1. Introduction
The Scandinavian (SCA) pattern is one of the leading modes of atmospheric variability stretching from the North Atlantic to the Eurasian continent during the boreal winter (Barnston and Livezey 1987). It is manifested as a tripole pattern with a primary anticyclonic anomaly over the Scandinavian Peninsula and cyclonic anomalies over the northeastern Atlantic and central Siberia during its positive phase (Bueh and Nakamura 2007). As a result, fluctuation of the SCA pattern exerts profound impacts on climate over the Eurasian continent. For instance, the positive SCA phase coincides with below-average surface air temperature over central Russia (Sohn et al. 2011; Liu et al. 2014; Yu et al. 2017) and facilitates southward intrusions of cold waves into East Asia and the South China Sea (Zhou et al. 2009; Bueh et al. 2011; Shi et al. 2019; Yuan and Li 2019; Pang et al. 2022; Zhang et al. 2022). It is also responsible for anomalous precipitation, which is enhanced over southern Europe but suppressed over northern Europe (Casanueva et al. 2014; Hernández et al. 2015; Cook et al. 2016; Łupikasza and Cielecka-Nowak 2020).
Previous studies have been devoted to understanding the dynamical processes and mechanisms of the SCA pattern. On the one hand, it is suggested that the SCA pattern is an inherent atmospheric mode, which is excited and maintained by feedback forcing from transient eddy activity along the storm track (Bueh and Nakamura 2007). The extraction of kinetic and available potential energy from basic flows is also an important source in the growth of the SCA pattern (Kim et al. 2021). On the other hand, the SCA pattern could be modulated by other climate factors, such as sea surface temperature (Jung et al. 2017) and convective heating (Wang and Tan 2020) over the western North Atlantic. Apart from the extratropical origins, tropical variability is thought to be an important source in inducing teleconnection patterns in the mid–high latitude (e.g., Hoskins and Karoly 1981; Okumura et al. 2001; Manola et al. 2013; Scaife et al. 2017; Maidens et al. 2021). However, the role of tropical forcing on the SCA pattern is yet to be fully clarified.
A number of studies have examined the variability of the SCA pattern on intraseasonal or interannual time scales, but few works reported its interdecadal variations. A possible decadal change in the SCA pattern occurred in the early 2000s after the pattern was mainly in its negative phase during the previous two decades (Liu et al. 2014). In addition, circulation anomalies relevant to the recent cooling over Eurasia resemble the positive phase of the SCA pattern (e.g., Mori et al. 2014; Crasemann et al. 2017; Sui et al. 2020). In this regard, it is worth examining whether the SCA varies at decadal time scales. If so, what is the possible physical mechanism responsible for this change? We note that the limited length of reanalyses causes a major challenge to explore the low-frequency variation and possible drivers, which may not be sufficient to draw robust conclusions. To address these issues, model outputs from phase 6 of the Coupled Model Intercomparison Project (CMIP6; Eyring et al. 2016) are analyzed in this work, which have been evaluated to yield realistic spatial patterns of atmospheric modes over the Euro-Atlantic sector, including the SCA (Cusinato et al. 2021).
This study aims to investigate the interdecadal variability and physical mechanisms of the SCA pattern based on observations and coupled models. The remainder of this paper is arranged as follows. Section 2 introduces the details of reanalysis, model outputs, and methods. Section 3 analyzes the interdecadal variations of the SCA and associated climatic changes in observations. Section 4 further explores the simulated results in CMIP6 models and proposes the corresponding drivers of the interdecadal variability. Finally, conclusions and discussions are presented in section 5.
2. Data, model outputs, and methods
a. Reanalysis
The monthly mean of observed atmospheric data used in this study are obtained from the fifth major global reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ERA5; Hersbach et al. 2020; Bell et al. 2021) covering the period 1950–2021. The variables are available at a horizontal resolution of 2.5° × 2.5°. In addition, the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR; Kalnay et al. 1996) and the Japanese 55-year Reanalysis (JRA-55; Kobayashi et al. 2015) dataset are also used to verify the robustness of interdecadal variations. The surface air temperature (Ts) and precipitation data are adopted from the Climatic Research Unit (CRU) time series, version 4.05, of high-resolution gridded data with a horizontal resolution of 0.5° × 0.5° (Harris et al. 2020).
b. Model outputs
The model outputs from CMIP6 are used in this study, including historical and preindustrial control (piControl) simulations (Eyring et al. 2016). The multimodel ensemble (MME) mean of historical simulation reveals the response to external forcing, while the results of piControl simulations reflect the internal variability of the climate system. Here, the MME mean is first employed by the single-model ensemble mean and then averaged over the 35 models. By averaging over the ensemble members from individual models, the internal variability will be largely offset and the response to external forcing will remain (e.g., Frankcombe et al. 2015; Kravtsov and Callicutt 2017; Frankcombe et al. 2018). As for the piControl simulation, the external forcing is set to be constant, and thus the internal variability is regarded as the only source. Given the different run length in different models, the periods of 1850–2014 for historical simulations and the last 500 years for piControl experiments are selected. Table 1 lists the details for the 35 models used in this study. All the variables of model outputs are interpolated to a horizontal resolution of 2.5° × 2.5°, consistent with the observations.
