Influence of high-latitude atmospheric circulation changes ...david/Ding_etal_2017.pdf · (ms 1per decade) in ERA-Interim reanalysis. c, Domain-averaged time series for September

ARTICLESPUBLISHED ONLINE: 13 MARCH 2017 | DOI: 10.1038/NCLIMATE3241

Influence of high-latitude atmospheric circulationchanges on summertime Arctic sea iceQinghua Ding1,2,3*, Axel Schweiger3, Michelle L’Heureux4, David S. Battisti5,6, Stephen Po-Chedley5,Nathaniel C. Johnson7, Eduardo Blanchard-Wrigglesworth5, Kirstin Harnos4, Qin Zhang4,Ryan Eastman5 and Eric J. Steig5,6

The Arctic has seen rapid sea-ice decline in the past three decades, whilst warming at about twice the global average rate.Yet the relationship between Arctic warming and sea-ice loss is not well understood. Here, we present evidence that trendsin summertime atmospheric circulation may have contributed as much as 60% to the September sea-ice extent declinesince 1979. A tendency towards a stronger anticyclonic circulation over Greenland and the Arctic Ocean with a barotropicstructure in the troposphere increased the downwelling longwave radiation above the ice by warming and moistening thelower troposphere. Model experiments, with reanalysis data constraining atmospheric circulation, replicate the observedthermodynamic response and indicate that the near-surface changes are dominated by circulation changes rather thanfeedbacks from the changing sea-ice cover. Internal variability dominates the Arctic summer circulation trend and may beresponsible for about 30–50% of the overall decline in September sea ice since 1979.

I t is well recognized that recent Arctic sea-ice decline hasboth natural and anthropogenic drivers1–3, but their relativeimportance is poorly known4–7. This uncertainty arises from

the fact that the contribution of atmospheric internal variabilityin the Arctic climate system is inadequately understood, and itis unclear how well models reproduce these processes8. Reliablydistinguishing the natural and anthropogenic contributions of sea-ice loss requires a comprehensive understanding of the mechanismsthat control the variability of sea ice. Although some progresshas been made in this area8–17, the answer is far from clear, andconflicting hypotheses exist18,19.

Earlier work indicated that the decline of sea ice before the1990s was in part owing to an upward trend in the North AtlanticOscillation (NAO) index20,21. However, since the early 1990s, theapparent link between the NAO and sea ice has largely disappeared,with Arctic sea ice declining further despite the reversal in the NAOtrend22. On the other hand, several studies have argued that therecent NAO trend may still play a key role in the sea-ice retreat23,24.Alternatively, the Arctic Dipole has been suggested as a driver ofsea-ice change in some regions of the Arctic25. Overall, the recenttrends in Arctic sea ice cannot be linked to simple indices ofclimate variability22.

In this paper, we examine the contribution of the atmosphericcirculation to Arctic sea-ice variability by utilizing an atmosphericgeneral circulationmodel (ECHAM5) coupledwith a simple ocean–sea-ice model in which the atmospheric circulation field is nudgedto observations. Specifically, we explore how the high-latitudesummertime atmospheric circulation impacts the September Arcticsea-ice extent, and estimate to what extent changes in atmosphericcirculation explain the observed sea-ice loss of the past few decades.

Observed linkage between circulation and sea iceTo examine the physical linkages, we focus on the connectionbetween September sea-ice extent and the preceding summer(June–July–August, JJA) atmospheric circulation. We choose thispreceding 3-month window because sea-ice extent anomalies havea ∼3-month decorrelation timescale26, and previous studies haveshown a strong link between summer circulation and sea-icevariability23,24,27,28. We focus on physical mechanisms, analysingtemperature, humidity, and downward longwave radiation (DLR),all of which are affected by atmospheric circulation and, in turn,affect sea-ice concentration. A key player is the radiation balance,which dominates the surface energy balance controlling the growthand melt of Arctic sea ice29,30.

The region with the greatest negative trend in Septembersea-ice concentration since 1979 is highlighted in Fig. 1a, andincludes the Beaufort, Chukchi, and East Siberian Seas, featuringan average decline >10%/decade. Concomitant with the trend insea ice are trends in the atmospheric circulation: 200 hPa and700 hPa geopotential heights have been rising over northeasternCanada and Greenland, and the surface winds have become moreanticyclonic (Figs 1b and 2d). To illuminate processes that linkthe circulation changes to sea-ice changes, we first construct anindex of September sea-ice coverage over the region with the fastestsea-ice decline from 1979 to 2014 (Fig. 1c), as well as indices ofJJA low-level temperature (surface to 750 hPa), DLR at surface,and integrated atmospheric water vapour (surface to 750 hPa)derived from ERA-Interim reanalysis (hereafter ERA-I); indices areaveraged over the Arctic, poleward of 70◦ N, and shown in Fig. 1c.The decreasing trend in sea-ice concentration is accompanied byincreasing trends in JJA Arctic temperature, DLR and water vapour.

1Department of Geography, University of California, Santa Barbara, California 93106, USA. 2Earth Research Institute, University of California,Santa Barbara, California 93106, USA. 3Polar Science Center, Applied Physics Laboratory, University of Washington, Seattle, Washington 98195, USA.4NOAA Climate Prediction Center, College Park, Maryland 20740, USA. 5Department of Atmospheric Sciences, University of Washington, Seattle,Washington 98195, USA. 6Department of Earth and Space Sciences, University of Washington, Seattle, Washington 98195, USA. 7Cooperative Institute forClimate Science, Princeton University, Princeton, New Jersey 08540, USA. *e-mail: [email protected]

NATURE CLIMATE CHANGE | ADVANCE ONLINE PUBLICATION | www.nature.com/natureclimatechange 1

© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

http://dx.doi.org/10.1038/nclimate3241

mailto:[email protected]

www.nature.com/natureclimatechange

ARTICLES NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE3241

September sea ice

Sea ice related Z200

a

d

b c

e f g

1 m s−1 20 kg m−1 s−1

0.4 0.5 0.6 0.7

Temp. related Z200 Hum. related Z200 LW related Z200

Sea iceSpe. Hum.

Temp.LW

GL-Z200

−1025

80604020

0−20−40−60−80

1980 1985 1990 1995Year

2000 2005 2010

60

40

20

0

−20

−40

−60

Sea ice

Temp.

