1. Introduction
There is mounting observational evidence of drastic climate change in the Arctic, ranging from considerable sea ice loss (e.g., Rothrock et al. 1999; Wadhams and Davis 2000; Comiso 2002; Serreze et al. 2003) to rapid surface warming (e.g., ACIA 2004; Serreze and Francis 2006; IPCC 2013). Still more changes are expected to occur in future decades, with comprehensive climate models projecting that Arctic surface air temperatures will warm by about 5°C by the end of the twenty-first century—faster than any other region on Earth (IPCC 2013)—and that there will be a complete disappearance of summer Arctic sea ice by midcentury (Holland et al. 2006).
While climate change in the Arctic is driven largely by increases in long-lived greenhouse gases (GHGs), increases in shorter-lived trace species and aerosols have also accelerated warming by altering the radiative and chemical properties of the Arctic. For example, in recent decades increased black carbon deposition on snow and ice has significantly enhanced surface longwave fluxes over the Arctic and may have been twice as effective as carbon dioxide at warming the Arctic surface (Koch and Hansen 2005). Simulations with comprehensive climate models also indicate that increased levels of ozone precursors, including nitrogen oxides and volatile organic compounds, have contributed as much as 30% to the observed positive trends in twentieth-century Arctic surface temperatures by increasing high-latitude tropospheric ozone (Shindell et al. 2006). Therefore, a comprehensive understanding of the current and future distributions of chemical and particulate tracers in the Arctic is key for understanding climate.
It is now well appreciated that nearly all of the pollution in the Arctic originates over Northern Hemisphere (NH) midlatitudes (Law and Stohl 2007). Since the distributions of trace species reflect the full interplay between emissions, chemistry, and transport, Arctic pollution in the future will reflect not only changes in species’ emissions and chemistry, but also changes in the large-scale circulation. However, while the climate-change signature on large-scale dynamics has been examined in both models and observations [e.g., shifts in the midlatitude tropospheric jets (e.g., Yin 2005; Miller et al. 2006; Barnes and Polvani 2013), the expansion and weakening of the Hadley cell (e.g., Lu et al. 2007), and trends in atmospheric variability (e.g., Hurrell 1995; Thompson et al. 2000; Zhou et al. 2001)], relatively little attention has been paid to assessing the large-scale response of transport into the Arctic.
Here we quantify tropospheric transport using idealized tracers that partition the air in the Arctic according to the regions where it last contacted the planetary boundary layer (PBL). In Orbe et al. (2015, hereinafter Part I), we presented the first model climatology of Arctic air-mass origin in terms of rigorously defined air-mass fractions f(r | Ωi) that quantify the fraction of air at location r that last contacted the PBL over the origin region Ωi. (Note that the term “origin” is used in reference to the region where air last contacted the PBL.) In practice f(r | Ωi) is calculated as a simple equilibrated tracer mixing ratio that shows where in the Arctic, and with what dilution, the air from an origin region can be found.
Air-mass origin climatologies for NH winter [December–February (DJF)] and NH summer [June–August (JJA)] were presented in Part I based on calculations from a time-slice integration of the Goddard Earth Observing System Chemistry–Climate Model (GEOSCCM) subject to forcings representative of the present-day climate [i.e., fixed 2010–19 time-averaged GHGs and ozone-depleting substances (ODS)]. It was shown that the Arctic middle and upper troposphere (i.e., above 700 hPa) consists largely of air that last contacted the PBL over latitudes between 25° and 60°N, defined herein as NH midlatitudes. Last contact at the midlatitude PBL occurs primarily over the oceans during NH winter and over land during NH summer, consistent with ventilation of the midlatitude boundary layer by the storm tracks and large-scale convection, respectively. It was also shown that during NH winter last contact at the midlatitude surface occurs primarily over the eastern Pacific, where strong poleward flow ensures that air is efficiently transported to the Arctic with little chance of reencountering the PBL. By comparison, during summer air of NH midlatitude origin last contacts the PBL primarily over Asia, consistent with strong convection and mean poleward flow over Siberia.