CMIP6 models used in this study.
c. Methods
3. Observed interdecadal variations of the SCA pattern
Figure 1 shows the large-scale circulation anomalies regressed onto the SCA index in observations. A clear teleconnection pattern is observed over the Eurasian continent in the upper troposphere, with a major positive height center over the Scandinavian Peninsula and two negative subcenters over western Europe and central Siberia, respectively (Fig. 1a). Hereafter, this phase is referred to as the positive phase of the SCA. Meanwhile, the lower troposphere is dominated by anticyclonic anomalies over a vast area from eastern Europe to western Siberia, along with cyclonic ones to the west of the Iberian Peninsula (Fig. 1b). Induced by the associated north easterlies in the lower troposphere, an extensive cooling occupies the majority of north Asia from the Ural Mountains to eastern Siberia (Fig. 1c). The anomalous anticyclone also suppresses precipitation over northern Europe (Fig. 1d). The above features are in agreement with the typical SCA pattern in previous studies (e.g., Bueh and Nakamura 2007; Sohn et al. 2011; Liu et al. 2014; Yu et al. 2017).
The SCA pattern and its surface climate signature. The regressions of (a) 300-hPa geopotential height (Z300; m), (b) sea level pressure (SLP; shading; hPa) and 850-hPa horizontal wind (vectors; m s−1), (c) surface air temperature (Ts; K), and (d) precipitation (Pr; mm month−1) anomalies onto the observed SCA index. Values significant at the 95% confidence level are dotted, and vectors are shown as thick and black when they are significant in at least one direction.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
Figure 2a displays the observed time series of the SCA index and its interdecadal variations. It is found that the SCA index shows clear low-frequency variability, suggesting a remarkable phase change on decadal time scales. Specifically, the SCA pattern is in the interdecadal negative phase (SCA−) from the late 1970s to early 2000s, but in its interdecadal positive phase (SCA+) afterward. A significant abrupt point is identified in 2003 by applying the moving t test with a 15-yr window (p < 0.05; Fig. 2b). A second interdecadal change seems to have occurred around the late 1970s, but it is weaker and less prominent. This result is also insensitive to removal of the linear trend (not shown). Hereafter, the corresponding change is calculated as the difference between 15-yr after and before 2003, that is, 2004–18 and 1988–2002. The above features are almost identical to those in the other two datasets (Figs. 2c–f). The interdecadal SCA index is also highly correlated among three reanalyses during their overlapping periods of 1958–2020 (r > 0.95). Because of the robustness in observations, only the results based on ERA5 are illustrated in the following analyses.
Low- and high-frequency variability of the SCA. (a) Time series of the SCA index (bars) and its interdecadal component (9-yr running average; lines). (b) Moving t test (15-yr window; solid lines) and 95% significance level (dashed lines) based on the ERA5 reanalysis. (c)–(f) As in (a) and (b), but for the NCEP–NCAR and JRA-55 reanalyses.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
Figure 3 demonstrates that differences in the regional Eurasian climate between the early twenty-first century and late twentieth century can be explained by the interdecadal change of the SCA index around 2003. In general, the circulation anomalies are identical to the regressions on the SCA index shown in Fig. 1. The upper troposphere is characterized as an SCA+ pattern, in agreement with the strengthening of the SCA index (Fig. 3a). Along with that, the lower troposphere exhibits concurrent change, featured as an anticyclone over northern Eurasia (Fig. 3b). Accompanying the changes of the large-scale circulation, extensive cold anomalies are observed over Siberia, and warm anomalies occur over the Arctic (Fig. 3c). The anomalies resemble to the warm-Arctic cold-Siberian pattern, which is reported as a pronounced trend of surface temperature since the 1990s (Cohen et al. 2014; Mori et al. 2014). However, the cooling remains insignificant when the long-term trend is removed over the periods of 1950–2021 (figure not shown). This result implies that the interdecadal variation of surface temperature over Siberia is unlikely to be resulted from global warming and instead is a result of the low-frequency change in SCA. The signals for rainfall are slightly different from the regression, with increased precipitation over southern Europe and decreased rainfall over the Ural Mountains (Fig. 3d).