Hum. LWZ200

2.01.51.00.50.0

−0.5−1.0−1.5

−2.0

86420−2−4−6−8

−0.4

−0.2

0.0

0.2

0.4

20

15

10

−20

−30

JJA Z200 Time series

2 m s−1 per decade

−10

Figure 1 | Relationship between the September Arctic sea ice and summer large-scale circulation. a, Linear trend (% per decade) of September sea-iceconcentration from the NSIDC passive microwave monthly sea-ice record (1979–2014). b, Linear trend (m per decade) of JJA Z200 and surface wind(m s−1 per decade) in ERA-Interim reanalysis. c, Domain-averaged time series for September sea-ice anomaly (%) averaged over the region circled by theblue contour in a, lower-level (1,000 hPa to 750 hPa) JJA temperature (◦C) and JJA specific humidity anomalies (g kg−1) in the Arctic (averaged over theregion north of 70◦ N), JJA downwelling longwave radiation (LW) anomaly at surface (W m−2) in the Arctic north of 70◦ N, and JJA Z200 anomaly (m)over Greenland (66◦–80◦ N, 310◦–330◦ E, indicated by the dot in b, and referred to as GL-Z200). d–g, Correlation of each time series in c with JJA Z200for the period 1979–2014. In d, regression of JJA surface wind with September sea-ice index is superposed. The sign is reversed in d for simplicity ofcomparison with other plots. In f, regression between the specific humidity index and vertically integrated water vapour flux is plotted. All linear trends areremoved in calculating the correlations in d–g. Black stippling in all plots indicates statistically significant correlation or trend at the 5% level; in d and f,vectors are plotted when regressions are statistically significant at the 5% level.

Regressing the domain-averaged sea-ice anomaly time series againstthe JJA geopotential height at 200 hPa (Z200) in ERA-I, we findthat decreasing sea ice is accompanied by increasing Z200, withmaximum amplitude over Greenland (Supplementary Fig. 1).

These same interrelationships between the trends in the indicesof JJA Z200, temperature, water vapour, DLR and September sea-ice concentration are also apparent using detrended indices: anArctic summer with higher than normal Z200 over Greenland,greater DLR at the surface, and increased low-level water vapourand temperature over the Arctic is followed by negative sea-ice anomalies in September (Fig. 1c); for example, detrendedindices of temperature and specific humidity are correlated atr = 0.89; detrended indices of sea-ice concentration and DLR arecorrelated at r =−0.75. Importantly, the detrended data showsthat the summertime anomalies in the indices of near-surfacetemperature, DLR, and water vapour are simultaneously associatedwith a common pattern of atmospheric circulation variability (seeFig. 1d–g), and that this circulation pattern is very similar to thecirculation pattern that is associated with the sea-ice interannualvariability in September (see Fig. 1d–g). These correlationmaps alsocompare well with the pan-Arctic Z200 and surface wind trends inJJA (Fig. 1b), which feature a strong high over northeastern Canada,Greenland, and the Arctic Ocean.

The statistical links between atmospheric circulation andtemperature, water vapour, and DLR anomalies over the Arcticin JJA and subsequent sea-ice anomalies in September, whichare robust at both interannual and decadal timescales, prompt

the following investigation to identify the underlying physicalmechanisms. In addition, in this statistical analysis, we find avery weak connection between large-scale atmospheric circulationand surface downwelling shortwave flux in the Arctic duringsummertime. However, we cannot rule out the possibility thatshortwave changes in the past 30 years, which are not directly linkedto the circulation change, may also contribute to the abrupt Arcticsea-ice retreat in September29.

The mechanism connecting anomalies in the large-scale upper-level circulation and lower-level thermodynamics is apparentin a zonally averaged vertical cross-section of the atmosphere(Fig. 2). The recent JJA circulation trend in the Arctic waslargest in the upper troposphere (200 hPa), while atmosphericwarming mainly occurred in the lower troposphere below 700 hPa(Fig. 2a). Associated with this trend, significant downwardmotion below 700 hPa and poleward of ∼75◦ N is observed(Fig. 2a and Supplementary Fig. 2). Year-to-year variability insummertime vertical motion in the lower troposphere is also highlycorrelated with similar circulation changes in the upper troposphere(Supplementary Fig. 2). Because circulation and vertical velocityare strongly coupled, adiabatic descent forced by the anticycloniccirculation aloft must be a contributor to the observed low-levelwarming. Increased DLR is a direct result of the warming of thelower atmosphere, as well as an increase inwater vapour in the lower(warm) atmosphere. DLR is also strongly affected by cloud cover.To examine the relationship between cloudiness and the circulation,we analyse the relationship between the detrended Z200 heights and

2


NATURE CLIMATE CHANGE | ADVANCE ONLINE PUBLICATION | www.nature.com/natureclimatechange



NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE3241 ARTICLES100

Obs.

Obs.6

6

6 12 12

6 6

66

6 6

6 66

6

12

Exp1 Exp2

Exp2 (sea ice)

a b c

e

h

fd

g Exp2 (time series)

Temp.

Spe. Hum.LW

Exp1 Exp2

200300400500600700800900

1,000

100200

8

8

6

1012

300400500600700800900

1,000

100200 8

3008

6

10

12400500600700800900

1,00060 70

Latitude (° N)80 90

6

9

10

14

1612

0.8

0.6

0.4

0.2

−0.2

−0.4

−0.6

−10

4

3

2

1

0

−1

−2

−3

−4

0.5 16

12

8

4

0

−4

−8

−12

−16

0.4

0.3

0.2

0.1

0.0

−0.1

−0.2

−0.3

−0.4

−0.5

Temp.

Hum.LW

−20

−30

1980 1985 1990 1995

Year

2000 2005 2010

1.2

0.9

0.6

0.3

−0.3

−0.6

−0.9

0.8

0.6

0.4

0.2

−0.2

−0.4

−0.660 70

Latitude (° N)80 90 60 70

Latitude (° N)80 90

Figure 2 | Simulated impact of atmospheric circulation on Arctic thermodynamic trends. a, Meridional cross-section of the linear trend of zonalmean JJA temperature (shading, ◦C per decade), geopotential height (black contour, m per decade) and vertical velocity (red/blue contour, interval:6× 10−5 Pa s−1 decade−1) in ERA-I (1979–2014). b,c, Same as a but for Exp-1 and Exp-2 simulations, respectively. d, Linear trend of lower tropospheric(1,000 hPa to 750 hPa) JJA temperature (shading, ◦C per decade) and geopotential height at 700 hPa (red contour, m/decade) in ERA-I (1979–2014).e,f, Same as d but for Exp-1 and Exp-2 simulations, respectively. g, Linear trend of September sea-ice concentration (% per decade) simulated in Exp-2.h, Domain-averaged time series for lower tropospheric (1,000 hPa to 750 hPa) JJA temperature (◦C) and specific humidity anomalies (g kg−1), and JJAdownwelling longwave radiation (LW) anomaly (W m−2) at surface in the Arctic (north of 70◦ N) simulated in Exp-2. In d to g stippling indicatesstatistically significant trends at the 5% level.

cloud anomalies across the Arctic. An anomalous high pressure overthe Arctic Ocean favours a decrease in cloudiness in the upper andmiddle levels of the atmosphere, possibly associated with decreasedstorm activity. However, low clouds increase slightly over Greenlandand the central Arctic (Supplementary Fig. 3). Because low cloudsdominate the Arctic, this circulation-driven shift towards lowerclouds is qualitatively consistent with the observed increase in DLR.