In Part I different features of the large-scale circulation were used to interpret the seasonal cycle of air-mass origin in the Arctic and its partitioning with respect to the different PBL regions. Here we ask how future changes in the midlatitude storm tracks, large-scale stationary waves, and large-scale vertical motions over NH midlatitudes will affect transport into the Arctic in terms of the regions where air last contacts the PBL. In particular, recent studies have shown that, while comprehensive climate models project that the zonal-mean midlatitude tropospheric jet will shift poleward by the end of the twenty-first century, the longitudinally varying response is highly variable across basins and between seasons, at places featuring robust equatorward shifts (Barnes and Polvani 2013; Simpson et al. 2014). And yet, while strong longitudinal variations in the jet response may have large impacts on regional transport and climate (Simpson et al. 2014), these transport changes have yet to be assessed.
In addition to future changes in the tropospheric midlatitude jet, comprehensive models also indicate that dry static stability over midlatitudes will increase in response to GHG-induced warming, with the largest increases occurring during NH summer (Wetherald and Manabe 1988; Frierson 2006). Although increases in tropospheric stability have been linked to the projected weakening and delayed onset of the North American monsoon in a warmer climate (Cook and Seager 2013), a systematic examination of the corresponding transport response in the Arctic has not been performed.
Here we examine how transport into the Arctic will change by the end of the twenty-first century by examining differences in the climatological air-mass fractions between two time-slice integrations of GEOSCCM: the present-day or reference (REF) integration presented in Part I and a future (FTR) integration forced with greenhouse gases and ozone-depleting substances representative of the end of the twenty-first century. After briefly describing the model and simulation in section 2, we present the model’s dynamical large-scale response to changes in greenhouse gases in section 3 and the projected changes in the air-mass fractions in section 4, followed by conclusions in section 5.
2. Model simulation and diagnostic tracers
Air-mass fractions for the future climate are calculated using a 20-yr-long time-slice integration of GEOSCCM subject to 2080–2100 time-averaged greenhouse gases and ozone-depleting substances under the SRES A1B and A1 scenarios respectively. As for the present-day integration presented in Part I, which was also integrated for 20 model years, sea surface temperatures and sea ice concentrations are taken from an integration of the NCAR Community Climate System Model, version 3.0, subject to A1B GHG forcing, except that time averages have now been taken over the model years 2080–2100. For more details about the model we refer the reader to section 2 in Part I.
For both the REF and FTR integrations, air-mass origin regions are defined with respect to the model’s PBL, which is first partitioned into three zonally symmetric origin regions: a “southern latitude patch” (ΩSTH) spanning latitudes south of 25°N, a “midlatitude patch” (ΩMID) between 25° and 60°N, and an “Arctic patch” poleward of 60°N (ΩARC; Fig. 1 in Part I). In addition, six nonoverlapping origin regions within ΩMID are defined over the eastern Pacific, North America, the Atlantic, Europe, Asia, and the western Pacific, denoted throughout using the labels EPAC, NAM, ATL, EUR, ASI, and WPAC, respectively.
Following spinup of the dynamical variables, nine tracers, corresponding to the nine Ωi origin regions, are integrated for 20 years for the future (FTR) integration. Once air masses have reached equilibrium, their annual (ANN), wintertime (DJF), and summertime (JJA) climatological mean fractions are calculated over the last 10 years of the integration, and are denoted as
Climate changes in the air-mass fractions are expressed in terms of the differences between the FTR and REF 10-yr averaged climatologies and denoted throughout using the notation Δf(r | Ωi) (where Δ ≡ FTR − REF). Note that throughout we refer to the FTR − REF changes as the responses to “increases in greenhouse gases” since the circulation changes at NH high latitudes incurred only by changes in ozone-depleting substances are relatively weaker. Statistical significance of the diagnosed climate changes in the air-mass fractions and the dynamical variables is assessed based on an independent two-sample Student’s t test using the local standard deviation at each grid point
3. Climate change in large-scale dynamics
During boreal winter the air in the Arctic that last contacts the midlatitude surface originates primarily over the oceans, owing to vigorous isentropic transport associated with the midlatitude storm tracks; by comparison, during summer, when the storm tracks weaken and large-scale convection over land intensifies, ΩMID air originates primarily over land (Part I). We therefore analyze the large-scale dynamical response to changes in greenhouse gases in terms of changes in the DJF and JJA climatological zonal winds, meridional transient eddies, and convective cloud fraction (Figs. 1 and 2). As with the air-mass fractions, differences in the dynamical variables are taken between 10-yr REF and FTR climatologies and statistical significance is assessed at the 90% confidence level.