Differences between the late twentieth and early twenty-first century associated with a change in the SCA pattern. Differences of (a) Z300 (m), (b) SLP (shading; hPa) and 850-hPa horizontal wind (vectors; m s−1), (c) Ts (K), and (d) Pr (mm month−1) between the periods of 2004–18 and 1988–2002. Values significant at the 95% confidence level are dotted, and vectors are shown as thick and black when they are significant in at least one direction.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
Figure 4 further examines the relationship between the interdecadal change of surface temperature and the SCA. The variance of low-frequency surface temperature explained by the SCA index appears to be large over an extensive area from the Caspian Sea to eastern Siberia (Fig. 4a). The maximum in explained variance can reach up to 80% over western Siberia. The area with large variance corresponds well to the region where cold anomalies occur (Fig. 1c). Accordingly, the Ts index is calculated as area-averaged Ts anomalies over (40°–65°N, 50°–130°E). It can be seen that the Ts index is closely related to the SCA index (Fig. 4b). The correlation coefficient between two indices is −0.69, which is significant at the 99% confidence level. Furthermore, the Ts index exhibits obvious variations on decadal time scale, which is synchronous to the change in SCA index. The correlation is much higher between their interdecadal components (−0.88), suggesting an important role of the low-frequency SCA on modulating the variability of surface temperature over Siberia. The impact on low-frequency precipitation is also examined (figure not shown). It seems that the relationship between their interdecadal variations is weaker and less prominent than the high-frequency ones.
SCA pattern explains low-frequency Eurasian temperature variability. (a) Variances (%) of 9-yr running averaged Ts explained by the interdecadal component of the SCA index. The box indicates the area for defining Ts index. (b) Time series of the normalized SCA (gray) and Ts (red) index (bars) and curves refer to their 9-yr running average, respectively.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
4. Simulated interdecadal variations of the SCA
a. The role of external forcing
We first assess the capability of models to reproduce the SCA pattern. Figure 5 shows a Taylor diagram of 300-hPa Z anomalies associated with the SCA for the historical simulations in 35 models and their MME mean. The diagram includes pattern correlation coefficient (R) and ratio of spatial standard deviations (σ) between individual model simulations and observations (Taylor 2001). The two parameters reveal the performance of models in capturing the spatial structure and amplitude of the observed SCA, respectively. The better simulations agree with observations, the closer they will be to the reference point (REF). The result generally shows a good performance in most models: the correlations are greater than 0.7 and standard deviations fall into the range of 0.75–1.25. Moreover, a noticeable improvement can be seen from the MME mean in reproducing both the spatial structure (R = 0.86) and amplitude (σ = 1.0) of the SCA. However, it should be noted that the results show some dispersion among individual models, which indicate their different abilities in simulating the SCA pattern.
Ability of climate models to represent the SCA pattern. Taylor diagram of Z300 anomalies regressed onto SCA index, with observations as REF. The red crosses and blue dot refer to the 35 models and MME mean, respectively.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
Figure 6 displays the circulation anomalies regressed onto the SCA index based on the MME mean historical simulation. Overall, the features of the SCA pattern are well represented, suggesting that the models can well capture the observed circulation and climate impact. Compared with the results in reanalysis (Fig. 1), the amplitudes of circulation are equivalent in both the upper and lower troposphere, while the spatial patterns shift slightly southeastward in models (Figs. 6a,b). This is consistent with the results shown in the Taylor diagram that the standard deviation is close to one unit, but the pattern correlation seems to be lower (Fig. 5). In addition, the simulated amplitudes of the SCA-related surface air temperature and precipitation anomalies are also comparable to the observations (Figs. 6c,d). The results in individual models are also examined that some models overestimate the intensity of the SCA, while others underestimate, indicating the role of internal variability (figure not shown).
Pattern of SCA anomalies in multimodel historical simulations. The regressions of (a) Z300 (m), (b) SLP (shaded; hPa), (c) Ts (K), and (d) Pr (mm month−1) anomalies onto SCA index based on the MME mean in historical simulation. Values where the multimodel consistency exceeds 90% are dotted, and vectors are shown as thick and black when they meet the criteria in at least one direction.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
Figure 7 illustrates the spatial differences of MME mean between the early twenty-first century and late twentieth century in response to external forcing. Here, the global mean has been removed from spatial differences as significant warming is present on the global scale. The differences are calculated between the periods of 2004–13 and 1988–2002 because of the short period of model outputs. The interdecadal changes of large-scale circulation are relatively weak in historical simulations and show large differences with the anomalies associated with the strengthening of the SCA index in observations (Fig. 3). The upper troposphere shows positive height anomalies over the Arctic and negative ones over the North Atlantic (Fig. 7a). Meanwhile, the lower troposphere is dominant by a weak cyclonic anomaly over the North Atlantic (Fig. 7b). Concurrently, the change of surface air temperature and precipitation bear some discrepancy with the observations as well (Figs. 7c,d). Assuming the model responses are realistic (but see e.g., Scaife and Smith 2018), it is indicated that the SCA pattern shows only a weak response to external forcing on these time scales.
Forced changes in climate estimated from the multimodel mean. The differences of (a) Z300 (m), (b) SLP (shaded; hPa), (c) Ts (K), and (d) Pr (mm month−1) between the periods of 2004–13 and 1988–2002 in MME mean. Values significant at the 95% confidence level are dotted, and vectors are shown as thick and black when they are significant in at least one direction.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
Figure 8 further compares the interdecadal changes of the SCA index between models and observation. It can be concluded that the SCA index exhibits a comparable variation to that in observations in some models but that it is very weak in the MME, in agreement with the results in Fig. 7. For the MME mean, the value only accounts for 5.7% of that in observations. Thus, it is inferred that the external forcing is unlikely to be the main driver of the recent interdecadal change in SCA, assuming the models have responded realistically to decadal forcing (Smith et al. 2020). The weak response is due to inconsistent change among single model ensemble (SME) means, which reveals the uncertainty of historical simulations under the given external forcing. For instance, only two models are able to simulate over a half magnitude of change as seen in observations, and the results are offset by those opposite responses when they are averaged as the MME mean. Moreover, the SME means are accompanied by a large spread among different ensemble members. Some of them exhibit comparable change as observations, but others show completely contrary change of the SCA index. Based on this result, the great diversity suggests that the observed interdecadal change of the SCA appears to be explained by unforced internal variability.