Atmospheric nudging experimentsWe have provided a plausible mechanism for how circulationchanges can impact Arctic sea ice, but it is difficult to determinecausality with observational evidence alone because of the feedbacksbetween sea ice and the atmosphere. For example, the removalof sea ice can increase surface air temperature, clouds, low-levelwater vapour and hence DLR31, and may therefore affect the

atmospheric circulation itself32,33. To better understand the directionof causality, we conduct two model experiments to examine theinfluence of the observed high-latitude circulation trends on seaice and other key variables. The first experiment (Exp-1) is a 36-yrhistorical run with the ECHAM5 atmospheric model, in whichthe global atmospheric circulation (vorticity and divergence fromthe surface to the top of the atmosphere) is weakly nudged to thecorresponding daily reanalysis data (see Methods). For the lowerboundary, climatological sea surface temperature (SST), and seaice are imposed everywhere. In the second experiment (Exp-2), allsettings are the same as those in Exp-1, except that the atmospheremodel is coupled to a slab ocean–sea-ice model in the Arctic (northof 60◦ N), permitting a feedback from the ice–ocean system to theprescribed atmosphere. Anthropogenic forcings are held constantin both experiments. The difference between the two experiments



3




Full forcing

Full forcing Modified forcing

Full forcing

PIOMAS

Obs

Modified forcing

Modified forcing

−10

3

2

1

0

−1

−2

−3

6

8

4

2

0

−2

−4

1980 1985 1990 1995 2000Year

2005 2010

1980

Unit: 106 km2

Unit: 1,000 km3

1985 1990 1995 2000Year

2005 2010

−6

−8

−20

−30

0.1

−0.1

−0.3

−0.5

Full forcing

a

d

b

e

c

fModified forcing

September sea ice extent

September sea ice volume

Figure 3 | Simulated impact of atmospheric circulation on summertime Arctic sea-ice trends. a,b,d,e, Linear trend of September sea-ice concentration(a,b, % per decade) and thickness (d,e, m per decade) in Exp-5 (denoted as ‘full forcing’) and Exp-6 (denoted as ‘modified forcing’). c, Anomalous totalarea of September sea-ice extent (area of ocean with ice concentration of at least 15%) in both simulations and NSIDC observations. f, Anomalous totalvolume of September sea ice (area of ocean with ice concentration of at least 15%) in both simulations and the Pan-Arctic Ice Ocean Modeling andAssimilation System (PIOMAS; ref. 46).

isolates the importance of tropospheric forcing (dynamics) fromthe amplifying effects of sea-ice–ocean feedbacks. If the mechanismresponsible for the strong correlations between the thermodynamicvariables (temperature, humidity and DLR) and Z200 circulation isprimarily due to a dynamical control of the circulation on radiation(via the temperature and moisture fields) rather than feedbacks,then the results from Exp-2 should be similar to those in Exp-1.

Exp-1 and Exp-2 reproduce the spatial distribution in theobserved trends in geopotential height, temperature and verticalvelocity, both in height and latitude (see Fig. 2b,cwith Fig. 2a).Whilethe simulated spatial patterns are similar to the reanalysis data, theheight increases are slightly smaller and temperature increases largerthan observed. The simulated lower tropospheric temperature inthe experiments is significantly correlated with the ERA-I reanalysisdata on both interannual and interdecadal timescales. The regionwith the greatest lower tropospheric warming over the Ocean in JJAis over the Beaufort, Chukchi, and East Siberian seas (Fig. 2d). Theresults from Exp-2 are very similar to those in Exp-1 and the ERA-Ireanalysis, indicating that the lower-level atmospheric warming islargely due to the changes in atmospheric circulation, while theinfluence of sea ice on the atmosphere is small (Fig. 2e,f). This isnot entirely surprising: in JJA, sea-ice extent is at a minimum, theice surface is covered with melt water ponds, and the temperaturecontrast between the sea ice and the atmosphere is small. Therefore,changes in sea-ice conditions in JJA cannot have a significantimpact on surface temperature, atmospheric humidity, and DLR.The sea-ice trends simulated in Exp-2 are similar in magnitude but‘upstream’ of those observed (see Fig. 2g with Fig. 1c), probably dueto a lack of advection of ice by the winds which are not modelled inthe slab ocean, thermodynamic-only sea-ice model used in Exp-1and Exp-2 (see Methods).

Is it possible that the observed circulation itself, especially inthe lower levels, is a response to the sea-ice change? If so, theobserved boundary layer warming would be expected as the resultof the thermal wind relationship when the circulation is nudgedto the observed winds as done in both Exp-1 and Exp-2. To testthis idea, in Exp-3, the model’s global winds are nudged to theobservations only above 700 hPa and a more active interactionbetween sea ice and atmosphere is allowed in the boundary layerof the Arctic. In this new experiment, we still simulate a similarbut slightly weaker low-level temperature and sea-ice response asin Exp-2 (Supplementary Fig. 4). In addition, model experimentsthat examine the impact of changes in sea-ice conditions in isolationshow no significant impact on the circulation during the summermonths32,34. Because the impact of sea-ice changes on the circulationmay be model dependent, we examine this impact in an additionalexperiment (Exp-4; see Methods), in which we specify the observedtime evolution (1979–2014) of Arctic SST and sea ice (north of60◦ N), and impose climatological SST elsewhere. In JJA, the modelresponse to the SST warming trend and sea-ice retreat in the Arcticis very small (Supplementary Fig. 5), in contrast to the responses inthe other three seasons. This is because, in summer, the temperaturecontrast between the open ocean and sea ice is relatively small(compared with fall and winter), and the impact of sea-ice losson the atmosphere is muted. Even in fall and winter, when themodel exhibits the largest response to sea-ice melt, the simulatedcirculation trend is only about 3m/decade in the upper troposphere,which is only 10 to 20% of the observed change in JJA. Thus, weconclude that the JJA vertical temperature trend in the Arctic in thepast 36 yr is mainly controlled by the changes in the mid- and uppertroposphere (top-down effect), rather than a bottom-up responsedue to the sea-ice melt.

4





NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE3241 ARTICLES

JJA

Z20

0 tr

end

JJA

Z70

0 tr

end

Obs. (1979−2014) CMIP5 (1979−2014) LENS (1979−2014)

25

20

15

10

5

−5

3

6

9

12

15

−3

a b c

Obs. (1979−2014)



CMIP5 (1979−2014) LENS (1979−2014)d e f

Figure 4 | Observed and estimated radiatively forced trends in upper and lower tropospheric geopotential height and winds. a–f, Linear trends of JJAgeopotential height (m per decade) and zonal and meridional winds at 200 hPa (a–c) and 700 hPa (d–f) for the period 1979–2014 from a and d ERA-I (a,d)(repeated from Fig. 1b) the 26-model ensemble mean from the CMIP5 project (b,e) and the 30-member ensemble mean from the LENS project (c,f).

