Influence of the recent high-latitude atmospheric circulation change on
summertime Arctic sea ice
Qinghua Ding1,2, Axel Schweiger2, Michelle L’Heureux3, David S. Battisti4,
Nathaniel C. Johnson5, Eduardo Blanchard Wrigglesworth4, Qin Zhang3, Kirstin Harnos3,
Ryan Eastman4, Eric J. Steig4,6
1.
Department of Geography, and Earth Research Institute, University of California, Santa Barbara, Santa
Barbara, California 93106, USA
2.
Polar Science Center, Applied Physics Laboratory, University of Washington, Seattle, Washington 98195,
USA
3.
NOAA Climate Prediction Center, College Park, Maryland 20740, USA,
4.
Department of Atmospheric Sciences, University of Washington, Seattle, Washington 98195, USA
5.
Cooperative Institute for Climate Science, Princeton University, Princeton, New Jersey 08540, USA
6.
Department of Earth and Space Sciences, University of Washington, Seattle, Washington 98195, USA
* Corresponding author ([email protected])
It is widely believed that the decline in Arctic sea ice coverage for the past three decades
is due to increasing greenhouse gases1-3, yet the mechanisms of this linkage are not well
understood4. Here, we present observational and modeling evidence that the rapid melting
of summertime sea ice was due to changes in the large-scale atmospheric circulation during
the Northern Hemisphere summer. A strengthening upper tropospheric anticyclonic
anomaly over Greenland and the Arctic Ocean increased the downwelling longwave
radiation above the ice by warming and moistening the lower Arctic troposphere. Our
experiments indicate that the circulation trend may have contributed as much as 60% to
the decline of the September sea ice extent minimum since 1979. Because the observed
circulation variability over the Arctic is inconsistent with the expected model response to
1
anthropogenic forcing, a significant component of sea ice loss over the last three decades
may have been driven by dynamical sources of natural climate variability.
It is well-recognized that recent Arctic sea ice decline has both natural and anthropogenic
drivers2,5, but their relative importance is poorly known6,7. This uncertainty arises from the fact
that the dynamical processes of intrinsic variability in Arctic climate system are poorly
understood; it is unclear how well models reproduce these processes8. Reliably distinguishing the
natural and anthropogenic contributions of sea ice loss requires a strong understanding of the
physical mechanisms that control the variability of sea ice. Although some progress has been
made in this area9-15, the answer is far from clear and conflicting hypotheses exist16,17.
Earlier work indicated that the decline of sea ice before the 1990s was in part owing to an
upward trend in the North Atlantic Oscillation (NAO) index18,19. However, since the early 1990s,
the apparent link between the NAO and sea ice has largely disappeared, with Arctic sea ice
declining further despite the reversal in the NAO trend20. On the other hand, several studies have
argued that the recent NAO trend may still play a key role in the sea ice retreat21,22. Another
mode of atmospheric circulation, the Arctic Dipole, has been suggested as a driver of sea ice
change in some regions of the Arctic23. In general, though, the success of these simple indices as
predictors of sea ice variability has been poor20.
In this paper we examine the contribution of the atmospheric circulation to Arctic sea ice
variability by utilizing an atmospheric general circulation model (ECHAM5) in which the
circulation field is nudged to observations. Specifically, we explore whether the high-latitude
summertime atmospheric circulation can significantly control the magnitude of the September
Arctic sea ice extent minimum, and whether the long-term circulation trend contributes
substantially to the observed sea ice loss.
2
To examine the physical linkages, we focus on the connection between September sea ice
extent and the preceding summer (June-July-August, JJA) atmospheric circulation. We choose
this preceding 3-month window because sea-ice extent anomalies have a ~3 month decorrelation
time-scale24, and previous studies have shown a strong link between summer circulation and sea
ice variability21,22. We focus on physical mechanisms, inspecting temperature, humidity, and
downward longwave radiation (DLR) that 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 growth and melt of Arctic sea ice25.
The region with the greatest negative trend in September sea ice concentration since 1979 is
highlighted in Fig. 1a, and includes the Beaufort, Chuchki, and East Siberian Seas, featuring an
average decline >10%/decade. Concomitant with the trend in sea ice are trends in atmospheric
circulation: 200 hPa geopotential heights have been rising over northeastern Canada and
Greenland and surface winds have become more anticyclonic (Fig. 1b). To illuminate process
that link the circulation changes to sea ice changes, we first construct an index of September sea
ice coverage over the region with the fastest sea ice decline from 1979 to 2014 (Fig. 1c), as well
as indices of JJA boundary layer temperature (surface to 750 hPa), DLR at surface, and
integrated atmospheric water vapor (surface to 750 hPa); indices are averaged over the Arctic,
poleward of 70ºN, and shown in Fig. 1c. The decreasing trend in sea ice concentration is
accompanied by increasing trends in JJA Arctic temperature, DLR and water vapor. Regressing
the trend in sea ice against the JJA geopotential height at 200hPa (Z200) in the ERA-Interim
reanalysis (hereafter ERA-I), we find decreasing sea ice is accompanied by increasing Z200,
with a maximum amplitude over Greenland (Fig. 1d; Supplementary Fig. 1).