Modeled and observed changes in the SCA pattern. The differences of the SCA index between the periods of 2004–13 and 1988–2002 in observations (red bar), the MME mean (gray bar), single model ensemble means (white bars), and each member (dots) under the historical simulations. The dashed line indicates 50% of the observed change.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
b. The role of internal variability
The role of internal variability in the interdecadal change of the SCA is investigated by the outputs of piControl simulations. Similar to the definition in observations, the interdecadal change is identified by the moving t test with a 15-yr window at the 95% significance level, but the time interval between two adjacent cases in the same model should be at least 30 years. The abrupt point in each case is defined as a reference when the t value reaches its peak. As a result, two types of interdecadal changes are detected, that is, the weakening cases refer to the phase shifting from SCA+ to SCA− and the strengthening cases for the reverse transition. Thus, these strengthening cases correspond to the change around 2003 in observations. Accordingly, there are 126 weakening cases and 131 strengthening cases, respectively. The numbers of cases in each model are listed in Table S1 in the online supplemental material. Hereafter, the multicase ensemble (MCE) mean is performed, and the relevant change is measured by the difference between 15-yr subperiods before and after the abrupt points.
Figure 9 illustrates the temporal evolutions of the SCA index based on the selected cases defined above. It can be found that the interdecadal components of the SCA index are steadily decreasing in the weakening cases (Fig. 9a), and increasing in the strengthening ones (Fig. 9b), indicating the interdecadal shifts between SCA phases. More importantly, the 9-yr running averaged indices in piControl simulations show similar features as those in observations, especially for the MCE means (Fig. 9b). In addition, the difference between two 15-yr subperiods in the MCE means (0.79) is equivalent to the change around 2003 in observations (0.71). The comparable change of the SCA index in piControl simulations is distinguished from the weak response to external forcing (Fig. 8). Thus, it is implied that the internal variability plays a dominant role on the interdecadal variations of the SCA and that the internal variability in models is similar in magnitude to that in observations.
Model transitions between interdecadal SCA phases. Time series of the interdecadal SCA index (gray) and the MCE means for selected (a) weakening and (b) strengthening cases in the piControl simulation. The red curve in (b) stands for the observed results around 2003.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
Figure 10 displays the spatial differences between two subperiods for these interdecadal cases. The piControl simulation can well represent the interdecadal variations of the SCA. For the weakening cases, the upper troposphere is characterized as a SCA− pattern, with negative height anomalies over the Scandinavian Peninsula (Fig. 10a). Consistent with the circulation anomalies are the seesaw patterns of surface air temperature with cold anomalies over the Arctic and warm ones over central Siberia (Fig. 10c). In contrast, for the strengthening cases, the associated circulation and surface temperature are characterized as opposite changes (Figs. 10b,d). The SCA+ anomalies in the upper troposphere and warm-Arctic cold-Siberian anomalies near the surface both exhibit close resemblance to the observed changes around 2003 (Figs. 3a,c), with the pattern correlations exceeding 0.7. In addition, the majority of cases display similar features to those demonstrated in the MCE mean, indicating robust changes of large-scale circulation and climatic impact relevant to the transition of the SCA pattern.
Climate impacts of model transitions in the SCA pattern. Differences of (a),(b) Z300 (m) and (c),(d) Ts (K) between two 15-yr subperiods for selected weakening (left) and strengthening (right) cases in the MCE mean. Dots mark the regions where values are significant at 95% confidence level and the multicase consistency exceeds 60%.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
c. Mechanism for the interdecadal variations
Figure 11 shows the differences of precipitation between two subperiods for the selected cases in piControl simulations. It is found that the change of precipitation is not only evident in the extratropics, but also prominent in the tropics. The differences in the mid–high latitudes are dominant over the North Atlantic–European sector, with alternate anomalies over the Greenland–Barents Seas, western Europe, and the Mediterranean, respectively. This is in agreement with the regression associated with the overall SCA (Fig. 6d). More importantly, anomalous precipitation appears over the tropical Atlantic, which is significantly decreased in the weakening SCA cases (Fig. 11a), but increased in the strengthening cases (Fig. 11b). Additionally, the changes display high consistency among cases, as more than 60% of them show the same sign with the MCE means. The above results suggest that the SCA is closely linked to the variability of precipitation over the tropical Atlantic on decadal time scales, which might have a potential contribution to the transitions of the SCA. Precipitation anomalies also seem to occur over the southern Indian Ocean and tropical western Pacific, but the changes are less significant and consistent among cases.