Simulated impact of atmospheric circulation on sea iceHaving demonstrated the impact of the circulation on near-surfacetemperature, water vapour and DLR, we estimate its contribution tothe overall downward trend in sea ice. As noted above, the patternof sea-ice decline simulated in Exp-2 is shifted upstream of theobserved sea-ice decline (see Figs 1a and 2g). Without sea-ice drift,the strongest reduction in sea ice in Exp-2 occurs over the CanadianBasin and north of Greenland, where the circulation exerts thelargest forcing on lower-level temperature and surface DLR overthe past 36 yr (Fig. 2h). To extend our results to include the effectsof ice advection and thus simulate a sea-ice response that can bemore directly compared to observations, we force a global coupledocean/sea-ice (POP2–CICE4) model with observed atmosphericfields (ERA-I reanalysis; see Methods) between 1979–2014. Twoexperiments are performed: Exp-5, which uses the complete set ofERA-I forcings, and Exp-6, which removes from the forcing theeffect of the circulation variability that is linearly related to Z200over Greenland (GL-Z200), identified above (see Methods).

The spatial patterns and time series of sea ice in the controlrun (Exp-5) exhibit a close correspondence to the observed sea-ice reduction (see Figs 1a and 3a) further demonstrating the skillof the POP2–CICE4 model to produce realistic sea-ice variabilitywhen driven with observed forcings. The pattern of the JJA trendin sea-ice melt in Exp-5 is similar to the extent change of sea icein Exp-2 (Fig. 2g). Also, an index of the domain-averaged JJA sea-ice melt in Exp-5 (north of 70◦ N) shows a very high correlationwith Z200 over Greenland and the Arctic on both interannual andlonger timescales (Supplementary Fig. 6). This indicates a similarimpact of the atmospheric circulation on thermodynamic anomaliesin the ECHAM5 and POP2–CICE4 models, and supports the ideathat ice advection transports thermodynamically generated sea-ice

anomalies downstream from the Beaufort and Chukchi seas to theEast Siberian Sea.

A comparison of sea-ice concentration and sea-ice volumepatterns and time series between Exps-5 and 6 supports ourhypothesis that trends in the atmospheric circulation havecontributed substantially to the 1979–2014 reduction of the sea-ice extent in September. In the absence of GL-Z200 circulationanomalies (Exp-6), the decline in sea-ice concentration andthickness are only 41% and 40% of the decline in the control run(Exp-5), respectively, suggesting that the summertime circulationcontributes to as much as 60% of the sea-ice loss since 1979. Themagnitude of this estimate is perhaps not surprising given priorestimates linking roughly 60% of September sea-ice extent varianceto sea-level pressure variability in the prior month23.

Estimation of internal and anthropogenic contributionsHaving demonstrated the effect of changes in the atmosphericcirculation on sea-ice trends, we now examine whether thecirculation trends are due to internal variability or anthropogenicforcing. First, to determine the role of anthropogenic warming onthe high-latitude circulation, we examine the ensemble averagetrends in JJA geopotential height and winds at 200 hPa and 700 hPaover the Arctic region from the historical forcing simulations ofthe Coupled Model Intercomparison Project Phase 5 (CMIP5;ref. 35) and the Large Ensemble Project (LENS; ref. 36). The averageover the 26-member (Supplementary Table 1) ensemble in CMIP5and the 30-member ensemble in LENS display positive trends innear-surface temperature and in Z200 and Z700 over our studyperiod (1979–2014) (Fig. 4 and Supplementary Fig. 7). Unlikeobservations, which show barotropic, anticyclonic circulationtrends over Greenland from the surface to the upper troposphere,



5




the circulation trends in CMIP5 and LENS ensembles are weakand unorganized in the Arctic, and particularly in the lowertroposphere (see Fig. 4 or Fig. 1b). An amorphous warming withlittle circulation change is the typical thermodynamic responsein the mid- and high latitudes to increasing CO2 (see alsoSupplementary Fig. 7). The strong gradients in the observedtrend pattern (Fig. 2a) support the idea that internal atmosphericvariability is an important contributor to the recent multidecadaltrends in the high-latitude circulation.

Second, to estimate the anthropogenic contribution to theobserved warming and sea-ice reduction in the Arctic, twoadditional experiments are conducted. Exp-7 and 8 are equivalentto Exp-2 but we remove the effects of global warming on thehigh-latitude winds, which are used to constrain the model inExp-2 (Supplementary Fig. 8). These results show the same stronggeopotential height increases as in Exp-2, with approximately 70%of Arctic low-level warming and sea-ice extent change (northof 70◦ N) relative to Exp-2. Hence, these experiments suggestthat ∼30% of the anomalous thermodynamic sea-ice extentreduction is attributable to anthropogenic influences on the Arcticcirculation. Applying this estimate to the overall circulation-drivensea-ice trend established in Exp-6 (60%), we estimate that about∼42% (70%×60%) of the sea-ice decline observed since 1979 inSeptember is due to internal variability.

What are the sources of internal variability on the Arcticcirculation? The upper tropospheric circulation in the Arctic hasa strong link to variability originating from the tropics27,37–40.Previous research39 identified a relationship between tropical SSTvariability and annual mean atmospheric circulation over the Arcticwith a centre of action over Greenland, and model experiments39show that about 50% of the circulation change and the associatedwarming over Greenland are attributable to internal variabilityoriginating from the tropical Pacific Ocean. Therefore, a substantialcontribution of tropical Pacific variability on sea ice loss via thisteleconnection is to be expected. A further examination of thisquestion will require a modelling framework that reproduces thetropics–high latitude linkage faithfully and efficiently.

Uncertainties in reanalysis dataWe have used ERA-I as a proxy for observational data. To assessthe robustness of this assumption we examined a suite of currentlyavailable reanalyses and station radiosonde data in Greenland. Wefind that all exhibit a very similar JJA circulation trend in thepast 36 yr with a clear anticyclonic maximum over Greenland andnortheastern Canada. The strong temporal coherence in GL-Z200among all reanalysis data sets (Supplementary Fig. 9) points totheir reliability to capture the circulation variability in and aroundGreenland. All data exhibit a similarly strong trend in circulationsince 1979 and coherent variations on interannual timescales.The magnitude of trends over Greenland vary, ranging from15m/decade to 25m /decade with ERA-I (19.3m/decade) in themiddle of this range and closest inmatching the trend in radiosondedata (18.8m/decade). Thus, the circulation trend centred overGreenland seen in all reanalysis produces is a robust feature of thedata, and unlikely the result of a changing observation networkthat is common to all reanalyses. The impact of observationaluncertainties on our conclusions can be estimated by linearlyscaling the results from Exp-6 by the GL-Z00 trends extractedfrom different reanalyses products. The atmospheric circulationcontribution to sea-ice loss ranges from 48% (NCEP-2) to 75%(MERRA-2). Using the 70% contribution of internal forcing to thecirculation variability established above, we can attribute 30–50% ofsea-ice loss to internal variability.

