Changes in Northern Hemisphere Winter Storm Tracks ... - AMS Journals

6 downloads 0 Views 4MB Size Report
May 15, 2017 - vided in the affirmative (e.g., Lorenz and DeWeaver. 2007; Honda et al. 2009; Francis and Vavrus 2012, ... temperature (Ts) (e.g., S. Lee et al.
15 MAY 2017

WANG ET AL.

3705

Changes in Northern Hemisphere Winter Storm Tracks under the Background of Arctic Amplification JIABAO WANG, HYE-MI KIM, AND EDMUND K. M. CHANG School of Marine and Atmospheric Sciences, Stony Brook University, Stony Brook, New York (Manuscript received 30 August 2016, in final form 25 January 2017) ABSTRACT An interdecadal weakening in the North Atlantic storm track (NAST) and a poleward shift of the North Pacific storm track (NPST) are found during October–March for the period 1979–2015. A significant warming of surface air temperature (Ts) over northeastern North America and a La Niña–like change in the North Pacific under the background of Arctic amplification are found to be the contributors to the observed changes in the NAST and the NPST, respectively, via modulation of local baroclinicity. The interdecadal change in baroclinic energy conversion is consistent with changes in storm tracks with an energy loss from eddies to mean flow over the North Atlantic and an energy gain over the North Pacific. The analysis of simulations from the Community Earth System Model Large Ensemble project, although with some biases in storm-track and Ts simulations, supports the observed relationship between the NAST and Ts over northeastern North America, as well as the link between the NPST and El Niño–Southern Oscillation. The near-future projections of Ts and storm tracks are characterized by a warmer planet under the influence of increasing greenhouse gases and a significant weakening of both the NAST and the NPST. The potential role of the NAST in redistributing changes in Ts over the surrounding regions is also examined. The anomalous equatorward moisture flux associated with the weakening trend of the NAST would enhance the warming over its upstream region and hinder the warming over its downstream region via modulation of the downward infrared radiation.

1. Introduction Storm tracks are defined as the preferred regions of extratropical synoptic-scale disturbances (Blackmon 1976). Extreme weather events in the midlatitudes, including extreme precipitation (Kunkel et al. 2012), droughts (Hoerling et al. 2014), coastal flooding (Colle et al. 2008), and extreme cold or heat events (Kocin et al. 1988; Chang et al. 2016) are caused or strongly modulated by storm tracks. Considering significant impacts of storm tracks on global extreme weather and climate, the variability of storm tracks, ranging from hourly to decadal time scale, has long been an intriguing research topic (e.g., Chang et al. 2002). On the subseasonal time scale, the main driver for the variability of Northern Hemisphere winter storm tracks (hereafter ‘‘storm tracks’’) is the Madden– Julian oscillation (MJO; Madden and Julian 1971, 1972, 1994). The eastward movement of anomalous

Corresponding author e-mail: Hye-Mi Kim, hyemi.kim@ stonybrook.edu

tropical convection associated with the MJO across the Indian Ocean and the western Pacific is accompanied by northeastward propagation of the North Pacific storm track (NPST) (Matthews and Kiladis 1999; Deng and Jiang 2011; Lee and Lim 2012; Takahashi and Shirooka 2014). On the seasonal time scale, the North Atlantic and Pacific storm tracks show distinct features: the NPST reaches its minimum in winter and peaks in spring and fall, which is termed the ‘‘midwinter suppression,’’ while the North Atlantic storm track (NAST) attains its maximum in winter (Nakamura 1992; Christoph et al. 1997). On an interannual time scale, the variation of storm tracks could be decomposed into changes in location and intensity (Lau 1988). The north–south shift of the NPST is modulated by El Niño–Southern Oscillation (ENSO) with equatorward and eastward movement during El Niño years, and vice versa during La Niña years (Trenberth and Hurrell 1994; Straus and Shukla 1997). The intensity of the NPST is found to be influenced by the East Asian winter monsoon (Nakamura et al. 2002; Harnik and Chang 2004; Lee

DOI: 10.1175/JCLI-D-16-0650.1 Ó 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