The zonal-mean changes in the wintertime zonal winds,
By comparison, the response aloft is statistically significant and collocated with an intensification in the transient eddy variance of the meridional velocity
While the zonal-mean response in GEOSCCM indicates that the midlatitude winter circulation will undergo a poleward shift with global warming, an examination of the changes in the DJF 300–900-hPa column-integrated zonal winds reveal strong zonal asymmetries, including a poleward shift over the Atlantic but an equatorward shift over the Pacific (Fig. 1c). Both responses are statistically significant relative to the model’s internal variability, although the significance of response over the eastern Pacific is weaker and confined to narrow regions at around 40°N and around 25°N where the zonal winds weaken and intensify, respectively. This response in the model is consistent with projected changes in the NH midlatitude jet among the CMIP5 models (Barnes and Polvani 2013; Delcambre et al. 2013; Simpson et al. 2014).
The response in the NH summer zonal-mean zonal winds,
Changes in the summertime convective cloud fraction provide a gross sense for how large-scale stability and convective transport over midlatitudes changes in the future climate (preferable to discerning changes in noisier fields, like the vertical velocity ω). The zonal-mean response to changes in greenhouse gases (Fig. 2b) shows a statistically significant decrease in convective cloud fraction throughout the troposphere as large-scale vertical motions over the lower and middle troposphere weaken in concert with an increase in tropospheric static stability (Wetherald and Manabe 1988; Senior and Mitchell 1993; Zelinka et al. 2013). In addition, the positive cloud fraction anomalies that extend down from the tropopause poleward of 60°N reflect a poleward shift in the REF climatology, while positive anomalies at the tropopause reflect a deepening of the troposphere as the high-latitude tropopause rises by approximately 10 hPa, consistent with CMIP5 multimodel projections (Zelinka et al. 2013). Finally, smaller-scale convective cloud fraction changes are not discussed as they are not statistically significant and more likely to hinge on model-specific cloud parameterizations and cloud–radiative feedbacks.
4. Climate change in transport to the Arctic
The air-mass fractions have characteristic seasonal-mean climatological distributions and responses to changes in greenhouse gases that we examine systematically in terms of
During boreal winter and summer, Arctic mid- and upper-tropospheric air originates primarily over the NH midlatitude surface, with
The changes
a. NH winter (DJF)
The large (~7%) positive anomalies in
The change
The changes
Compared to the upper-tropospheric increases in
In a similar sense we have confirmed that the upper-level changes in
b. NH summer (JJA)
In contrast to winter, the summertime response to increases in greenhouse gases,
To interpret the changes
By comparison, the negative anomalies in
Finally, the thin band of large (~7%) and statistically significant anomalies in
5. Changes in PBL origin of Arctic air: Large-scale circulation constraints
To aid in the interpretation of the air-mass-fraction responses to future warming, we now discuss changes in the large-scale circulation over the midlatitudes, including changes in large-scale stationary waves, transient eddy variance, and large-scale convection. Throughout, we focus on circulation changes that GEOSCCM represents with fidelity compared to other comprehensive climate models subject to A1B GHG forcing (i.e., the dynamical changes discussed in section 2).