Reanalysis products are reliable representations of the observedcirculation, humidity and temperature. The reanalysis data probablyalso reliably represent the variability in total cloud cover in the

satellite era, particularly in summer41. However, there is evidencethat the relationship between sea ice and the vertical distributionof cloud fraction in reanalysis does not agree well with that inobservations31. Therefore, to the extent that changes in cloudinesscontribute to the DLR trends associated with the trends incirculation, the connection between large-scale circulation andcloud variability is subject to considerable uncertainty and needs tobe explored in future studies.

ConclusionOur conclusions are based on experiments involving several modelsrather than one integrated model. POP2–CICE4 simulations areused to allow for a realistic simulation of sea-ice variability, whileECHAM5 simulations allow for realistic simulations of the linkagebetween circulation and thermodynamic drivers of sea-ice loss,while permitting nudging to observed wind fields. The closeagreement of patterns of thermodynamic forcing on sea ice betweenthe ECHAM5 and POP2–CICE4 provides confidence in combiningthese two tools.

How sea-ice variability and trends can impact the Arcticatmospheric circulation is an area of vigorous research42. Studiessuggest numerous mechanisms in which sea-ice loss modulatesthe large-scale circulation in the lower troposphere in winter43–45.This paper, instead, puts the spotlight on how the high-latitudecirculation impacts sea ice. Although positive feedbacks betweensea ice and the Arctic circulation exist, we find that these are smallduring summer. Instead, circulation variations over the Arctic havebeen a significant factor in driving sea-ice variability since 1979, andhave had a non-trivial contribution to the total surface temperaturetrend over Greenland and northeastern Canada39. The potentiallylarge contribution of internal variability to sea-ice loss over thenext 40 years27 reinforces the importance of natural contributionsto sea-ice trends over the past several decades. The similarity ofhigh-latitude circulation variability associated with sea-ice loss tothe teleconnections with the tropical Pacific suggests a contributionof sea-ice losses from SST trends across the tropical Pacific Ocean.Decadal trends in the hemispheric circulation are an importantdriver of Arctic climate change, and therefore a significant source ofuncertainty in projections of sea ice. Better understanding of theseteleconnections and their representation in global models underincreasing greenhouse gases may help increase predictability onseasonal to decadal timescales.

MethodsMethods, including statements of data availability and anyassociated accession codes and references, are available in theonline version of this paper.

Received 26 July 2016; accepted 3 February 2017;published online 13 March 2017

References1. Min, S. K., Zhang, X. B., Zwiers, F. W. & Agnew, T. Human influence on

Arctic sea ice detectable from early 1990s onwards. Geophys. Res. Lett. 35,L21701 (2008).

2. Kay, J. E., Holland, M. M. & Jahn, A. Inter-annual to multi-decadal Arctic seaice extent trends in a warming world. Geophys. Res. Lett. 38, L15708 (2011).

3. Notz, D. & Marotzke, J. Observations reveal external driver for Arctic sea-iceretreat. Geophys. Res. Lett. 39, L08502 (2012).

4. Stroeve, J., Holland, M. M., Meier, W., Scambos, T. & Serreze, M. Arctic sea icedecline: faster than forecast. Geophys. Res. Lett. 34, L09501 (2007).

5. Day, J. J., Hargreaves, J. C., Annan, J. D. & Abe-Ouchi, A. Sources of multi-decadal variability in Arctic sea ice extent. Environ. Res. Lett. 7, 034011 (2012).

6. Rampal, P., Weiss, J., Dubois, C. & Campin, J.-M. IPCC climate models do notcapture Arctic sea ice drift acceleration: consequences in terms of projected seaice thinning and decline. J. Geophys. Res. 116, C00D07 (2011).

7. Swart, N. C., Fyfe, J. C., Hawkins, E., Kay, J. E. & Jahn, A. Influence of internalvariability on Arctic sea-ice trends. Nat. Clim. Change 5, 86–89 (2015).

6






NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE3241 ARTICLES8. Zhang, R. Mechanisms for low frequency variability of summer Arctic sea ice

extent. Proc. Natl Acad. Sci. USA 112, 1422296112 (2015).9. Francis, J. A. & Hunter, E. New insight into the disappearing Arctic sea ice.

Eos 87, 509–511 (2006).10. Graversen, R. G. &Wang, M. Polar amplification in a coupled climate model

with locked albedo. Clim. Dynam. 33, 629–643 (2009).11. Chylek, P., Folland, C. K., Lesins, G., Dubey, M. K. &Wang, M.-Y. Arctic air

temperature change amplification and the Atlantic multidecadal oscillation.Geophys. Res. Lett. 36, L14801 (2009).

12. Screen, J. A. & Simmonds, I. The central role of diminishing sea ice in recentArctic temperature amplification. Nature 464, 1334–1337 (2010).

13. Serreze, M. C. & Barry, R. G. Processes and impacts of Arctic amplification:a research synthesis. Glob. Planet. Change 77, 85–96 (2011).

14. Bintanja, R., Graversen, R. & Hazeleger, W. Arctic winter warming amplified bythe thermal inversion and consequent low infrared cooling to space.Nat. Geosci. 4, 758–761 (2011).

15. Vaughan, D. G. et al . in Climate Change 2013: The Physical Science Basis(eds Stocker, T. F. et al .) 317–382 (IPCC, Cambridge Univ. Press, 2013).

16. Park, H.-S., Lee, S., Kosaka, Y., Son, S.-W. & Kim, S.-W. The impact ofArctic winter infrared radiation on early summer sea ice. J. Clim. 28,6281–6296 (2015).

17. Park, D.-S., Lee, S. & Feldstein, S. B. Attribution of the recent wintersea-ice decline over the Atlantic sector of the Arctic Ocean. J. Clim. 28,4027–4033 (2015).

18. Screen, J. A., Deser, C. & Simmonds, I. Local and remote controls on observedArctic warming. Geophys. Res. Lett. 39, L10709 (2012).

19. Perlwitz, J., Hoerling, M. & Dole, R. Arctic tropospheric warming: causes andlinkages to lower latitudes. J. Clim. 28, 2154–2167 (2015).

20. Deser, C., Walsh, J. E. & Timlin, M. S. Arctic sea ice variability in the context ofrecent atmospheric circulation trends. J. Clim. 13, 617–633 (2000).

21. Hu, A., Rooth, C., Bleck, R. & Deser, C. NAO influence on sea ice extent in theEurasian coastal region. Geophys. Res. Lett. 29, 2053 (2002).