3706

JOURNAL OF CLIMATE

et al. 2010; Song et al. 2016). A weaker NPST is observed during strong East Asian winter monsoon years due to narrowing of the subtropical Pacific jet (Harnik and Chang 2004) as well as a change in local baroclinicity (Lee et al. 2010). The NAST shows extension and meridional shift during different phases of the North Atlantic Oscillation (NAO; Walker 1924). Burkhardt and James (2006) and Chang (2009) found an intensified NAST over the eastern Atlantic during the positive phase of the NAO. Hurrell et al. (2003) and Bader et al. (2011) suggested enhanced (reduced) storm-track activity (STA) to the north (south) of the NAST during the positive phase of the NAO. Among various time scales, the STA on the decadal time scale has received more scientific attention in recent decades because of the growing interest in climate change. While studies have shown that both the NAST and the NPST have had a significant shift of their intensity around the midtwentieth century, the years and intensity of the change differ among the studies (Geng and Sugi 2001, 2003; Graham and Diaz 2001; Chang and Fu 2002; Wang et al. 2013; Chang and Yau 2016). The change of storm tracks in recent decades and its future projection is even more uncertain. Several studies have shown a northward shift of both the NAST and the NPST in recent decades (Wang et al. 2006; Bender et al. 2012) and in the future (Fischer-Bruns et al. 2005; Yin 2005; Bengtsson et al. 2006; O’Gorman 2010; Chang et al. 2012; Lehmann et al. 2014), while others have demonstrated a recent and projected weakening trend in the NAST (Lee et al. 2012; Zappa and Shaffrey et al. 2013; Colle et al. 2015) or a strengthening and extension of southern flank of the NAST with increasing greenhouse gases (Woollings et al. 2012; Harvey et al. 2015). The decadal change of storm tracks, particularly in the North Atlantic, has yet to be fully addressed. Arctic amplification, the warming trend in the Arctic, which is almost twice as large as the global mean temperature, is known to be a significant result of recent climate change. However, whether Arctic amplification has an impact on the midlatitude weather is still disputable. Convincing evidence has been provided in the affirmative (e.g., Lorenz and DeWeaver 2007; Honda et al. 2009; Francis and Vavrus 2012, 2015; Liu et al. 2012; Tang et al. 2013). For example, with the aid of a general circulation model (GCM), Lorenz and DeWeaver (2007) argued that under the background of global warming, the tropopause height rises and is accompanied by a strengthening in the transient eddies with a poleward shift over the globe. On the other hand, several studies argued that an increase in frequency of the negative NAO associated

VOLUME 30

with Arctic amplification could induce an equatorward shift of the NAST (Strong et al. 2009; Overland and Wang 2010; Liu et al. 2012; Woollings and Blackburn 2012; Nakamura et al. 2015). A detailed review regarding the influence of Arctic amplification on the midlatitudes is provided by Cohen et al. (2014). However, some investigators have argued that Arctic amplification may not be a dominant factor controlling the midlatitude weather because a response is not noticeable and is often obscured by the atmospheric internal variability (Screen and Simmonds et al. 2013; Barnes and Screen 2015; Perlwitz et al. 2015). Therefore, any possible role of Arctic amplification in causing changes in the STA is still an open question. This is the focus of our study. The results of previous studies could not reach a consensus on how storm tracks have changed in recent decades or how they will be changed in the future. One reason for discrepancies among investigators is a difference of analysis methods used in previous studies. Some studies were based solely on numerical model experiment (e.g., Fischer-Bruns et al. 2005; Yin 2005; Lehmann et al. 2014); others included reanalysis data but employed different single-statistic representatives of storm tracks (e.g., Wang et al. 2006; Lee et al. 2012; Zappa and Shaffrey et al. 2013). In this study, we will revisit the interdecadal change in storm tracks based on three statistical metrics and verify the possible mechanism with a model simulation. Previous studies also leave us uncertain whether the recent change in storm tracks is related to the Arctic warming. Our hypothesis is that Arctic amplification would result in changes in baroclinicity and hence influence the STA. We will test our hypothesis using a high-resolution reanalysis and simulations from the Community Earth System Model (CESM) Large Ensemble (CESM-LE; Kay et al. 2015) project. In addition, although several studies suggested a possible contribution of midlatitude variation to changes in the Arctic surface air temperature (Ts) (e.g., S. Lee et al. 2011; Yoo et al. 2012; Woods et al. 2013; Park et al. 2015), their analyses were focused primarily on an intraseasonal time scale. In this study, we will examine the potential feedback of changes in the NAST on Arctic amplification on the interdecadal time scale. The paper is organized as follows. Section 2 contains an introduction of datasets and statistical metrics used to define storm tracks. Section 3 describes the interdecadal change of storm tracks using different statistical parameters. The influence of Arctic amplification on the interdecadal change in storm tracks and its plausible mechanism will be presented in section 4, along with the model simulations. The possible feedback of the NAST

15 MAY 2017

3707

WANG ET AL.

on Arctic amplification is discussed in section 5. Section 6 gives a summary and discussion.