Precipitation anomalies of model transitions in the SCA pattern. Differences of precipitation (mm month−1) between two 15-yr subperiods for selected (a) weakening and (b) strengthening cases in the MCE mean. Dots mark the regions where values are significant at the 95% confidence level and the multicase consistency exceeds 60%. The boxes refer to the tropical Atlantic (5°S–5°N, 30°W–15°E) used to define the PI.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
Figure 12 examines the relationship between the low-frequency variability of the SCA and tropical precipitation in piControl simulations. A precipitation index (PI) is defined as the domain-averaged anomalies over the tropical Atlantic (5°S–5°N, 30°W–15°E) shown as boxes in Fig. 11. The 9-yr running average is applied to both SCA and PI indices. It is shown that the correlation coefficients are positive in almost all models (34/35) and prominent (p < 0.05) in 23/35 of them (Fig. 12a). In other words, the weakening (strengthening) of the SCA is significantly related to the reduced (enhanced) convection over the tropical Atlantic, in agreement with the results in MCE mean (Fig. 11). In addition, the lead-to-lag correlations are further analyzed to confirm this inferred linkage (Fig. 12b). It turns out that the relationship is maximized and significant at lag 0 in the MME mean, which implies that the tropical convection exhibits a synchronous transition with the SCA pattern at decadal time scales. The simultaneous correlation is consistent with the relatively short time for Rossby waves to propagate from the tropics to the extratropics and is also found in most models, confirming the robust linkage between the interdecadal variations of tropical convection and the SCA pattern (figure not shown).
Relationship between SCA and tropical convection over the Atlantic. (a) The cross-correlation coefficients between 9-yr running averaged PI and SCA indices for (a) each model at lag 0 and (b) the MME mean in piControl simulations. Gray bars in (a) and dashed line in (b) represent values significant at 95% confidence level. The positive value of the x axis in (b) indicates that the PI index leads the SCA index, and vice versa.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
Figure 13 further represents the ensemble means of low-frequency variables regressed onto the PI index. Here, the 9-yr running average is employed to retain the interdecadal component. The results represent the extratropical changes accompanying the variation of precipitation anomalies over the tropical Atlantic. The change of PI index is associated with enhanced convection over the tropical Atlantic and suppressed convection over the south Indian Ocean (Fig. 13a), being consistent with the tropical changes relevant to the MCE mean interdecadal strengthening cases (Fig. 11b). These can also be verified by the outgoing longwave radiation since the tropical precipitation intensifies the deep convection (Fig. 13b). Accompanied with that, the diabatic heating causes intensified divergent flows in the upper troposphere and further induce anomalous RWS over the northern Africa (Fig. 13c). As a result, a poleward propagation of Rossby wave is triggered and a wavelike pattern extending northeastward from western Africa to the Mediterranean and Scandinavia with alternating anomalies is clearly observed in the upper troposphere (Fig. 13d). The responses in the upper troposphere show similar structures with the SCA+ anomalies and thus favor the decadal shift of the SCA phases. The above processes how tropical convection affects extratropical circulation are in agreement with previous studies (e.g., Hoskins and Karoly 1981; Scaife et al. 2017; Dunstone et al. 2018; Li et al. 2020) and a clear pathway for Rossby wave propagation can be seen from the tropical Atlantic to the Scandinavia in response to the precipitation variations over tropical Atlantic.
Circulation anomalies coincident with tropical convection changes. The MME mean of the regression maps of (a) Pr (mm month−1), (b) top-of-atmospheric outgoing longwave radiation (OLR; W m−2), (c) 200-hPa Rossby wave source (RWS; shaded; 10−12 s−2) and divergent flows (vectors; m s−1), and (d) 300-hPa streamfunction (shaded; 106 m2 s−1) and horizontal winds (vectors; m s−1) onto the PI in piControl simulations. Values are shown where the multimodel consistency exceeds 60%. Both variables and PI index are performed by 9-yr running average before the regressions.
Citation: Journal of Climate 36, 10; 10.1175/JCLI-D-22-0494.1
5. Summary and discussion
In this study, the interdecadal variations of the SCA are investigated based on both the observational datasets and simulated outputs from coupled models within CMIP6. The results show that the SCA pattern undergoes a significant phase shift from negative to positive in the early 2000s, which is highly consistent in the different reanalysis. The corresponding changes in the upper troposphere are featured as positive height anomalies over the Scandinavian Peninsula, along with negative and weak anomalies over western Europe and central Siberia, respectively, in consistent with the SCA+ anomalies. As a result, the interdecadal change of the SCA contributes to an extensive cooling over the Eurasia after the early 2000s. The variance of surface air temperature explained by the interdecadal component of the SCA reaches up to 60% over the western Siberia, suggesting its crucial role in modulating the surface climate on the decadal time scale.