22. Deser, C. & Teng, H. Evolution of Arctic sea ice concentration trends and therole of atmospheric circulation forcing 1979–2007. Geophys. Res. Lett. 35,L02504 (2008).

23. Ogi, M., Rigor, I. G., McPhee, M. G. &Wallace, J. M. Summer retreat of Arcticsea ice: role of summer winds. Geophys. Res. Lett. 35, L24701 (2008).

24. Ogi, M., Yamazaki, K. & Wallace, J. M. Influence of winter and summersurface wind anomalies on summer Arctic sea ice extent. Geophys. Res. Lett. 37,L07701 (2010).

25. Watanabe, E., Wang, J., Sumi, A. & Hasumi, H. Arctic dipole anomaly and itscontribution to sea ice export from the Arctic Ocean in the 20th century.Geophys. Res. Lett. 33, L23703 (2006).

26. Blanchard-Wrigglesworth, E., Armour, K. C., Bitz, C. M. & DeWeaver, E.Persistence and inherent predictability of Arctic sea ice in a GCM ensembleand observations. J. Clim. 24, 231–250 (2011).

27. Wettstein, J. J. & Deser, C. Internal variability in projections of twenty-firstcentury Arctic sea ice loss: role of the large-scale atmospheric circulation.J. Clim. 27, 527–550 (2014).

28. Screen, J. A., Simmonds, I. & Keay, K. Dramatic interannual changes ofperennial Arctic sea ice linked to abnormal summer storm activity. J. Geophys.Res. 116, D15105 (2011).

29. Kay, J. E., L’Ecuyer, T., Gettelman, A., Stephens, G. & O’Dell, C. Thecontribution of cloud and radiation anomalies to the 2007 Arctic sea ice extentminimum. Geophys. Res. Lett. 35, L08503 (2008).

30. Francis, J. A. & Hunter, E. Drivers of declining sea ice in the Arctic winter:a tale of two seas. Geophys. Res. Lett. 34, L17503 (2007).

31. Schweiger, A. J., Lindsay, R. W., Vavrus, S. & Francis, J. A. Relationshipsbetween Arctic clouds and sea ice during autumn. J. Clim. 21,4799–4810 (2008).

32. Bhatt, U. S. et al . The Atmospheric Response to Realistic Reduced Summer ArcticSea Ice Anomalies, in Arctic Sea Ice Decline: Observations, Projections,Mechanisms, and Implications (American Geophysical Union, 2008).

33. Liu, J. P. et al . Has Arctic sea ice loss contributed to increased surface melting ofthe Greenland ice sheet? J. Clim. 29, 3373–3386 (2016).

34. Deser, C., Tomas, R., Alexander, M. & Lawrence, D. The seasonal atmosphericresponse to projected Arctic sea ice loss in the late 21st century. J. Clim. 23,333–351 (2010).

35. Taylor, K. E., Stouffer, R. J. & Meehl, G. A. An overview of CMIP5 and theexperiment design. Bull. Am. Meteorol. Soc. 93, 485–498 (2012).

36. Kay, J. E. et al . The Community Earth System Model (CESM) largeensemble project: a community resource for studying climate change in thepresence of internal climate variability. Bull. Am. Meteorol. Soc. 96,1333–1349 (2015).

37. Hoerling, M. P., Hurrell, J. W. & Xu, T. Y. Tropical origins for recent NorthAtlantic climate change. Science 292, 90–92 (2001).

38. Lee, S. Testing of the tropically excited Arctic warming (TEAM) mechanismwith traditional El Nino and La Nina. J. Clim. 25, 4015–4022 (2012).

39. Ding, Q. H. et al . Tropical forcing of the recent rapid Arctic warming innortheastern Canada and Greenland. Nature 509, 209–212 (2014).

40. Trenberth, K. E., Fasullo, J. T., Branstator, G. & Phillips, A. S. Seasonalaspects of the recent pause in surface warming. Nat. Clim. Change 4,911–916 (2014).

41. Liu, Y. & Key, J. Assessment of Arctic cloud cover anomalies in atmosphericreanalysis products using satellite data. J. Clim. 29, 6065–6083 (2016).

42. Barnes, E. A. & Screen, J. The impact of Arctic warming on the midlatitudejetstream: Can it? Has it? Will it?WIREs Clim. Change 6, 277–286 (2015).

43. Overland, J. E. & Wang, M. Increased variability in the early winter subarcticNorth American atmospheric circulation. J. Clim. 28, 7297–7305 (2015).

44. Francis, J. A. & Skific, N. Evidence linking rapid Arctic warming tomid-latitude weather patterns. Phil. Trans. R. Soc. A 373, 20140170 (2015).

45. Cohen, J. An observational analysis: tropical relative to Arctic influence onmidlatitude weather in the era of Arctic amplification. Geophys. Res. Lett. 43,5287–5294 (2016).

46. Zhang, J. & Rothrock, D. A. Modeling global sea ice with a thickness andenthalpy distribution model in generalized curvilinear coordinates.Mon. Weath. Rev. 131, 845–861 (2003).

AcknowledgementsThis study was supported by NOAA’s Climate Program Office, Climate Variability andPredictability Program (NA15OAR4310162). We thank the Max Planck Institute forMeteorology and National Center for Atmospheric Research model developers formaking the ECHAM5 and CESM available and M. Steele, J. M. Wallace, C. Bitz, Q. Fu,M. Wang, D. L. Hartmann and D. Frierson for discussions. We acknowledge the CESMLarge Ensemble Community Project and supercomputing resources provided byNSF/CISL/Yellowstone. Q.D. acknowledges support from the University of Washington’sPolar Science Center, the UW-Future of Ice Initiative, the Tamaki Foundation and UCSBCenter for Scientific Computing at CNSI. A.S. is grateful for funding from the NationalScience Foundation through grant ARC-1203425. D.S.B. acknowledges support from theTamaki Foundation. R.E. acknowledges support from NASA NNXBAQ35G.

Author contributionsQ.D. led this work with contributions from all authors. Q.D. made the calculations,implemented the general circulation model experiments, created the figures, and ledwriting of the paper. All authors contributed to the experimental design, interpretingresults and writing the paper.

Additional informationSupplementary information is available in the online version of the paper. Reprints andpermissions information is available online at www.nature.com/reprints.Correspondence and requests for materials should be addressed to Q.D.

Competing financial interestsThe authors declare no competing financial interests.



