doi:.
Magnusdottir
,
G.
,
C.
Deser
, and
R.
Saravanan
,
2004
:
The effects of North Atlantic SST and sea ice anomalies on the winter circulation in CCM3. Part I: Main features and storm track characteristics of the response
.
J. Climate
,
17
,
857
876
, doi:.
Maykut
,
G. A.
,
1982
:
Large-scale heat exchange and ice production in the central Arctic
.
J. Geophys. Res.
,
87
,
7971
7984
, doi:.
Mori
,
M.
,
M.
Watanabe
,
H.
Shiogama
,
J.
Inoue
, and
M.
Kimoto
,
2014
:
Robust Arctic sea-ice influence on the frequent Eurasian cold winters in past decades
.
Nat. Geosci.
,
7
,
869
873
, doi:.
North
,
G. R.
,
T. L.
Bell
,
R. F.
Cahalan
, and
F. J.
Moeng
,
1982
:
Sampling errors in the estimation of empirical orthogonal functions
.
Mon. Wea. Rev.
,
110
,
699
706
, doi:.
Orsolini
,
Y. J.
,
R.
Senan
,
R. E.
Benestad
, and
A.
Melsom
,
2012
:
Autumn atmospheric response to the 2007 low Arctic sea ice extent in coupled ocean–atmosphere hindcasts
.
Climate Dyn.
,
38
,
2437
2448
, doi:.
Outten
,
S.
, and
I.
Esau
,
2012
:
A link between Arctic sea ice and recent cooling trends over Eurasia
.
Climatic Change
,
110
,
1069
1075
, doi:.
Outten
,
S.
,
R.
Davy
, and
I.
Esau
,
2013
:
Eurasian winter cooling: Intercomparison of reanalyses and CMIP5 data sets
.
Atmos. Oceanic Sci. Lett.
,
6
,
324
331
, doi:.
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
9
, doi:.
Overland
,
J. E.
,
K. R.
Wood
, and
M.
Wang
,
2011
:
Warm Arctic – cold continents: Climate impacts of the newly open Arctic Sea
.
Polar Res.
,
30
,
15 787
, doi:.
Park
,
T.-W.
,
C.-H.
Ho
, and
S.
Yang
,
2011
:
Relationship between the Arctic Oscillation and cold surges over East Asia
.
J. Climate
,
24
,
68
83
, doi:.
Parkinson
,
C. L.
, and
D. J.
Cavalieri
,
2012
:
Arctic sea ice variability and trends, 1979–2010
.
Cryosphere Discuss.
,
6
,
931
956
, doi:.
Peng
,
S. L.
,
W. A.
Robinson
, and
M. P.
Hoerling
,
1997
:
The modeled atmospheric response to midlatitude SST anomalies and its dependence on background circulation states
.
J. Climate
,
10
,
971
987
, doi:.
Petoukhov
,
V.
, and
V. A.
Semenov
,
2010
:
A link between reduced Barents–Kara Sea ice and cold winter extremes over northern continents
.
J. Geophys. Res.
,
115
,
D21111
, doi:.
Prinsenberg
,
S. J.
,
I. K.
Peterson
,
S.
Narayanan
, and
J. U.
Umoh
,
1997
:
Interaction between atmosphere, ice cover, and ocean off Labrador and Newfoundland from 1962 to 1992
.
Can. J. Fish. Aquat. Sci.
,
54
,
30
39
, doi:.
Rigor
,
I. G.
,
J. M.
Wallace
, and
R. L.
Colony
,
2002
:
Response of sea ice to the Arctic Oscillation
.
J. Climate
,
15
,
2648
2663
, doi:.
Sato
,
K.
,
J.
Inoue
, and
M.
Watanabe
,
2014
:
Influence of the Gulf Stream on the Barents Sea ice retreat and Eurasian coldness during early winter
.
Environ. Res. Lett.
,
9
,
101003
, doi:.
Scaife
,
A. A.
, and
J. R.
Knight
,
2008
:
Ensemble simulations of the cold European winter of 2005–2006
.
Quart. J. Roy. Meteor. Soc.
,
134
,
1647
1659
, doi:.
Schröder
,
D.
,
T.
Vihma
,
A.
Kerber
, and
B.
Brümmer
,
2003
:
On the parameterization of turbulent surface fluxes over heterogeneous sea ice surfaces
.
J. Geophys. Res.
,
108
,
3195
, doi:.
Screen
,
J. A.
, and
I.
Simmonds
,
2010a
:
Increasing fall-winter energy loss from the Arctic Ocean and its role in Arctic temperature amplification
.
Geophys. Res. Lett.
,
37
,
L16707
, doi:.
Screen
,
J. A.
, and
I.
Simmonds
,
2010b
:
The central role of diminishing sea ice in recent Arctic temperature amplification
.
Nature
,
464
,
1334
1337
, doi:.
Screen
,
J. A.
, and
I.
Simmonds
,
2013
:
Exploring links between Arctic amplification and midlatitude weather
.
Geophys. Res. Lett.
,
40
,
959
964
, doi:.