a. Enhanced oceanic PBL origin during NH winter
Recall from Part I that the ΩMID air-mass fraction in the lower Arctic reflects transport by large-scale stationary waves over NH midlatitudes that control the low-level convergence and poleward transport of recently labeled ΩMID air into the Arctic. More precisely, it is shown that the individual ΩMID air-mass fractions originating over regions of mean cyclonic flow tend to be large over their corresponding origin regions Ωi, since low-level convergence ensures that air is less likely to relabeled elsewhere at the PBL. In addition to modifying the conditions under which air is (re)labeled at the PBL, changes in large-scale stationary waves also affect meridional transport into the Arctic. Thus, when interpreting the changes
To begin, we examine the NH winter sea level pressure (SLP) response to increases in greenhouse gases. Over the Pacific the Aleutian low deepens by approximately 3 hPa (Fig. A2a, left panel), a response that closely resembles the sea level pressure changes that occur during El Niño (Trenberth and Hurrell 1994; Zhang et al. 1997). This response is consistent with CMIP3 and CMIP5 multimodel projections, some of which reveal sea level pressure decreases over the North Pacific in excess of 4 hPa (IPCC 2013). As the Aleutian low deepens, low-level convergence and mean ascent shifts farther over ΩEPAC. Associated with these changes, the near-surface westerlies weaken over the eastern Pacific and obtain a more northward component over Alaska, ensuring that ΩEPAC air in the warmer climate is less likely to be advected westward over North America, where it is relabeled (Fig. A2b, left panel). Correspondingly, large positive anomalies in
In addition to the wind changes that impact the (re)labeling of eastern Pacific air at the PBL, we find that the near-surface wind response over the Pacific is associated with a barotropic cyclonic anomaly that extends throughout the troposphere. This circulation anomaly is manifest as negative anomalies in the 500–900-hPa-integrated eddy geopotential height
Over the Atlantic, by comparison, the sea level pressure response reveals a westward and northward shift in the Icelandic low out of midlatitudes and over Greenland, consistent with CMIP3 and CMIP5 multimodel mean projections, which results in lower pressures over the poles and higher pressures over midlatitudes (i.e., a trend in the northern annular mode toward its high index polarity) (Thompson et al. 2000) (Fig. A2a, right panel). This is associated with stronger westerlies over the North Atlantic, a statistically significant response relative to the model’s internal variability (Fig. A2b, right panel, and Fig. 1c). Correspondingly, stronger westerlies ensure that ΩATL air is less likely to encounter its origin region and more likely to be advected east over Europe, where it is relabeled, resulting in the negative anomalies in
As over the Pacific, the stationary wave response over the Atlantic is not confined to the surface but rather is related to a barotropic anticyclonic anomaly that extends throughout the lower and middle troposphere over the eastern Atlantic and western coast of Europe. This circulation feature appears as positive anomalies in the 300–700-hPa column-integrated eddy geopotential height and projects strongly onto the REF climatology (Fig. 6a). Correspondingly, enhanced longitudinal gradients in
Finally, while changes in large-scale stationary waves over the NH appear to control much the response of the Ωi air-mass fractions in the lower and middle troposphere, in the upper Arctic (i.e., above 500 hPa) interactions with the PBL are weaker and the poleward transport of Ωi air during winter is largely mediated by the transient eddy variance of the meridional velocity (Part I). More precisely, by recasting the passive tracer equation in terms of the residual mean circulation we show in Part I that the eddy transport term
A comparison of
b. Enhanced land PBL origin during NH summer
Air that is labeled over ΩASI and convectively lifted out of the PBL during boreal summer is either transported equatorward to the subtropical upper troposphere via the Asian monsoon or eastward across the Pacific by the mean westerly flow (see Fig. 11, left panel, in Part I). As shown in Part I, strong poleward flow over Siberia ensures that ΩASI air that is convectively transported out of the PBL first enters the Arctic before crossing the Pacific, resulting in the large fractions of ΩASI air that dominate the Arctic middle and upper troposphere. By comparison, ΩNAM air is deflected southward away from the Arctic by mean equatorward flow and is more likely to be relabeled at the PBL, resulting in weaker fractions
In response to increases in greenhouse gases, convection shifts poleward into the equatorward edge of ΩARC (Fig. 2b), with most of this shift occurring over Europe and Asia (not shown). Correspondingly, as convection shifts poleward less ΩASI air is vertically lofted away from the PBL into the upper troposphere, consistent with an overall reduction in
To interpret the positive anomalies in
The changes in
6. Conclusions
There is growing evidence that changes in the long-range transport of midlatitude pollutants have impacted Arctic climate over recent decades (e.g., Hansen and Nazarenko 2004; Lubin and Vogelmann 2006; Shindell et al. 2008). It is therefore natural to ask how long-range transport from midlatitudes to the Arctic will respond to large-scale circulation changes over the twenty-first century. Here, we have assessed how the composition of Arctic air (in terms of its last PBL origin) changes in response to increases in greenhouse gases. Changes in the air-mass fractions reveal the following:
Our model projections indicate that (~10%) more air in the Arctic will originate at the NH midlatitude PBL. The largest increases of midlatitude air during NH winter are concentrated in the upper and middle Arctic, where they reflect an intensification of the transient eddy meridional wind that shifts poleward and upward in response to future increases in greenhouse gases. During summer, by comparison, enhanced fractions of midlatitude air are concentrated below 500 hPa and extend down to the Arctic surface.