3
These same interrelationships between the trends in the indices of JJA Z200, temperature,
water vapor, DLR and September sea ice concentration are also seen in the detrended indices: an
Arctic summer with, higher than normal Z200 over Greenland, greater DLR at surface, and
increased boundary layer water vapor and temperature over the Arctic is followed by negative
sea ice anomalies in September (Fig. 1c; e.g., indices of temperature and specific humidity are
correlated at r=0.94; indices of sea ice concentration and DLR are correlated at r = -0.85).
Importantly, the detrended data shows that the summertime anomalies in the indices of nearsurface temperature, DLR, and water vapor are associated with a common pattern of atmospheric
circulation variability (cf. Figs 1d-g), and that this circulation pattern is very similar to the
circulation pattern that is associated with the sea ice trend (cf. Figs. 1d-g and Supplementary Fig.
1a-d). These correlation maps also compare well with the pan-Arctic Z200 and surface wind
trends in JJA (Fig. 1b), which feature a strong high over northeastern Canada, Greenland and the
Arctic Ocean.
That there is a common pattern of atmospheric circulation anomaly that is linked to
temperature, water vapor, and DLR over the Arctic in JJA and subsequently linked to sea ice
anomalies in September, and that the trends in this same circulation pattern is also similarly
linked to trends in these variables, suggests the September sea ice trends are causally linked to
tends in atmospheric circulation.
The linkage between the large-scale upper level circulation and lower level thermodynamics is
apparent in the zonal-vertical cross-section of the atmosphere (Fig. 2). The recent JJA circulation
trend in the Arctic was largest in the upper troposphere (200hPa), while atmospheric warming
mainly occurred in the lower troposphere below 700hPa (Fig. 2a). Associated with this trend,
there was significant downward motion below 700hPa and poleward of ~75°N (Supplementary
4
Fig. 2). Year-to-year variability in summertime vertical motion in the lower troposphere is also
highly correlated with similar circulation changes in the upper troposphere (Supplementary Fig.
2). Given a strong coupling between the circulation and vertical velocity it is clear that the near
surface warming is due, in part, to adiabatic descent forced by the anticyclonic circulation aloft.
The warming of the surface (increased DLR) is a direct result of the warming of the lower
atmosphere and the increase in water vapor in the lower (warm) atmosphere. DLR is also
strongly affected by cloud cover. To examine the relationship between cloudiness and circulation,
we examine the relationship between the detrended Z200 heights and cloud anomalies in the
Arctic. An anomalous high pressure over the Arctic Ocean favors a decrease in cloudiness in the
upper and middle levels of the atmosphere, possibly associated with decreased storm activity.
However, low clouds increase slightly over Greenland and the central Arctic (Supplementary Fig.
3), likely due to a reduction in mixing of air out of the boundary layer. Because low clouds
dominate in the Arctic, this circulation-driven downward shift in cloudiness is qualitatively
consistent with the observed increase in DLR.
We have provided a plausible mechanism for how circulation changes can impact Arctic sea
ice, but it is difficult to determine causality with observational evidence alone because of the
feedbacks between sea ice and the atmosphere. For example, the removal of sea ice can increase
surface air temperatures, clouds, low level water vapor and hence DLR26, and hence may affect
the atmospheric circulation itself27,28. To better understand the direction of causality, we conduct
two model experiments to examine the influence of the observed high-latitude circulation trend
on the sea ice and other key variables. The first experiment (Exp-1) is a thirty-six-year historical
run with the ECHAM5 atmospheric model, in which the global 3-D atmospheric circulation
(vorticity and divergence from the surface to the top of the atmosphere) is loosely nudged to the
5
corresponding daily reanalysis data (see Methods). For the lower boundary, climatological SST,
and sea ice are used everywhere. In the second experiment (Exp-2), all settings are the same as
those in Exp-1, except that a slab ocean-sea ice model is used in the Arctic (north of 60º N),
permitting a feedback from the ice-ocean system to the prescribed atmosphere. Anthropogenic
forcings are held constant in both experiments. The difference between the two experiments
quantifies the role of the circulation in controlling the thermodynamics versus the amplifying
effects of sea ice-ocean feedbacks. If the mechanism behind the strong correlations of
thermodynamical variables (temperature, humidity and DLR) with the Z200 circulation is owing
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.