Furthermore, the results based on model outputs are demonstrated from the historical and piControl simulations. The performance of coupled models is first assessed and the observed large-scale circulation and climatic effects of the SCA can be well captured. However, the interdecadal change of the SCA in historical simulations can only explain a small account of observed variability, along with a large spread of responses among different models. It is indicated that the interdecadal variation of the SCA is unlikely to be driven by external forcing but is likely dominated by internal variability. Therefore, the interdecadal SCA cases are selected from the 500-yr piControl experiments. The simulated changes of the SCA index and relevant circulation in the strengthening cases are comparable to the observed cases around the early 2000s. Additionally, it is found that the change of tropical convection over the Atlantic is a likely cause for the interdecadal variation of the SCA. The enhanced precipitation over the tropical Atlantic intensifies the geopotential height over the Scandinavian Peninsula by exciting a northeastward propagating Rossby wave train in the upper troposphere.
The present results suggest a role of tropical Atlantic convection in the interdecadal transition of the SCA pattern. The observed changes in precipitation over the tropical Atlantic are enhanced after the early 2000s as well, bearing a similarity to those in models (Fig. 11b). It is inferred that the intensified tropical convection may be related to the sea surface warming in the tropical Atlantic, consistent with the warm phase of the Atlantic multidecadal oscillation (AMO) in recent decades (e.g., Tokinaga and Xie 2011; Servain et al. 2014; Li et al. 2016). However, the sea surface temperature anomalies associated with the interdecadal transition of the SCA pattern are weak over the North Atlantic (Fig. S1). Besides, the correlations between the SCA and AMO indices exhibit strong uncertainty in different models based on piControl simulations, and thus the relationship tends to be not robust (Fig. S2). In addition, it has also been suggested that the recent cooling over the midlatitude Eurasia is linked to the loss of sea ice over the Arctic (e.g., Cohen et al. 2014). However, the simulated responses in models are inconsistent and it is hard to establish causality (e.g., Cohen et al. 2020; Smith et al. 2022) and some studies point to tropical convection for year-to-year changes in sea ice (e.g., Warner et al. 2020). Further evidence is needed to investigate whether these decadal variations arise mainly from the Arctic or the tropics, as suggested here.
Acknowledgments.
The authors appreciate the editor and three anonymous reviewers for their constructive and detailed comments, which greatly improved the presentation. This research was jointly sponsored by the National Natural Science Foundation of China (Grants 42105020, 41721004, and 42075052). AAS was supported by the U.K.–China Research and Innovation Partnership Fund through the Met Office Climate Science for Service Partnership (CSSP) China as part of the Newton Fund.
Data availability statement.
All the data used in this study are openly available. The ERA5 data are obtained from https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5, the NCEP–NCAR data are obtained from https://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.html, and the JRA-55 data are obtained from https://jra.kishou.go.jp/JRA-55/index_en.html. The CRU data are retrieved from https://data.ceda.ac.uk/badc/cru/data/cru_ts/cru_ts_4.05. Model outputs within CMIP6 are obtained from https://esgf-index1.ceda.ac.uk/search/cmip6-ceda/.
REFERENCES
Barnston, A. G., and R. E. Livezey, 1987: Classification, seasonality and persistence of low-frequency atmospheric circulation patterns. Mon. Wea. Rev., 115, 1083–1126, https://doi.org/10.1175/1520-0493(1987)115<1083:CSAPOL>2.0.CO;2.
Bell, B., and Coauthors, 2021: The ERA5 global reanalysis: Preliminary extension to 1950. Quart. J. Roy. Meteor. Soc., 147, 4186–4227, https://doi.org/10.1002/qj.4174.
Bretherton, C. S., M. Widmann, V. P. Dymnikov, J. M. Wallace, and I. Bladé, 1999: The effective number of spatial degrees of freedom of a time-varying field. J. Climate, 12, 1990–2009, https://doi.org/10.1175/1520-0442(1999)012<1990:TENOSD>2.0.CO;2.
Bueh, C., and H. Nakamura, 2007: Scandinavian pattern and its climatic impact. Quart. J. Roy. Meteor. Soc., 133, 2117–2131, https://doi.org/10.1002/qj.173.
Bueh, C., N. Shi, and Z. Xie, 2011: Large-scale circulation anomalies associated with persistent low temperature over southern China in January 2008. Atmos. Sci. Lett., 12, 273–280, https://doi.org/10.1002/asl.333.
Casanueva, A., C. Rodríguez-Puebla, M. D. Frías, and N. González-Reviriego, 2014: Variability of extreme precipitation over Europe and its relationships with teleconnection patterns. Hydrol. Earth Syst. Sci., 18, 709–725, https://doi.org/10.5194/hess-18-709-2014.
Chu, J.-E., K.-J. Ha, J.-Y. Lee, B. Wang, B.-H. Kim, and C. E. Chung, 2014: Future change of the Indian Ocean basin-wide and dipole modes in the CMIP5. Climate Dyn., 43, 535–551, https://doi.org/10.1007/s00382-013-2002-7.
Cohen, J., and Coauthors, 2014: Recent Arctic amplification and extreme mid-latitude weather. Nat. Geosci., 7, 627–637, https://doi.org/10.1038/ngeo2234.