Increased fractions of midlatitude air during winter primarily reflect increases in air of eastern Pacific and Atlantic origin, indicating that transport changes alone in the future may lead to “cleaner” Arctic winters (i.e., less air from polluted boundary layers over industrial regions). Future increases in air of eastern Pacific PBL origin reflect anomalous poleward flow along the west coast of North America, a robust dynamical response among comprehensive climate models.
The NH summer air-mass origin response to increases in GHGs is characterized by about 5% increases in air of Asian and North American PBL origin throughout the lower and middle Arctic, indicating that transport changes may enhance Arctic pollution during summer. The enhanced fractions of Asian air are consistent with weaker large-scale convection over NH midlatitudes and stronger poleward flow over Siberia so that less ΩASI air is convectively transported southward into the subtropical upper troposphere and more efficiently steered poleward into the Arctic.
Enhanced poleward transport of midlatitude air may have various impacts on climate by modifying the radiative and chemical properties of the Arctic. Our results indicate that this will depend strongly on season, with higher fractions of oceanic air that are relatively diluted in anthropogenic aerosols dominating the NH winter response; conversely, increases in air originating over Asia, where there are large industrial emissions, could lead to enhanced aerosol loading during summer. To further relate the species-independent transport diagnostics presented here to particulate and gaseous tracers (e.g., black carbon and hydrocarbons) we plan on expanding our analysis using tracers similar to the Ωi air-mass fractions, but also subject to idealized chemical and/or physical loss.
When relating our results to studies of pollution transport, other considerations must be made, including how processes within the PBL may impact chemical constituents. In this study the air-mass fractions, by construction, track air since last PBL contact and air that travels low-level paths into the Arctic become relabeled as Arctic air along the way. However, while our boundary region Ω has been defined for convenience as the modeled PBL, the choice of Ω may be refined to account for species that are sensitive to transport pathways within the PBL. For example, one may be interested in examining the surface origin of short-lived ozone depleting substances residing in the tropical upper troposphere that may be sensitive to low-level cross-equatorial paths associated with seasonal changes in tropical convection, in which case it may be appropriate to use zero-flux boundary conditions over part of Earth’s surface and/or mixing-ratio boundary conditions that are rapidly pulsed in time (Holzer 2009).
When considering future changes in tropospheric chemical composition, it may also be important to keep track of the relative amounts of air that are of tropospheric and stratospheric origin. This would be relevant to ozone, for example, which has both tropospheric and stratospheric sources. While the air-mass fractions defined here trace all air back to its last contact with the PBL, one can readily generalize the setup to also include stratospheric regions of last contact (e.g., Orbe et al. 2013) so that the air-mass fractions sum to unity only when both tropospheric and stratospheric fractions are included. A future decrease in the mass exchange between the PBL and the free troposphere would manifest as a decrease in the PBL air-mass fraction with a corresponding increase in the stratospheric fraction.
Finally, the modeled transport response in GEOSCCM underscores how longitudinal variations in the NH midlatitude circulation response to climate change may lead to significantly different regional responses in meridional transport. In particular, our results show that, while the zonal-mean midlatitude jet is projected to shift poleward in response to increases in greenhouse gases, the dominant transport response—enhanced fractions of eastern Pacific air—is linked to projected changes in large-scale stationary waves in the Pacific where the jet shifts equatorward. This suggests that it may be limiting to consider only shifts in jet latitude and/or strength when interpreting future distributions of tropospheric constituents, and that more research is needed to quantify future changes in large-scale stationary waves over NH midlatitudes. The mechanisms underlying the stationary wave changes described herein, however, are not well understood and will be investigated in future work.
Acknowledgments
This research was supported by an appointment to the NASA Postdoctoral Program at the Goddard Space Flight Center, administered by Oak Ridge Associated Universities through a contract with NASA. The authors also acknowledge support from ARC Grant DP120100674 (M.H.) and NSF Grants AGS-1403676 (D.W.) and AGS-1402931 (M.H., L.M.P.).