Both Exp-1 and Exp-2 produce similar changes in geopotential height and temperatures
with respect to height and latitude, as upper-level geopotential heights and lower-level
temperature exhibit positive trends in the Arctic (Figs. 2b and c). While the simulated spatial
patterns are similar to the reanalysis data, the height increases are somewhat smaller and
temperature increases larger than observed. The simulated lower tropospheric temperature in the
experiments is significantly correlated with the ERA-I reanalysis data on both interannual and
interdecadal time scales. The region with the strongest lower-tropospheric warming in JJA is
over the Beaufort, Chuchki, and East Siberian Seas (Fig. 2d). The results from Exp-2 are very
similar to those in Exp-1 and Exp-2, indicating that the lower level atmospheric warming is
largely due to the changes in atmospheric circulation, while the forcing of the atmosphere by the
sea ice appears to be small (Figs. 2e and f). This is not entirely surprising: temperature contrasts
between the sea ice and the atmosphere are small in JJA, the sea ice concentration is reduced and
the surface is riddled with melt water ponds. Hence, changes in sea ice conditions in JJA only
6
have a small impact on surface temperature, atmospheric humidity and DLR. The sea ice trends
simulated in Exp-2 are similar in magnitude but “upstream” of those observed (cf. Fig. 2g with
1c), likely due to a lack of advection of ice by the mean winds (which is precluded by our use of
a slab ocean, thermodynamic-only sea ice model).
It is possible that the circulation itself is a response to the sea ice change and the prescribed
winds are the result of the thermal wind relationship. If so, the observed boundary layer warming
would be expected when the circulation is nudged to the observed winds as done in both Exp-1
and Exp-2. However, model experiments that examine the impact of changes in sea ice
conditions in isolation show no significant impact on the circulation during the summer
months27,29. Because the impact of sea ice changes on the circulation may be model dependent,
we examine this impact in an additional experiment (Exp-3; see Methods), in which we specify
the observed time evolution (1979-2014) of Arctic SST and sea ice (north of 60ºN), and use
climatological SST elsewhere. In JJA, the model response to the SST warming trend and sea ice
retreat in the Arctic is very small (Supplementary Fig. 4), in contrast to the responses in the other
three seasons. This is because in summer, temperature contrasts between the open ocean and sea
ice are relatively small (compared with fall and winter), and the impact of sea ice loss on the
atmosphere is muted. Even in fall and winter when the model exhibits the largest responses to
sea ice melt, the simulated circulation trend is only about 3m/decade in the upper troposphere,
which is only 10 to 20% of the observed change in JJA. Thus, we conclude that the JJA vertical
temperature trend in the Arctic in the past 36 years is mainly controlled by the changes in the
mid and upper troposphere (top-down effect), rather than a bottom-up response due to the sea ice
melt.
7
Having demonstrated the impact of the circulation on near surface temperature, water vapor
and DLR, we aim to quantify its contribution to the overall downward trend in sea ice. As noted
above, the pattern of sea ice decline simulated in Exp-2 is shifted upstream of the observed sea
ice decline (Figs. 1a and 2g). Without sea ice drift, the strongest reduction in sea ice in Exp-2
occurs in the Beaufort Sea, where the circulation exerts the largest forcing on lower level
temperature and DLR at the surface over the last 36 years (Fig. 2h). To illustrate the role of sea
ice advection, we force a global coupled ocean/sea-ice (POP2-CICE4) model with observed
atmospheric fields (ERA-I reanalysis; see Methods) between 1979-2014, Two experiments are
performed: Exp-4 which uses the complete set of ERA-I forcings, and Exp-5 which removes
from the forcing the effect of the circulation variability and trends that are linearly related to
Z200 over Greenland, identified above (see Methods).
The spatial patterns and time series of sea ice in the control run (Exp-4) exhibit a close
correspondence to the observed sea ice reduction (cf. Figs. 1a and 3a). A comparison of sea ice
concentration and sea ice volume patterns and time series supports our hypothesis. In the absence
of forcing due to the Greenland circulation pattern (Exp-5), the decline in sea ice concentration
and thickness are only 41% and 40% of the decline in the control run (Exp-4), respectively,
suggesting that the summertime circulation contributes to as much as a 60% of the sea ice loss
since 1979.
Having attributed a substantial part of September sea ice decline to variability in the high
latitude circulation during the preceding summer, we now examine whether long-term changes in
the circulation are driven by natural or anthropogenic variability. Upper tropospheric circulation
change in the Arctic is subject to strong natural variability originating from the tropics30-33.