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
, doi:.
Screen
,
J. A.
,
C.
Deser
,
I.
Simmonds
, and
R.
Tomas
,
2014
:
Atmospheric impacts of Arctic sea-ice loss, 1979–2009: Separating forced change from atmospheric internal variability
.
Climate Dyn.
,
43
,
333
344
, doi:.
Seierstad
,
I. A.
, and
J.
Bader
,
2009
:
Impact of a projected future Arctic sea ice reduction on extratropical storminess and the NAO
.
Climate Dyn.
,
33
,
937
943
, doi:.
Serreze
,
M. C.
,
M. M.
Holland
, and
J. C.
Stroeve
,
2007
:
Perspectives on the Arctic’s shrinking sea-ice cover
.
Science
,
315
,
1533
1536
, doi:.
Serreze
,
M. C.
,
A. P.
Barrett
,
J. C.
Stroeve
,
D. M.
Kindig
, and
M. M.
Holland
,
2009
:
The emergence of surface-based Arctic amplification
.
Cryosphere
,
3
,
11
19
, doi:.
Simmonds
,
I.
, and
K.
Keay
,
2009
:
Extraordinary September Arctic sea ice reductions and their relationships with storm behavior over 1979–2008
.
Geophys. Res. Lett.
,
36
,
L19715
, doi:.
Singarayer
,
J. S.
,
J. L.
Bamber
, and
P. J.
Valdes
,
2006
:
Twenty-first-century climate impacts from a declining Arctic sea ice cover
.
J. Climate
,
19
,
1109
1125
, doi:.
Smoliak
,
B. V.
,
2009
: A Eurasian pattern of Northern Hemisphere wintertime sea level pressure variability. M.S. thesis, Dept. of Atmospheric Sciences, University of Washington, 116 pp.
Sorteberg
,
A.
, and
B.
Kvingedal
,
2006
:
Atmospheric forcing on the Barents Sea winter ice extent
.
J. Climate
,
19
,
4772
4784
, doi:.
Takaya
,
K.
, and
H.
Nakamura
,
2005
:
Mechanisms of intraseasonal amplification of the cold Siberian high
.
J. Atmos. Sci.
,
62
,
4423
4440
, doi:.
Tang
,
Q.
,
X.
Zhang
,
X.
Yang
, and
J. A.
Francis
,
2013
:
Cold winter extremes in northern continents linked to Arctic sea ice loss
.
Environ. Res. Lett.
,
8
,
014036
, doi:.
Thompson
,
D. W. J.
, and
J. M.
Wallace
,
1998
:
The Arctic Oscillation signature in the wintertime geopotential height and temperature fields
.
Geophys. Res. Lett.
,
25
,
1297
1300
, doi:.
Wallace
,
J. M.
, and
D. S.
Gutzler
,
1981
:
Teleconnections in the geopotential height field during the Northern Hemisphere winter
.
Mon. Wea. Rev.
,
109
,
784
812
, doi:.
Wallace
,
J. M.
,
I. M.
Held
,
D. W. J.
Thompson
,
K. E.
Trenberth
, and
J. E.
Walsh
,
2014
:
Global warming and winter weather
.
Science
,
343
,
729
730
, doi:.
Wettstein
,
J. J.
, and
J. M.
Wallace
,
2010
:
Observed patterns of month-to-month storm track variability and their relationship to the background flow
.
J. Atmos. Sci.
,
67
,
1420
1437
, doi:.
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
, doi:.
Woollings
,
T.
,
B.
Harvey
, and
G.
Masato
,
2014
:
Arctic warming, atmospheric blocking and cold European winters in CMIP5 models
.
Environ. Res. Lett.
,
9
,
014002
, doi:.
Wu
,
Q.
, and
X.
Zhang
,
2010
:
Observed forcing-feedback processes between Northern Hemisphere atmospheric circulation and Arctic sea ice coverage
.
J. Geophys. Res.
,
115
,
D14119
, doi:.
Yu
,
L.
,
X.
Jin
, and
R. A.
Weller
,
2008
: Multidecade global flux datasets from the objectively analyzed air-sea fluxes (OAFlux) project: Latent and sensible heat fluxes, ocean evaporation, and related surface meteorological variables. Woods Hole Oceanographic Institution OAFlux Project Tech. Rep. OA-2008-01, 64 pp.
Zhang
,
X.
,
A.
Sorteberg
,
J.
Zhang
,
R.
Gerdes
, and
J. C.
Comiso
,
2008
:
Recent radical shifts of atmospheric circulations and rapid changes in Arctic climate system
.
Geophys. Res. Lett.
,
35
,
L22701
, doi:.

Footnotes

1

The first EOF (40%) of Eurasian SAT is associated with the temperature signature of the North Atlantic/Arctic Oscillation (cf. Thompson and Wallace 1998).