2. Data and method a. Data The data used in the present study is the European Centre for Medium-Range Weather Forecasts (ECMWF) interim reanalysis (ERA-Interim) dataset (Dee et al. 2011), which includes daily averaged and monthly quantities from 1979 to 2015. The pressure-level data include horizontal winds, vertical velocity, and air temperature with a horizontal resolution of 2.58 longitude 3 2.58 latitude and 19 vertical levels from 1000 to 100 hPa. The surface quantities are sea level pressure (SLP) and downward infrared radiation (IR). Considering the fact that storm tracks peak from fall to spring, we will focus our analysis on October–March (hereafter referred to as the cool season). To identify the influence of tropical sea surface temperature (SST) on storm tracks, ENSO years are selected based on the Niño-3.4 index using the Extended Reconstructed Sea Surface Temperature, version 4 (ERSST.v4; Huang et al. 2015), provided by the National Oceanic and Atmospheric Administration (NOAA). To test our hypothesis and suggested mechanism, we make use of daily output of SLP and monthly output of Ts from 33 ensemble members derived from the CESM-LE. The CESM-LE contains 40 simulations that are conducted with the CESM, version 1, which uses the Community Atmosphere Model, version 5.2 (CAM5.2), as its atmospheric component, coupled with the ocean, land, and sea ice component models. All historical simulations are subject to the identical external radiative forcing (i.e., greenhouse gas, short-lived gases and aerosols, and ozone) but are initiated by a slightly different atmospheric state, which enables assessment of the atmospheric internal variability. The historical forcing from 1920 to 2005 and the representative concentration pathway 8.5 (RCP8.5) forcing, which corresponds to an anthropogenic forcing of 8.5 W m22 by 2100 from 2006 to 2100, are applied in the CESM-LE. RCP8.5 describes the climate change with the highest greenhouse gas emissions and CO2 concentrations without any climate mitigation (Riahi et al. 2011). A detailed description of the CESM-LE is provided by Kay et al. (2015). Near-future projections of Ts and storm tracks under different climate change scenarios are derived from 13 ensemble simulations (members 3–15) forced by RCP4.5, which corresponds to a medium radiative forcing of 4.5 W m22 by 2100. RCP4.5 derived from the CESM medium ensemble project is compared with the simulations forced by RCP8.5. The comparison of the RCP4.5 and RCP8.5 scenarios not only provides

examination of near-future projections of Ts and storm tracks but also enables us to evaluate changes in storm tracks under the influence of Arctic amplification. To facilitate comparison between the reanalysis data and model simulations, historical runs are selected from 1979 to 2005. For the near-future projection, we analyze the model simulation output from 2016 to 2040.

b. Methodology Different choices of variables used to represent the STA will result in different results. For example, Chang et al. (2012) found an increasing trend for the North Atlantic STA in the period of 1980–2010 when using 300-hPa meridional velocity as a measure of the STA, while a decreasing trend is found when using SLP as a measure of the STA. For cross validation, we examine the STA using three different statistical metrics. The first metric is the 24-h difference-filtered variance to extract the synoptic variability (Wallace et al. 1988; Chang et al. 2012; Chang and Yau 2016), calculated as follows: vv300 5

1 N

å [y(t 1 24 h) 2 y(t)]2 ,

(1)

N

where vv300 is the 24-h difference-filtered variance of meridional velocity y at 300 hPa and N is the number of days within the cool season, which is 181. This 24-h filter highlights the synoptic variation between periods of 1.2–6 days. The second metric is the high-frequencyfiltered variance. A Lanczos filter (Duchon 1979) is applied to the daily SLP to extract the high-frequency signal in 2–8 days with 17 daily weights, and the square of the quantity, denoted as ‘‘pp,’’ is then calculated to represent the intensity of storm tracks (Chang et al. 2012; Chang and Yau 2016). The third metric is the eddy kinetic energy (EKE). A 2–8-day Lanczos filter is applied to daily horizontal winds, and then the EKE is computed on each vertical level using the following expression: 1 EKE 5 (u0 2 1 y 0 2 ) , 2