Previous research32 identified a relationship between tropical SST variability and annual mean
8
atmospheric circulation over the Arctic with a center of action over Greenland. Model
experiments32 show that about 50% of the circulation change and the associated warming over
Greenland is attributable to natural variability originating from the tropical Pacific Ocean. An
“ad hoc” attribution through a combination of these components suggests that ~30% (= 50% ×
60%) of the sea ice decline observed since 1979 during JJA is attributable to natural variability in
the tropical Pacific. An alternate hypothesis is that changes in the Arctic circulation itself are the
result of anthropogenic forcing. To test this hypothesis, we examine the JJA Z200 trends over the
Arctic region in simulations from the Large Ensemble Project (LENS) using the Community
Earth System Model (CESM). Each simulation in the ensemble was forced with identical
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
initial conditions for each simulation were slightly different. The average over the 30-member
ensemble displays positive trends in near-surface temperature and in Z200 over the course of our
study period (1979-2014). Unlike in the observations, however, the trends associated with the
historical forcing are smooth and uniform (cf, Fig. 4b with 4a or 1b) and are typical of the
thermodynamic response in the mid- and high latitudes to CO2 forcing (Supplementary Fig. 5).
The stronger gradients in the observed trend pattern (Fig. 2a) support the idea that natural
variability is an important contributor to the multidecadal variability of the high-latitude
circulation and sea ice over the past four decades. This conclusion assumes that the CESM
accurately reflects the response of the high latitude circulation to anthropogenic forcing34.
Attribution estimates of Arctic sea ice loss based on CESM therefore need to be viewed with
caution.
9
In addition, it is important to recall that we have used reanalysis data as a proxy for
observational data. Reanalysis products are reliable representations of the observed circulation
humidity and temperature, but less so for cloudiness. The variability in total cloud cover in the
satellite data is well represented in the reanalysis data seems, particularly in summer35. However,
there is evidence that the relationship between sea ice and the vertical distribution cloud fraction
in reanalysis does not agree well with that in observations26. Therefore, to the extent that changes
in cloudiness contribute to the DLR trends associated with the trends in circulation, the
connection between large-scale circulation and cloud variability is subject to considerable
uncertainty and needs to be explored in future studies. We also must consider the limitations in
the design of our experiment (Exp-5), which constructs atmospheric forcing fields based on a
linear regression of a Z200 index, to quantify the circulation impact on sea ice loss.
How sea ice variability and trends can impact the Arctic atmospheric circulation is an area of
vigorous research36. Studies suggest numerous mechanisms in which sea ice loss modulates the
large-scale circulation in the lower troposphere in winter37-39. This paper, instead, focuses on
how the high-latitude circulation impacts sea ice. Although positive feedbacks between sea ice
and the Arctic circulation exist, we find that these are small during summer. Instead, circulation
variations over the Arctic have been a dominant factor in driving sea ice variability since 1979,
and has had a non-trival (~50%) contribution to the total surface temperature trend over
Greenland and northeastern Canada32. The forcing of the summer sea ice by the trends in largescale circulation, which are likely due mostly to natural variability, represents an important
driver of the observed Arctic climate change. Therefore, future changes of the high latitude,
large-scale circulation represent a significant source of uncertainty for projections of the Arctic
summertime sea ice, with ramifications for the rest of the Arctic and the global climate system.
10
Reference
1. 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, 6 (2008).
2. Kay, J. E., Holland, M. M. & Jahn, A. Inter-annual to multi-decadal Arctic sea ice extent
trends in a warming world. Geophys. Res. Lett. 38, L15708 (2011).
3. Notz, D. & Marotzke, J. Observations reveal external driver for Arctic sea-ice retreat.
Geophys. Res. Lett. 39, L08502 (2012).
4. Stroeve, J., Holland, M. M., Meier, W., Scambos, T. & Serreze, M. Arctic sea ice decline:
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. Environmental Research Letters 7, 034011 (2012).
6. Rampal, P., Weiss, J., Dubois, C. & Campin, J.-M. IPCC climate models do not capture
Arctic sea ice drift acceleration: Consequences in terms of projected sea ice thinning and
decline. J. Geophys. Res. 116, C00D07 (2011).
7. Swart, N. C., Fyfe, J. C., Hawkins, E., Kay, J. E. & Jahn, A. Influence of internal variability
on Arctic sea-ice trends. Nature Climate Change 5, 86-89 (2015)
8. Zhang, R. Mechanisms for low frequency variability of summer Arctic sea ice extent. Proc.
Nat. Aca. Sci. 112, 1422296112 (2015).
9. Francis, J. A. & Hunter, E. New insight into the disappearing Arctic sea ice. Eos Trans. AGU
87, 509–511 (2006).
10. Graversen, R. G. & Wang, M. Polar amplification in a coupled climate model with locked
albedo. Climate Dyn. 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 recent Arctic
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 by the thermal
inversion and consequent low infrared cooling to space. Nat. Geosci. 4, 758–761 (2011).
15. Vaughan, D.G. et al., Observations: Cryosphere. In: Climate Change 2013: The Physical Science Basis. Cambridge University Press (2013).