7



http://www.nature.com/reprints



MethodsData.We use the monthly circulation, temperature, cloud and radiation data fromthe ERA-Interim reanalysis47 during the period 1979–2015. We analyse theobserved monthly sea-ice concentration data derived from Goddard edited passivemicrowave retrievals, which are compiled by the National Snow and Ice DataCenter. Monthly geopotential height records at 200 hPa from six stations along thecoast of Greenland in the Integrated Global Radiosonde Archive (IGRA2) were alsoused, except for 1989 and 1994. All stations have sufficient daily observations (atleast 20 days) to construct reliable monthly averages during the period 1979–2014.We also analyse the output from the thirty realizations, each with slightly perturbedinitial conditions, of historical and future climate projection simulations from theCESM Large Ensemble Community Project (LENS; ref. 36), and the output from26 climate models (Supplementary Table 1) archived in the CMIP5 (ref. 35)database to estimate the climate response to external forcing in the historicalrecord. In both cases, we focus our analysis on the historical satellite era from1979–2014. In both cases, each simulation in the ensemble was forced withidentical historical forcing data (greenhouse gases, solar, volcanic and land use)over 1920–2005 and the 21st century representative concentration pathway 8.5(RCP8.5) scenario through 2100; the experiments that comprise the ensemblesdiffer only in their atmospheric initial conditions.

Models. The general circulation model used to perform the experiments (Exps-1 to4, 7 and 8) in this study is the ECHAM5 atmospheric general circulation model48,with a horizontal resolution of T42 (∼2.8◦ latitude×2.8◦ longitude) and19 vertical levels. In Exps-2, 3, 7 and 8, we coupled the ECHAM5 to a slab ocean inthe high latitudes to assess the role of prescribed circulation in driving the SST andsea ice in the Arctic. The slab ocean is represented as boxes of water of uniformspecified depth (50m). A simple, thermodynamic-only sea-ice model is includedwhen the ocean temperature reaches the freezing point. The ocean temperature orsea-ice condition at each grid point is affected only by heat exchange across theair–sea interface; there is no direct communication between adjacent ocean gridpoints, nor is there any representation of the deep ocean. In addition, acyclo-stationary climatological heat flux is added to the ocean temperaturetendency equation to maintain a seasonal cycle of ocean temperature and sea-iceconditions that is close to the observed climatology during 1979–1995.

In Exps-5 and 6, we use the coupled sea-ice–ocean component (POP2+CICE4)that is from the latest version of CESM (ref. 49). This is a model that includes boththe thermodynamics and dynamics of the ocean and sea ice; the model is run usingtripole 1-degree ocean/sea-ice grids. Specific details on each model experiment arelisted in a later section.

Significance of correlation. The statistical significance of the correlationcoefficient accounts for the autocorrelation in the time series by using an ‘effectivesample size’ N ∗:

N ∗=N1− r1r21+ r1r2

where N is the number of available time steps and r1 and r2 are lag-oneautocorrelation coefficients of each variable50.

Experimental design. Exp-1: ECHAM5’s circulation nudged to observed state.Because we are interested in the impact of the long-term trend of the observedcirculation in our model simulations, and to facilitate our computation efforts, weinterpolate observed monthly ERA-I data to daily fields for nudging. We use a veryweak damping term to nudge the three-dimensional (from surface to the top ofatmosphere) divergence and vorticity fields of the model to the observed monthly(smoothly interpolated to daily) fields in the past 36 yr; this weak damping allowsthe model to generate its own day-to-day variability but constrains the model to bevery close to the observed circulation on monthly and longer timescales. In thelower boundary, we impose climatological SST/sea-ice everywhere. Anthropogenicforcings are held constant in this experiment.Exp-2: Same as Exp-1, except that a simple slab ocean/sea-ice model is used in theArctic, north of 60◦ N.Exp-3: Same as Exp-2, except that we nudge the model’s wind fields above 700 hPato the observed winds.Exp-4: The ECHAM5 atmosphere model is forced by observed history of sea-iceconcentration and SST in the Arctic (north of 60◦ N) and the climatologicalSST/sea-ice everywhere else during the period 1979–2014. Twenty realizations are

run, each with different atmospheric initial conditions; the 20-member ensemblemean represents the model response to the prescribed history of sea ice and SST.Anthropogenic forcings are held constant in this experiment.Exp-5: POP2+CICE4 control run. POP2+CICE4 is used to simulate the sea-icevariability over the past 36 yr using prescribed ERA-Interim atmospheric forcing.The forcing fields include daily data of 10m wind, downwelling shortwave andlongwave radiation, 2m temperature and humidity, sea-level pressureand precipitation.Exp-6 : Same as Exp-5 except that the atmospheric forcing is modified to excise theforcing associated with the trends in the Greenland circulation pattern.

To remove the circulations trend from the observations, we first construct thethirty-six-year seasonal (JJA) averaged time series of the Z200 index overGreenland, Z200GL (GL-Z200 in Fig. 1c). We then linearly regress a key variable Bagainst this time series to obtain spatial pattern β(x,y) of the variable associatedwith the Greenland circulation index. Specifically, for the variable B we have

B(x ,y , t

)=β

(x ,y

)×Z200GL(t) (1)

where B represents a forcing field (for example, 10m zonal wind, DLR, temperature,and so on), x and y indicate the location, t indicates time (JJA), Z200GL is theGreenland Z200 index (GL-Z200 in Fig. 1c), and β is the regression coefficient.

In the second step, the seasonal mean anomalous value of each forcing field issubtracted from the observed daily (or 6-hourly) forcing data during thesummer—rendering a modified forcing that does not include variability or trendsin variables that are associated with Z200GL. In the nine non-summer months, theforcing is exactly the same as that used in the Exp-5 control experiment. Given astrong correlation between circulation and surface winds, temperature, specifichumidity, sea-level pressure, and downwelling longwave radiation in the Arctic,variability and trends in these six variables that are associated with Z200GL areprocessed and removed from the forcing. The initial states of ocean, sea ice andatmosphere in Exp-5 and Exp-6 are exactly the same.Exp-7: Same as Exp-2, except that ECHAM5 is nudged to a modified observedwind patterns in which the long-term trends of simulated winds (zonal andmeridional winds) in the ensemble mean (26 members) of CMIP5 during 1979 to2014 are removed from observation.Exp-8: Same as Exp-2, except that the wind patterns used to nudge ECHAM5 arethe residual wind fields resulting from the removal of the long-term wind trendsderived from LENS (30 members) from ERA-I winds.

Data availability. All reanalysis data used in this study were obtained from publiclyavailable sources: ERA-I reanalysis data can be obtained from the ECMWF publicdata sets web interface, http://apps.ecmwf.int/datasets; CFSR can be obtained fromNCDC/NOMADS data access, https://nomads.ncdc.noaa.gov/#cfsr; NCEP-2reanalysis can be obtained from NOAA/ESRL/PSD data portal,https://www.esrl.noaa.gov/psd/data. MERRA-1 and MERRA-2 data are available atthe Modelling and Assimilation Data and Information Services Center,https://gmao.gsfc.nasa.gov/reanalysis/MERRA. JRA55 can be downloaded from theJapan Meteorological Agency (JMA) data dissemination system,http://jra.kishou.go.jp/JRA-55/index_en.html. IGRA2 radiosonde data can beaccessed via the NOAA/NCDC data portal,https://www.ncdc.noaa.gov/data-access. Simulated global circulation andtemperature under anthropogenic forcing were obtained from the CMIP5 andLENS archives accessed through the Earth System Grid Federation data portal,http://esgf.llnl.gov. Output data from the model simulations (Exps 1–8) areavailable from the corresponding author upon request.