(2)

where u and y are the zonal and meridional wind, respectively. The vertically integrated EKE between 925 and 250 hPa is derived to describe the STA (Takahashi and Shirooka 2014). The analysis of the STA in this study will mostly rely on the EKE. However, because the interdecadal change in the observed EKE is similar to that in the pp and because the horizontal winds are only available at three vertical levels (850, 500, and 200 hPa) in the model simulation, the assessment of model performance and near-future change of the STA will be based on the pp. To calculate the anomalies, the

3708

JOURNAL OF CLIMATE

VOLUME 30

FIG. 1. The warm–cold differences (shading; 1997/98–2014/15 mean minus 1979/80–96/97 mean) of (a) Ts (interval: 0.5 K) and storm tracks defined by the (b) vv300 (interval: 10 m2 s22), (c) pp (interval: 1 hPa2), and (d) EKE (interval: 1 m2 s22) during the cool season (October–March). Dotted areas indicate differences exceeding the 95% significance level according to the Student’s t test.

climatology of the seasonal cycle of the corresponding field is subtracted from the total field.

3. Interdecadal change in storm tracks To examine the interdecadal change in storm tracks over the past 36 years under the background of Arctic amplification, a difference of the 18-yr mean (warm– cold difference) before and after 1996/97 is compared. As suggested by Chylek et al. (2009) and Walsh (2014), 1996/97 is considered as the transition for the Arctic (608–908N) Ts during the period 1979/80–2014/15 (figure not shown). The period from1979/80 to 1996/97 is defined as a cold period, and the period from 1997/98 to 2014/15 is the warm period. The results of Ts and storm tracks based on the vv300, pp, and EKE are shown in Fig. 1. The warm–cold difference of Ts (Fig. 1a) shows significant Arctic warming with a La Niña–like pattern in the Pacific and a weak cooling trend over northern Asia. The La Niña–like pattern is significant, especially the cooling in the eastern Pacific, which has been documented in previous studies (e.g., Luo et al. 2012; Kosaka and Xie 2013). The cooling trend over northern Asia has been found in previous studies as well (e.g., Overland et al. 2011; Wang and Chen 2014; Horton et al. 2015; Sun et al. 2016). Two maxima in surface warming are discerned, with one located over eastern Canada and western Greenland and the other over the Barents and Kara Seas. A significant poleward shift of the NPST is found, which is consistent with some previous studies (e.g., Wang et al. 2006; Chang et al. 2012) (Figs. 1b–d).

Meanwhile, a striking weakening trend in the NAST is suggested with the center located over the northern North Atlantic (around 608N). In general, three different statistical methods/parameters indicate the same characteristic trend of storm tracks in both the North Pacific and Atlantic, albeit with slight differences in the amplitude and location. Therefore, we conclude that the interdecadal change of the STA is a robust phenomenon. The interdecadal change of the zonal mean storm track and its Pacific and Atlantic components are examined at various vertical levels (Fig. 2). The areas averaged in the Pacific and Atlantic cases are chosen based on the regions of their significant observed changes, found in Fig. 1 (1208E–1208W for the Pacific and 908W–08 for the Atlantic). Overall, a decreased temperature gradient (possibly due to Arctic amplification) is found throughout the troposphere. The interdecadal change in the global zonal mean STA reaches its maximum near 250 hPa and is characterized by an intensification at its southern flank and an attenuation along its northern flank (Fig. 2a). Meanwhile, an upward shift of storm tracks toward the tropopause is also detected, which is consistent with the conclusion of Yin (2005). However, by analyzing interdecadal changes in the North Atlantic and Pacific storms tracks, respectively, it is shown that the observed change in the global zonal mean STA is effectively a superposition of changes in the Atlantic and Pacific storm tracks. The change in the NPST (Fig. 2b) is characterized by a shift both poleward and upward as suggested by previous studies. In the Atlantic

15 MAY 2017

WANG ET AL.