16. Screen, J. A., Deser, C. & Simmonds, I. Local and remote controls on observed Arctic
warming. Geophys. Res. Lett. 39, L10709 (2012).
17. Perlwitz, J., Hoerling, M. & Dole, R. Arctic tropospheric warming: Causes and linkages to
lower latitudes. J. Climate 28, 2154–2167 (2015).
18. Deser, C., Walsh, J. E. & Timlin, M. S. Arctic sea ice variability in the context of recent
atmospheric circulation trends. J. Climate 13, 617-633 (2000).
19. Hu, A., Rooth, C., Bleck, R. & Deser, C. NAO influence on sea ice extent in the Eurasian
coastal region. Geophys. Res. Lett. 29, 2053 (2002).
20. Deser, C. & Teng, H. Evolution of Arctic sea ice concentration trends and the role of
atmospheric circulation forcing, 1979–2007. Geophys. Res. Lett. 35, L02504 (2008).
21. Ogi, M., Rigor, I. G., McPhee, M. G. & Wallace, J. M. Summer retreat of Arctic sea ice:
Role of summer winds. Geophys. Res. Lett. 35, L24701 (2008).
11
22. Ogi, M., Yamazaki, K. & Wallace, J. M. Influence of winter and summer surface wind
anomalies on summer Arctic sea ice extent. Geophys. Res. Lett. 37, L07701 (2010).
23. Watanabe, E., Wang, J., Sumi, A. & Hasumi, H. Arctic dipoleanomaly and its contribution to
sea ice export from the Arctic Ocean in the 20th century. Geophys. Res. Lett. 33, L23703
(2006).
24. Blanchard-Wrigglesworth E, Armour, K. C., Bitz, C. M. & DeWeaver, E. Persistence and
inherent predictability of Arctic sea ice in a GCM ensemble and observations. J. Climate, 24,
231-250 (2011).
25. Kay, J. E., L'Ecuyer, T., Gettelman, A., Stephens, G. & O'Dell, C. The contribution of cloud
and radiation anomalies to the 2007 Arctic sea ice extent minimum. Geophys. Res. Lett. 35,
L08503 (2008).
26. Schweiger, A. J., Lindsay, R.W., Vavrus, S. & Francis, J. A. Relationships between Arctic
clouds and sea ice during autumn. J. Climate 21, 4799-4810 (2008).
27. Bhatt, U. S. et al. The atmospheric response to realistic reduced summer Arctic sea ice
anomalies, in Arctic sea ice decline: Observations, projections, mechanisms, and
implications, American Geophysical Union, Washington, D.C. (2008).
28. Liu, J. P. et al. Has Arctic sea ice loss contributed to increased surface melting of the
Greenland ice sheet? J. Climate, 29, 3373-3386 (2016).
29. Deser, C., Tomas, R., Alexander, M. & Lawrence, D. The seasonal atmospheric response to
projected Arctic sea ice loss in the late 21st century. J. Climate 23, 333-351 (2010).
30. Hoerling, M. P., Hurrell, J. W. & Xu, T. Y. Tropical origins for recent North Atlantic climate
change. Science 292, 90-92 (2001).
31. Lee, S. Testing of the tropically excited Arctic warming (TEAM) mechanism with traditional
El Nino and La Nina. J. Climate 25, 4015-4022 (2012).
32. Ding, Q. H. et al. Tropical forcing of the recent rapid Arctic warming in northeastern Canada
and Greenland. Nature 509, 209-212 (2014).
33. Trenberth, K. E., Fasullo, J. T., Branstator, G. & Phillips, A. S. Seasonal aspects of the recent
pause in surface warming. Nature Climate Change 4, 911–916 (2014)
34. Wettstein, J. J. & Deser, C. Internal variability in projections of twenty-first century Arctic
sea ice loss: Role of the large-scale atmospheric circulation. J. Climate 27, 527-550 (2014).
35. Liu, Y. & Key, J. Assessment of Arctic cloud cover anomalies in atmospheric reanalysis
products using satellite data. J. Climate. doi:10.1175/JCLI-D-15-0861.1 (2016).
36. Barnes, E. A. & Screen, J. The impact of Arctic warming on the midlatitude jetstream: Can
it? Has it? Will it?. WIREs Climate Change 6, (2015).
37. Overland, J. E. & Wang, M. Increased variability in the early winter subarctic North
American atmospheric circulation. J. Climate 28, 7297–7305 (2015).
38. Francis, J. A. & Skific, N. Evidence linking rapid Arctic warming to mid-latitude weather
patterns. Phil. Trans. R. Soc. A, 373, (2015).
39. Cohen, J. An observational analysis: Tropical relative to Arctic influence on midlatitude
weather in the era of Arctic amplification. Geophys. Res. Lett. 43, 5287–5294 (2016).