APPENDIX
Air-mass Fractions in Reference (REF) Climate
The climatological mean zonally averaged air-mass fractions for the reference (REF) climate (Fig. A1) facilitate the interpretation of the air-mass fraction changes
REFERENCES
ACIA, 2004: Impacts of a Warming Arctic—Arctic Climate Impact Assessment. Arctic Climate Impact Assessment, Cambridge University Press, 144 pp. [Available online at http://www.acia.uaf.edu/.]
Barnes, E. A., and L. Polvani, 2013: Response of the midlatitude jets, and of their variability, to increased greenhouse gases in the CMIP5 models. J. Climate, 26, 7117–7135, doi:10.1175/JCLI-D-12-00536.1.
Comiso, J. C., 2002: A rapidly declining perennial sea ice cover in the Arctic. Geophys. Res. Lett., 29, 1956, doi:10.1029/2002GL015650.
Cook, B., and R. Seager, 2013: The response of the North American monsoon to increased greenhouse gas forcing. J. Geophys. Res. Atmos., 118, 1690–1699, doi:10.1002/jgrd.50111.
Delcambre, S. C., D. J. Lorenz, D. J. Vimont, and J. E. Martin, 2013: Diagnosing Northern Hemisphere jet portrayal in 17 CMIP3 global climate models: Twentieth-century intermodel variability. J. Climate, 26, 4910–4929, doi:10.1175/JCLI-D-12-00337.1.
Frierson, D. M., 2006: Robust increases in midlatitude static stability in simulations of global warming. Geophys. Res. Lett., 33, L24816, doi:10.1029/2006GL027504.
Hansen, J., and L. Nazarenko, 2004: Soot climate forcing via snow and ice albedos. Proc. Natl. Acad. Sci. USA, 101, 423–428, doi:10.1073/pnas.2237157100.
Holland, M. M., C. M. Bitz, and B. Tremblay, 2006: Future abrupt reductions in the summer Arctic sea ice. Geophys. Res. Lett., 33, L23503, doi:10.1029/2006GL028024.
Holzer, M., 2009: The path density of interhemispheric surface-to-surface transport. Part II: Transport through the troposphere and stratosphere diagnosed from NCEP data. J. Atmos. Sci., 66, 2172–2189, doi:10.1175/2009JAS2895.1.
Holzer, M., and G. J. Boer, 2001: Simulated changes in atmospheric transport climate. J. Climate, 14, 4398–4420, doi:10.1175/1520-0442(2001)014<4398:SCIATC>2.0.CO;2.
Hurrell, J. W., 1995: Decadal trends in the North Atlantic Oscillation: Regional temperatures and precipitation. Science, 269, 676–679, doi:10.1126/science.269.5224.676.
IPCC, 2013: Climate Change 2013: The Physical Science Basis. Cambridge University Press, 1535 pp., doi:10.1017/CBO9781107415324.
Koch, D., and J. Hansen, 2005: Distant origins of Arctic black carbon: A Goddard Institute for Space Studies ModelE experiment. J. Geophys. Res., 110, D04204, doi:10.1029/2004JD005296.
Law, K. S., and A. Stohl, 2007: Arctic air pollution: Origins and impacts. Science, 315, 1537–1540, doi:10.1126/science.1137695.
Lu, J., G. A. Vecchi, and T. Reichler, 2007: Expansion of the Hadley cell under global warming. Geophys. Res. Lett., 34, L06805, doi:10.1029/2006GL028443.
Lubin, D., and A. M. Vogelmann, 2006: A climatologically significant aerosol longwave indirect effect in the Arctic. Nature, 439, 453–456, doi:10.1038/nature04449.
Miller, R., G. Schmidt, and D. Shindell, 2006: Forced annular variations in the 20th century Intergovernmental Panel on Climate Change Fourth Assessment Report models. J. Geophys. Res., 111, D18101, doi:10.1029/2005JD006323.
Orbe, C., M. Holzer, L. M. Polvani, and D. Waugh, 2013: Air-mass origin as a diagnostic of tropospheric transport. J. Geophys. Res. Atmos., 118, 1459–1470, doi:10.1002/jgrd.50133.