3709

FIG. 2. Latitude–altitude cross section of the warm–cold differences of (a) zonal mean air temperature (shading; interval: 0.2 K) and storm tracks defined by the EKE (red and blue contours; interval: 2 m2 s22) and their counterparts averaged over (b) the Pacific (1208E–1208W) and (c) the Atlantic (908W–08). Gray contours indicate the climatology of zonal mean STA in (a) and the NPST and NAST in (b) and (c), respectively (interval: 50 m2 s22). The zero contour is omitted in all plots.

(Fig. 2c), the Arctic warming signal is strong throughout all vertical layers and extends toward the sub-Arctic region at around 508N. The NAST shows a clear weakening trend with the center located at around 400 hPa, approximately 100 hPa lower than its climatology, which suggests that the attenuation in the NAST is mainly in the midtroposphere when analyzed from the perspective of horizontal winds. Furthermore, although a strengthening is found at the southern flank of the climatological NAST, as suggested by Woollings et al. (2012) and Harvey et al. (2015), the majority of the storm track is characterized by a weakening trend. Therefore, we are inclined to conclude that the NAST has been weakening recently, which is consistent with Lee et al. (2012) and Zappa et al. (2013). In general, the pronounced changes in storm tracks based on the analysis of the EKE are at mid-to-higher altitudes in the vicinity of their climatology. It should be noted that, because changes in the North Atlantic and Pacific storm tracks are distinct from each other, they should be examined separately instead of using simple analysis of the global zonal mean (e.g., Yin 2005). The significant interdecadal changes in the Arctic temperature, NPST, and NAST motivate us to hypothesize that the changes in storm tracks may be a result of recent Arctic amplification. To test our hypothesis, changes in the occurrence frequency of negative phase of the NAST and the poleward shift of the NPST are compared in different

conditions. A negative phase is defined to be days when the normalized daily variation of the NAST (defined by pp) averaged over 208–808N, 908W–108E has negative values, which indicates a weak North Atlantic STA. To identify the dominant areas and characteristics of the NAST and NPST variability, empirical orthogonal function (EOF) analysis is applied to daily EKE anomalies for the period 1979/80–2014/15 during the cool season over 108–808N, 908W–508E for the NAST and 108–808N, 1208E– 1008W for the NPST. Figure 3 shows the three leading eigenvectors for both regions. The first modes of the NAST and the NPST, which are characterized by a monopole pattern over the selected domains, represent changes in storm-track intensity (Figs. 3a,d). Since the significant loading of daily variation of the NAST is over 208–808N, 908W–08, this region will be used to represent daily changes in intensity of the NAST in the later calculation. The second mode shows an east–west (northeast– southwest) shift of the NPST (NAST) (Figs. 3b,e). The third EOF mode of the NPST depicts its north–south shift (Fig. 3c), and thus the principal component of this mode (PC3) will be used later to represent the poleward shift of the NPST, while the third EOF mode of the NAST, which is characterized by a tripole pattern, describes its aggregation or separation (Fig. 3f). In the North Pacific, the relative frequency of poleward-shifted NPST is defined as N/Ntot, where N is the number of days of poleward-shifted NPST selected

3710

JOURNAL OF CLIMATE

VOLUME 30

FIG. 3. Spatial pattern of the first three EOF modes of the daily mean (a)–(c) NPST and (d)–(f) NAST anomalies during the cool season for the period 1979–2015. The percentage shown in each panel’s title represents the explanation variance of each mode.

based on the PC3 of the NPST that has positive values. The value of Ntot is the total number of days during a specific condition. We compare the change in the occurrence frequency of poleward-shifted NPST between the warm and cold periods (Ntot 5 3276 days each). The frequency of the poleward-shifted NPST is increased in the warm Arctic period (Fig. 4a, leftmost bars), consistent with the results shown in Figs. 1 and 2. One might argue that the SST change related to the ENSO has some influence on the variability of the NPST (Trenberth and Hurrell 1994; Straus and Shukla 1997, and others). To examine the relative contribution of the Arctic and tropical temperature change on the NPST, we compare the warm and cold Arctic years during which the ENSO is close to neutral (hereafter ‘‘pure Arctic warming years’’). We select significant warm or