40. Zhang, J. & Rothrock, D. A. Modeling global sea ice with a thickness and enthalpy
distribution model in generalized curvilinear coordinates. Mon. Wea. Rev. 131, 845–861
(2003).
12
Methods
Data
We use the monthly circulation, temperature, cloud and radiation data from the ERA-Interim
reanalysis41 during the period 1979-2015. We also analyze observed monthly sea ice
concentration data derived from Goddard edited passive microwave retrievals and compiled by
the National Snow and Ice Data Center. We also make extensive use of available experiments
from the CESM Large Ensemble Community Project42. Thirty realizations, each with slightly
perturbed initial conditions, of CESM1 historical and future projection simulations are available
from 1920 to 2100. We focus on the historical satellite era from 1979-2014.
Models
The general circulation model used to perform the experiments (Exps-1 to 3) in this study was
the ECHAM5 atmospheric general circulation model43, with a horizontal resolution of T42 (~2.8°
latitude × 2.8° longitude) and 19 vertical levels. In Exp-2, we coupled the ECHAM5 to a slab
ocean in the high-latitudes to assess the role of prescribed circulation in driving the SST and sea
ice in the Arctic. The slab ocean is represented as boxes of water of uniform specified depth (50
m). A simple, thermodynamic-only sea ice model is included when the ocean temperature
reaches the freezing point. The ocean temperature or sea ice condition at each grid point is
affected only by heat exchange across the air–sea interface; there is no direct communication
between adjacent ocean grid points, nor is there any representation of the deep ocean. In addition,
a cyclo-stationary climatological heat flux is added to the ocean temperature tendency equation
to maintain a seasonal cycle of ocean temperature and sea ice conditions that are close to the
observed climatology during 1979-1995.
13
In Exps-4 and 5, the sea-ice-ocean component (POP2+CICE4) in the latest version of CESM44
is used. This is a model of the ocean and sea ice that includes thermodynamics and dynamics; the
model is run using tripole 1-degree ocean/sea ice grids. Specific details on each model
experiment are listed in a later section.
Significance of correlation
For the correlation significance tests in this study, the degrees of freedom are adjusted by the
computation of an “effective sample size”, which accounts for the autocorrelation in the time
series. This effective sample size N* is calculated as
N* = N
1 − r1r2
1 + r1r2
where N is the number of available time steps and r1 and r2 are lag-one autocorrelation
coefficients of each variable45.
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 observed circulation in
our model simulations, and to facilitate our computation efforts, we interpolate observed monthly
ERA-I data to daily fields for nudging. We also use a very weak damping term to nudge the 3-D
(from surface to the top of atmosphere) divergence and vorticity fields of the model to the
observed monthly (smoothly interpolated to daily) fields in the last 36 years; this weak damping
allows the model to generate its own day-to-day variability but constraints the model to be very
close to the observed circulation on monthly and longer time scales. In the lower boundary, we
used climatological SST/sea ice everywhere. Anthropogenic forcings were held constant in this
experiment.
14
Exp-2: Same as Exp1, except that a simple slab ocean/sea ice model was used in the Arctic
(north of 60ºN).
Exp-3: In ECHAM5, the model is forced by observed history of sea ice concentration and SST in
the Arctic (north of 60ºN) and climatological SST/sea ice everywhere else during the period
1979-2014. Twenty realizations were run, each with different atmospheric initial conditions; the
20-member ensemble mean represents the model response. Anthropogenic forcings were held
constant in this experiment.
Exp-4: PoP2+CICE4 control run
POP2+CICE4 is used to simulate the sea ice variability over the past 36 years using prescribed
ERA-Interim atmospheric forcing. The forcing fields include daily data of 10m wind,
downwelling shortwave and longwave radiation, 2m temperature and humidity, sea level
pressure and precipitation.
Exp-5: Same as Exp-4 except that the atmospheric forcing is modified to excise the forcing
associated with the trends in the Greenland circulation pattern
To accomplish this goal, we construct the thirty-six-year seasonal (JJA) mean time series of
the Z200 index over Greenland, 𝑍200!" (GL-Z200 in Fig. 1c). We then linearly regress a key
variable B against this time series to obtain spatial pattern β(x,y) of the variable associated with
the Greenland circulation index. Specifically, for the variable B we have
𝐵 𝑥, 𝑦, 𝑡 = 𝛽×𝑍200!" (𝑡) ,
Eq.1
where B represents the forcing field (e.g., 10 m zonal wind, DLR, temperature, etc.), x and y
indicate the location, t indicates time (JJA), Z200GL is the Greenland Z200 index (GL-Z200 in
Fig. 1c), and β is the regression coefficient.