Cohen, J., and Coauthors, 2020: Divergent consensuses on Arctic amplification influence on midlatitude severe winter weather. Nat. Climate Change, 10, 20–29, https://doi.org/10.1038/s41558-019-0662-y.
Cook, B. I., K. J. Anchukaitis, R. Touchan, D. M. Meko, and E. R. Cook, 2016: Spatiotemporal drought variability in the Mediterranean over the last 900 years. J. Geophys. Res. Atmos., 121, 2060–2074, https://doi.org/10.1002/2015JD023929.
Crasemann, B., D. Handorf, R. Jaiser, K. Dethloff, T. Nakamura, J. Ukita, and K. Yamazaki, 2017: Can preferred atmospheric circulation patterns over the North-Atlantic-Eurasian region be associated with Arctic sea ice loss? Polar Sci., 14, 9–20, https://doi.org/10.1016/j.polar.2017.09.002.
Cusinato, E., A. Rubino, and D. Zanchettin, 2021: Winter Euro-Atlantic climate modes: Future scenarios from a CMIP6 multi-model ensemble. Geophys. Res. Lett., 48, e2021GL094532, https://doi.org/10.1029/2021GL094532.
Dunstone, N., and Coauthors, 2018: Predictability of European winter 2016/2017. Atmos. Sci. Lett., 19, e868, https://doi.org/10.1002/asl.868.
Eyring, V., S. Bony, G. A. Meehl, C. A. Senior, B. Stevens, R. J. Stouffer, and K. E. Taylor, 2016: Overview of the coupled model intercomparison project phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev., 9, 1937–1958, https://doi.org/10.5194/gmd-9-1937-2016.
Frankcombe, L. M., M. H. England, M. E. Mann, and B. A. Steinman, 2015: Separating internal variability from the externally forced climate response. J. Climate, 28, 8184–8202, https://doi.org/10.1175/JCLI-D-15-0069.1.
Frankcombe, L. M., M. H. England, J. B. Kajtar, M. E. Mann, and B. A. Steinman, 2018: On the choice of ensemble mean for estimating the forced signal in the presence of internal variability. J. Climate, 31, 5681–5693, https://doi.org/10.1175/JCLI-D-17-0662.1.
Harris, I., T. J. Osborn, P. Jones, and D. Lister, 2020: Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Sci. Data, 7, 109, https://doi.org/10.1038/s41597-020-0453-3.
Hernández, A., and Coauthors, 2015: Sensitivity of two Iberian lakes to North Atlantic atmospheric circulation modes. Climate Dyn., 45, 3403–3417, https://doi.org/10.1007/s00382-015-2547-8.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
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, https://doi.org/10.1175/1520-0469(1981)038<1179:TSLROA>2.0.CO;2.
Jung, O., M.-K. Sung, K. Sato, Y.-K. Lim, S.-J. Kim, E.-H. Baek, J.-H. Jeong, and B.-M. Kim, 2017: How does the SST variability over the western North Atlantic Ocean control Arctic warming over the Barents–Kara Seas? Environ. Res. Lett., 12, 034021, https://doi.org/10.1088/1748-9326/aa5f3b.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437–471, https://doi.org/10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.
Kim, M., C. Yoo, M.-K. Sung, and S. Lee, 2021: Classification of wintertime atmospheric teleconnection patterns in the Northern Hemisphere. J. Climate, 34, 1847–1861, https://doi.org/10.1175/JCLI-D-20-0339.1.
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.
Kravtsov, S., and D. Callicutt, 2017: On semi-empirical decomposition of multidecadal climate variability into forced and internally generated components. Int. J. Climatol., 37, 4417–4433, https://doi.org/10.1002/joc.5096.
Li, R. K. K., T. Woollings, C. O’Reilly, and A. A. Scaife, 2020: Tropical atmospheric drivers of wintertime European precipitation events. Quart. J. Roy. Meteor. Soc., 146, 780–794, https://doi.org/10.1002/qj.3708.
Li, X., S.-P. Xie, S. T. Gille, and C. Yoo, 2016: Atlantic-induced pan-tropical climate change over the past three decades. Nat. Climate Change, 6, 275–279, https://doi.org/10.1038/nclimate2840.
Liu, Y., L. Wang, W. Zhou, and W. Chen, 2014: Three Eurasian teleconnection patterns: Spatial structures, temporal variability, and associated winter climate anomalies. Climate Dyn., 42, 2817–2839, https://doi.org/10.1007/s00382-014-2163-z.
Łupikasza, E. B., and K. Cielecka-Nowak, 2020: Changing probabilities of days with snow and rain in the Atlantic sector of the Arctic under the current warming trend. J. Climate, 33, 2509–2532, https://doi.org/10.1175/JCLI-D-19-0384.1.
Maidens, A., J. R. Knight, and A. A. Scaife, 2021: Tropical and stratospheric influences on winter atmospheric circulation patterns in the North Atlantic sector. Environ. Res. Lett., 16, 024035, https://doi.org/10.1088/1748-9326/abd8aa.