References47. Dee, D. P. et al . The ERA-Interim reanalysis: configuration and

performance of the data assimilation system. Q. J. R. Meteorol. Soc. 137,553–597 (2011).

48. Roeckner, E. et al .Max Planck Institut für Meteorologie Report(Max-Planck-Institut für Meteorologie, 2003).

49. Hurrell, J. W. et al . The community earth system model: a framework forcollaborative research. Bull. Am. Meteorol. Soc. 94, 1339–1360 (2013).

50. Bretherton, C. S., Widmann, M., Dymnidov, V. P., Wallace, J. M. & Blade, I.The effective number of spatial degrees of freedom of a time-varying field.J. Clim. 12, 1990–2009 (1999).


NATURE CLIMATE CHANGE | www.nature.com/natureclimatechange


http://apps.ecmwf.int/datasets

https://nomads.ncdc.noaa.gov/#cfsr

https://www.esrl.noaa.gov/psd/data

https://gmao.gsfc.nasa.gov/reanalysis/MERRA

http://jra.kishou.go.jp/JRA-55/index_en.html

https://www.ncdc.noaa.gov/data-access

http://esgf.llnl.gov


In the format provided by the authors and unedited.ARTICLES

PUBLISHED ONLINE: 13 MARCH 2017 | DOI: 10.1038/NCLIMATE3241












ARTICLESPUBLISHED ONLINE: 13 MARCH 2017 | DOI: 10.1038/NCLIMATE3241













SUPPLEMENTARY INFORMATIONDOI: 10.1038/NCLIMATE3241

NATURE CLIMATE CHANGE | www.nature.com/natureclimatechange 1


Supplementary Figures for Ding et al., “Influence of high-latitude atmospheric …”

2

This document includes the Supplementary Figures that are referred to in the main text.

Supplementary Fig. 1 Same as Fig. 1 d) to g), but using raw data in calculating the correlation.Stippling indicates statistical significant correlation at the 5% level.






3

Supplementary Fig 2 a) Meridional cross section of linear trend of zonal mean JJA vertical velocity (10-5 Pa/s/decade) in ERA-I (1979-2014). b) Domain averaged JJA lower level vertical velocity (1000hPa to 700hPa, unit: 10-5 Pa/s) in the Arctic (north of 70ºN). c) Correlation of omega index in (b) with JJA Z200 in 1979-2014. The trends are removed before the calculation. In c) stippling indicates statistical significant correlation at the 5% level.






4

Supplementary Fig 3. Correlation of a domain (north of 70ºN) averaged JJA cloudiness index in the upper level (HCC), middle level (MCC) and lower level (LCC) with JJA Z200 (a to c) in 1979-2014. The trends are removed before the calculation. Cloud data is from the ERA-Ireanalysis. Stippling indicates statistical significant correlation at the 5% level.






5

Supplementary Fig 4: Linear trends of a) meridional cross section of zonal mean JJA temperature (shading, ºC per decade) and geopotential height (black contour, m per decade, b) JJA lower level temperature (100hPa-750hPa) and c) September sea ice (% per decade) simulated in Exp-3 in which the model is nudged to observed ERA-I winds above 700hPa. Stippling indicates statistical significant trends at the 5% level.






6

Supplementary Fig. 5 Meridional cross section of linear trend of zonal mean temperature (shading: ºC per decade) and geopotential height (contour: m per decade) in Exp-4 (1979-2014) in each season.






7

Supplementary Fig. 6: a) Linear trend of JJA total sea ice melting in POP2-CICE4 run forced by ERA-I forcing during 1979-2014 period (Exp-5).b) Correlation between the domain averaged JJA total melting in the Arctic (north of 70ºN) and JJA Z200 during the period 1979-2014. c) same as b) but using the detrended components of the melting index and JJA Z200 in the Arctic.






8

Supplementary Fig. 7 a) & c) Meridional cross section of the linear trend of zonal mean JJA temperature (shading: ºC per decade) and geopotential height (contour: m per decade) in CMIP5 projects (1979-2014, upper panels) and CESM LENS (1979-2014, lower panels); b) &d) Linear trend of lower tropospheric (surface to 750hPa) JJA temperature (ºC per decade) in CMIP5 (1979-2014) and CESM LENS (1979-2014).






9

Supplementary Fig. 8: Linear trends of a) & b) meridional cross section of zonal mean JJA temperature (shading, ºC per decade) and geopotential height (black contour, m per decade, c) & d) JJA lower level temperature (1000hPa-750hPa) and e) & f) September sea ice (% per decade)simulated in two ECHAM5 nudged experiments (Exp-7 and 8) in which the global wind patterns (zonal and meridional) forced by anthropogenic forcing in CMIP5 (left column) and LENS (right column) projects are removed from ERA-I observed winds.






10

Supplementary Fig. 9: a) to f) Linear trends of JJA Z200 from six different reanalysis datasets and the time series of g) GL-Z200 derived from each reanalysis and six IGRA2 radiosonde stations (location of the stations is marked in panel a, 1989&1994 data are not used because data in this two years doesn’t pass the quality control). Linear trend of each index (m/decade) is denoted below its name.






11

Supplementary Table 1: 26 climate models in the CMIP5 historical experiment. List of 26CMIP5 CGCMs used in Fig. 4 to examine the forced response of the climate system to anthropogenic and natural external forcing, along with the number of atmospheric horizontal grids.

CMIP5 model designation nx ny1. ACCESS1-02. ACCESS1-33. bcc-csm1-1

192192128

14414464

4. bcc-csm1-1-m5. BNU-ESM

320128

16064

6. CCSM4 288 1927. CNRM-CM5 256 1288. CSIRO-Mk3-6-0 192 969. CanESM2 128 6410. FGOALS-g2 128 6011. GFDL-CM3 144 9012. GFDL-ESM2G 144 9013. GFDL-ESM2M 144 9014. GISS-E2-H 144 8915. GISS-E2-R 144 8916. HadGEM2-AO 192 14417. inmcm4 180 12018. IPSL-CM5A-LR 96 9619. IPSL-CM5A-MR 144 14320. IPSL-CM5B-LR 96 9621. MIROC-ESM 128 6422. MIROC5 256 12823. MPI-ESM-LR 192 9624. MPI-ESM-MR 192 9625. MRI-CGCM3 320 16026. NorESM1-ME 144 96





Documents

Influence of high-latitude atmospheric circulation changes ...david/Ding_etal_2017.pdf · (ms 1per decade) in ERA-Interim reanalysis. c, Domain-averaged time series for September