15
In the second step, the seasonal mean anomalous value of each forcing field is subtracted from
the observed daily (or 6-hourly) forcing data during the summer – rendering a modified forcing
that does not include variability or trends in variables that are associated with Z200GL. In the nine
non-summer months, the forcing is exactly the same as that used in the Exp-4 control experiment.
Given a strong correlation between circulation and surface winds, temperature, specific humidity,
sea level pressure, and downwelling longwave radiation in the Arctic, variability and trends in
these six variables that are associated with Z200GL are processed and removed from the forcing.
The initial states of ocean, sea ice and atmosphere in Exp-4 and Exp-5 are exactly the same.
41. 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).
42. Kay, J. E. et al. The Community Earth System Model (CESM) Large Ensemble Project: A
community resource for studying climate change in the presence of internal climate
variability. Bull. Amer. Meteor. Soc. 96, 1333-1349 (2015).
43. Roeckner, E. et al. Max Planck Institut für Meteorologie Report 349, 140 (2003).
44. Hurrell, J. W. et al. The community earth system model: A framework for collaborative
research. Bull. Amer. Meteor. Soc. 94, 1339–1360 (2013).
45. 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. Climate 12, 1990–2009
(1999).
Acknowledgements. This work was supported by a NOAA Climate Programs Office/CVP grant
(NA15OAR4310162). We thank the Max Planck Institute for Meteorology and National Center
for Atmospheric Research model developers for making the ECHAM5 and CESM available and
J. M. Wallace, C. Bitz, Q. Fu, M. Wang, D. L. Hartmann and D. Frierson for discussion. We
acknowledge the CESM Large Ensemble Community Project and supercomputing resources
provided by NSF/CISL/Yellowstone. Q.-H. D. acknowledges support from the University of
Washington’s Polar Science Center and the Future of ice program, and the Tamaki Foundation.
A. S. is grateful for funding from the National Science Foundation through grant ARC-1203425.
D. B. acknowledges support from the Tamaki Foundation. R. E. acknowledges support from
NASA NNXBAQ35G.
Author Contributions. Q.D. led this work with contributions from all authors. Q.-H. D. made
the calculations, implemented the general circulation model experiments, created the figures, and
led writing of the paper. All authors contributed to the experimental design, interpreting results
and writing the paper.
16
Figure Captions
Fig. 1 Relationship between the September Arctic sea ice and summer large-scale
circulation.
a) Linear trend (% per decade) of September sea ice concentration 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 per decade) in ERA-Interim reanalysis; c) Domain averaged time series for
September sea ice anomaly (unit: %) averaged over the region circled by the blue contour in (a),
lower level (1000hPa to 750 hPa) JJA temperature (ºC) and JJA specific humidity anomalies
(g/kg) in the Arctic (averaged over the region north of 70ºN), JJA downwelling longwave
radiation anomaly at surface (W/m2) in the Arctic north of 70ºN, and JJA Z200 anomaly (m)
over Greenland (66-80ºN, 310-330ºE, indicated by the dot in (b), and referred as GL-Z200); d) to
g) correlation of each time series in (c) with JJA Z200 for 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 of comparison with other plots. In f), regression between the specific
humidity index and vertically integrated water vapor flux are plotted. All linear trends are
removed in calculating the correlations in d) to g). Black stippling in all plots indicates
statistically significant correlation or trend at the 5% level; in panels d and f, vectors are plotted
when regressions are statistically significant at the 5% level.
Fig. 2 Simulated impact of atmospheric circulation on Arctic thermodynamic trends.
a) Meridional cross section of the linear trend of zonal mean JJA temperature (shading, ºC per
decade) and geopotential height (contour, m per decade) in ERA-I (1979-2014); b) and c), same
as a) but for Exp-1 and Exp-2 simulations; d) Linear trend of lower tropospheric (1000hPa to
750hPa) JJA temperature in ERA-I (1979-2014); e) and f), same as d) but for Exp-1 and Exp-2
simulations; g) Linear trend of September sec ice concentration (% per decade) simulated in
Exp-2; h) Domain averaged time series for lower tropospheric (1000hPa to 750hPa) JJA
temperature (ºC) and specific humidity anomalies (unit: g/kg), and JJA downwelling longwave
radiation anomaly (W/m2) at surface in the Arctic (north of 70ºN) simulated in Exp-2. In d) to g)
stippling indicates statistical significant trend at the 5% level.
Fig. 3 Simulated impact of atmospheric circulation on summertime Arctic sea ice trends.
a), b), d) and e) Linear trend of September sea ice concentration (upper panels, % per decade)
and thickness (lower panel, m per decade) in Exp-4 (denoted as “full forcing”) and Exp-5
(denoted as “modified forcing”); c) Anomalous total area of September sea ice extent (area of
ocean with ice concentration of at least 15%) in both simulations and NSIDC observations. d)
anomalous total volume of September sea ice (area of ocean with ice concentration of at least
15%) in both simulations and the Pan-Arctic Ice Ocean Modeling and Assimilation System
(PIOMAS40).