Orbe, C., P. A. Newman, D. W. Waugh, M. Holzer, L. D. Oman, F. Li, and L. M. Polvani, 2015: Airmass origin in the Arctic. Part I: Seasonality. J. Climate, 28, 4997–5014, doi:10.1175/JCLI-D-14-00720.1.
Rothrock, D. A., Y. Yu, and G. A. Maykut, 1999: Thinning of the Arctic sea-ice cover. Geophys. Res. Lett., 26, 3469–3472, doi:10.1029/1999GL010863.
Senior, C., and J. Mitchell, 1993: Carbon dioxide and climate. The impact of cloud parameterization. J. Climate, 6, 393–418, doi:10.1175/1520-0442(1993)006<0393:CDACTI>2.0.CO;2.
Serreze, M. C., and J. A. Francis, 2006: The Arctic amplification debate. Climatic Change, 76, 241–264, doi:10.1007/s10584-005-9017-y.
Serreze, M. C., and Coauthors, 2003: A record minimum Arctic sea ice extent and area in 2002. Geophys. Res. Lett., 30, 1110, doi:10.1029/2002GL016406.
Shindell, D., G. Faluvegi, A. Lacis, J. Hansen, R. Ruedy, and E. Aguilar, 2006: Role of tropospheric ozone increases in 20th-century climate change. J. Geophys. Res., 111, D08302, doi:10.1029/2005JD006348.
Shindell, D., and Coauthors, 2008: A multi-model assessment of pollution transport to the Arctic. Atmos. Chem. Phys., 8, 5353–5372, doi:10.5194/acp-8-5353-2008.
Simpson, I. R., T. A. Shaw, and R. Seager, 2014: A diagnosis of the seasonally and longitudinally varying midlatitude circulation response to global warming. J. Atmos. Sci., 71, 2489–2515, doi:10.1175/JAS-D-13-0325.1.
Thompson, D. W., J. M. Wallace, and G. C. Hegerl, 2000: Annular modes in the extratropical circulation. Part II: Trends. J. Climate, 13, 1018–1036, doi:10.1175/1520-0442(2000)013<1018:AMITEC>2.0.CO;2.
Trenberth, K. E., and J. W. Hurrell, 1994: Decadal atmosphere–ocean variations in the Pacific. Climate Dyn., 9, 303–319, doi:10.1007/BF00204745.
Wadhams, P., and N. R. Davis, 2000: Further evidence of ice thinning in the Arctic Ocean. Geophys. Res. Lett., 27, 3973–3975, doi:10.1029/2000GL011802.
Wetherald, R., and S. Manabe, 1988: Cloud feedback processes in a general circulation model. J. Atmos. Sci., 45, 1397–1416, doi:10.1175/1520-0469(1988)045<1397:CFPIAG>2.0.CO;2.
Woollings, T., and M. Blackburn, 2012: The North Atlantic jet stream under climate change and its relation to the NAO and EA patterns. J. Climate, 25, 886–902, doi:10.1175/JCLI-D-11-00087.1.
Wu, Y., M. Ting, R. Seager, H.-P. Huang, and M. A. Cane, 2011: Changes in storm tracks and energy transports in a warmer climate simulated by the GFDL CM2.1 model. Climate Dyn., 37, 53–72, doi:10.1007/s00382-010-0776-4.
Yin, J. H., 2005: A consistent poleward shift of the storm tracks in simulations of 21st century climate. Geophys. Res. Lett., 32, L18701, doi:10.1029/2005GL023684.
Zelinka, M. D., S. A. Klein, K. E. Taylor, T. Andrews, M. J. Webb, J. M. Gregory, and P. M. Forster, 2013: Contributions of different cloud types to feedbacks and rapid adjustments in CMIP5. J. Climate, 26, 5007–5027, doi:10.1175/JCLI-D-12-00555.1.
Zhang, Y., J. M. Wallace, and D. S. Battisti, 1997: ENSO-like interdecadal variability: 1900–93. J. Climate, 10, 1004–1020, doi:10.1175/1520-0442(1997)010<1004:ELIV>2.0.CO;2.
Zhou, S., A. J. Miller, J. Wang, and J. K. Angell, 2001: Trends of NAO and AO and their associations with stratospheric processes. Geophys. Res. Lett., 28, 4107–4110, doi:10.1029/2001GL013660.