Manola, I., R. J. Haarsma, and W. Hazeleger, 2013: Drivers of North Atlantic Oscillation events. Tellus, 65A, 19741, https://doi.org/10.3402/tellusa.v65i0.19741.
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, https://doi.org/10.1038/ngeo2277.
Okumura, Y., S.-P. Xie, A. Numaguti, and Y. Tanimoto, 2001: Tropical Atlantic air-sea interaction and its influence on the NAO. Geophys. Res. Lett., 28, 1507–1510, https://doi.org/10.1029/2000GL012565.
Pang, B., R. Lu, and R. Ren, 2022: Impact of the Scandinavian pattern on long-lived cold surges over the South China Sea. J. Climate, 35, 1773–1785, https://doi.org/10.1175/JCLI-D-21-0607.1.
Sardeshmukh, P. D., and B. J. Hoskins, 1988: The generation of global rotational flow by steady idealized tropical divergence. J. Atmos. Sci., 45, 1228–1251, https://doi.org/10.1175/1520-0469(1988)045<1228:TGOGRF>2.0.CO;2.
Scaife, A. A., and D. Smith, 2018: A signal-to-noise paradox in climate science. npj Climate Atmos. Sci., 1, 28, https://doi.org/10.1038/s41612-018-0038-4.
Scaife, A. A., and Coauthors, 2017: Tropical rainfall, Rossby waves and regional winter climate predictions. Quart. J. Roy. Meteor. Soc., 143 (702), 1–11, https://doi.org/10.1002/qj.2910.
Servain, J., G. Caniaux, Y. K. Kouadio, M. J. McPhaden, and M. Araujo, 2014: Recent climatic trends in the tropical Atlantic. Climate Dyn., 43, 3071–3089, https://doi.org/10.1007/s00382-014-2168-7.
Shi, N., D. Zhang, Y. Wang, and S. Tajie, 2019: Subseasonal influences of teleconnection patterns on the boreal wintertime surface air temperature over southern China as revealed from three reanalysis datasets. Atmosphere, 10, 514, https://doi.org/10.3390/atmos10090514.
Smith, D. M., and Coauthors, 2020: North Atlantic climate far more predictable than models imply. Nature, 583, 796–800, https://doi.org/10.1038/s41586-020-2525-0.
Smith, D. M., and Coauthors, 2022: Robust but weak winter atmospheric circulation response to future Arctic sea ice loss. Nat. Commun., 13, 727, https://doi.org/10.1038/s41467-022-28283-y.
Sohn, S.-J., C.-Y. Tam, and C.-K. Park, 2011: Leading modes of East Asian winter climate variability and their predictability: An assessment of the APCC multi-model ensemble. J. Meteor. Soc. Japan, 89, 455–474, https://doi.org/10.2151/jmsj.2011-504.
Sui, C., L. Yu, and T. Vihma, 2020: Occurrence and drivers of wintertime temperature extremes in northern Europe during 1979–2016. Tellus, 72A, 1–19, https://doi.org/10.1080/16000870.2020.1788368.
Taylor, K. E., 2001: Summarizing multiple aspects of model performance in a single diagram. J. Geophys. Res., 106, 7183–7192, https://doi.org/10.1029/2000JD900719.
Tokinaga, H., and S.-P. Xie, 2011: Weakening of the equatorial Atlantic cold tongue over the past six decades. Nat. Geosci., 4, 222–226, https://doi.org/10.1038/ngeo1078.
Wang, M., and B. Tan, 2020: Two types of the Scandinavian pattern: Their formation mechanisms and climate impacts. J. Climate, 33, 2645–2661, https://doi.org/10.1175/JCLI-D-19-0447.1.
Warner, J. L., J. A. Screen, and A. A. Scaife, 2020: Links between Barents-Kara sea ice and the extratropical atmospheric circulation explained by internal variability and tropical forcing. Geophys. Res. Lett., 47, e2019GL085679, https://doi.org/10.1029/2019GL085679.
Yu, L., C. Sui, D. H. Lenschow, and M. Zhou, 2017: The relationship between wintertime extreme temperature events north of 60°N and large-scale atmospheric circulations. Int. J. Climatol., 37, 597–611, https://doi.org/10.1002/joc.5024.
Yuan, C., and W. Li, 2019: Variations in the frequency of winter extreme cold days in northern China and possible causalities. J. Climate, 32, 8127–8141, https://doi.org/10.1175/JCLI-D-18-0771.1.
Zhang, D., N. Shi, and S. Tajie, 2022: Mechanisms of the subseasonal influences of Scandinavian events on winter surface air temperature over eastern China. Atmos. Res., 268, 105994, https://doi.org/10.1016/j.atmosres.2021.105994.
Zhou, W., J. C. L. Chan, W. Chen, J. Ling, J. G. Pinto, and Y. Shao, 2009: Synoptic-scale controls of persistent low temperature and icy weather over southern China in January 2008. Mon. Wea. Rev., 137, 3978–3991, https://doi.org/10.1175/2009MWR2952.1.