Fig. 4: Observed and estimated radiatively forced trends of upper tropospheric
geopotential height.
Linear trend of JJA Z200 (m per decade) for the period 1979-2014 from a) ERA-I (repeated from
Fig. 1b) and b) the CESM Large Ensemble project (LENS).
17
Fig. 1 Relationship between the September Arctic sea ice and summer large-scale
circulation.
a) Linear trend (% per decade) of September sea ice concentration 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 per decade) in ERA-Interim reanalysis; c) Domain averaged time series for
September sea ice anomaly (unit: %) averaged over the region circled by the blue contour in (a),
lower level (1000hPa to 750 hPa) JJA temperature (ºC) and JJA specific humidity anomalies
(g/kg) in the Arctic (averaged over the region north of 70ºN), JJA downwelling longwave
radiation anomaly at surface (W/m2) in the Arctic north of 70ºN, and JJA Z200 anomaly (m)
over Greenland (66-80ºN, 310-330ºE, indicated by the dot in (b), and referred as GL-Z200); d) to
g) correlation of each time series in (c) with JJA Z200 for 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 of comparison with other plots. In f), regression between the specific
humidity index and vertically integrated water vapor flux are plotted. All linear trends are
removed in calculating the correlations in d) to g). Black stippling in all plots indicates
statistically significant correlation or trend at the 5% level; in panels d and f, vectors are plotted
when regressions are statistically significant at the 5% level.
18
Fig. 2 Simulated impact of atmospheric circulation on Arctic thermodynamic trends.
a) Meridional cross section of the linear trend of zonal mean JJA temperature (shading, ºC per
decade) and geopotential height (contour, m per decade) in ERA-I (1979-2014); b) and c), same
as a) but for Exp-1 and Exp-2 simulations; d) Linear trend of lower tropospheric (1000hPa to
750hPa) JJA temperature in ERA-I (1979-2014); e) and f), same as d) but for Exp-1 and Exp-2
simulations; g) Linear trend of September sec ice concentration (% per decade) simulated in
Exp-2; h) Domain averaged time series for lower tropospheric (1000hPa to 750hPa) JJA
temperature (ºC) and specific humidity anomalies (unit: g/kg), and JJA downwelling longwave
radiation anomaly (W/m2) at surface in the Arctic (north of 70ºN) simulated in Exp-2. In d) to g)
stippling indicates statistical significant trend at the 5% level.
19
Fig. 3 Simulated impact of atmospheric circulation on summertime Arctic sea ice trends.
a), b), d) and e) Linear trend of September sea ice concentration (upper panels, % per decade)
and thickness (lower panel, m per decade) in Exp-4 (denoted as “full forcing”) and Exp-5
(denoted as “modified forcing”); c) Anomalous total area of September sea ice extent (area of
ocean with ice concentration of at least 15%) in both simulations and NSIDC observations. d)
anomalous total volume of September sea ice (area of ocean with ice concentration of at least
15%) in both simulations and the Pan-Arctic Ice Ocean Modeling and Assimilation System
(PIOMAS40).
20
Fig. 4: Observed and estimated radiatively forced trends of upper tropospheric
geopotential height.
Linear trend of JJA Z200 (m per decade) for the period 1979-2014 from a) ERA-I (repeated from
Fig. 1b) and b) the CESM Large Ensemble project (LENS).
21
Supplementary Figures for Ding et al., “Influence of the Recent High-Latitude …”
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.
1
Supplementary Figures for Ding et al., “Influence of the Recent High-Latitude …”
Supplementary Fig 2 a) Meridional cross section of linear trend of zonal mean JJA vertical
velocity (10-5 Pa/s) 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.
2
Supplementary Figures for Ding et al., “Influence of the Recent High-Latitude …”
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-I
reanalysis. Stippling indicates statistical significant correlation at the 5% level.
3
Supplementary Figures for Ding et al., “Influence of the Recent High-Latitude …”
Supplementary Fig. 4 Meridional cross section of linear trend of zonal mean temperature
(shading: ºC per decade) and geopotential height (contour: m per decade) in Exp-3 (1979-2014)
in each season.
4
Supplementary Figures for Ding et al., “Influence of the Recent High-Latitude …”
Supplementary Fig. 5 a) Meridional cross section of the linear trend of zonal mean JJA
temperature (shading: ºC per decade) and geopotential height (contour: m per decade) in CESM
LENS (1979-2014); b) Linear trend of lower tropospheric (surface to 750hPa) JJA temperature
(ºC per decade) in CESM LENS (1979-2014).
5