cold Arctic years based on the normalized Ts anomalies averaged over the entire Arctic (608–908N) whose standard deviations are greater than 0.8 or lower than 20.8. Then we remove significant ENSO years, which are defined as years in which the mean Niño-3.4 SST of the cool season is larger than 0.8 K or less than 20.8 K. The cold Arctic years without ENSO influence are 1989/90, 1990/91, 1992/93, and 1993/94 for a total of 728 days; the warm years without ENSO influence are 2005/06, 2006/07, 2011/12, 2013/14, and 2014/15 for a total of 910 days. Results shown in Fig. 4a (center bars) suggest a 4.07% decrease in the relative occurrence frequency of poleward-shifted NPST during pure Arctic warming years. To further examine the combined effect of the Arctic warming and ENSO on changes of the NPST, we

15 MAY 2017

WANG ET AL.

3711

FIG. 4. (a) Relative occurrence frequency (%) of poleward-displaced NPST over the cold and warm periods (leftmost bars), cold and warm years selected based on the entire Arctic warming trend (zonal bands averaged over 608–908N) without influence of strong ENSO events (center bars), and with influence of strong El Niño and La Niña events (rightmost bars). (b) Relative occurrence frequency (%) of negative phases of the NAST over the cold and warm periods (leftmost bars), cold and warm Arctic years selected based on the entire Arctic warming trend (center bars), and on the local Arctic (608–908N, 1608–208W) warming trend (rightmost bars). The blue bars are for cold cases, while the red bars are for warm cases. The numbers on top of the bars represent the value of each bar (%).

compare the warm years that have significant La Niña events (2007/08 and 2010/11) and cold years that have significant El Niño events (1982/83, 1986/87, and 1987/88). The result is given in Fig. 4a (rightmost bars). During the warm Arctic years with strong La Niña events, the frequency of the poleward-shifted NPST is significantly increased (19.05% increase) compared to that in the cold Arctic years with El Niño events. This result is consistent with previous findings that the NPST displays an equatorward shift during El Niño years,

and vice versa during La Niña years (Trenberth and Hurrell 1994; Straus and Shukla 1997). We also compare the warm years that have significant El Niño events (2002/03 and 2009/10) and cold years that have significant La Niña events (1984/85), and a 23.36% decrease of poleward-displaced NPST during the former condition is found (figure not shown). To further investigate the impact of pure ENSO on the change in the NPST, we compare the frequency during the years that have significant ENSO events but no significant

3712

JOURNAL OF CLIMATE

changes in the Arctic temperature. Without the influence of the Arctic warming, there is still an increase in the frequency of the poleward-shifted NPST in La Niña years, although the results are not significant (figure not shown). The above results suggest the larger contribution of La Niña to the observed poleward shift of the NPST over the Arctic warming. While the results confirm the findings from previous studies and provide some new insights about the relative contribution of the Arctic and tropical temperature to the changes of the NPST, there are still uncertainties related to the lack of sufficient sample size for the selected years. To overcome this sampling issue and to confirm our observed findings, modeling experiments will be conducted in the future to clarify this issue. In the North Atlantic, the relative frequency of weakened NAST is defined as N/Ntot, where N is the number of days of negative phase. To examine the contribution of the Arctic warming on the weakened NAST, we compare changes in the NAST frequency among three different measures in selecting period or years. We compare the change in frequency for 1) the warm and cold periods that were defined earlier, 2) the warm and cold years based on the seasonal mean Ts averaged over the entire Arctic (608–908N), and 3) the warm and cold years over the local Arctic region west of the North Atlantic (608–908N, 1608–208W) where a warming center is found. The cold years, based on the Arctic temperature, are 1982/83, 1984/85, 1986/87, 1987/88, 1989/90, 1990/91, 1992/93, and 1993/94 for a total of 1456 days; the corresponding warm years are 2002/03, 2005/06, 2006/07, 2007/08, 2009/10, 2010/11, 2011/12, 2013/14, and 2014/15 for a total of 1638 days. Based on the local Arctic mean temperature anomalies, the cold years are 1982/83, 1983/84, 1986/87, 1988/89, 1989/90, 1990/91, 1991/92, 1992/93, 1993/94, and 1994/95 for a total of 1820 days, while the warm years are 1998/99, 2002/03, 2005/06, 2009/10, and 2010/11 for a total of 910 days. The change in the NAST from the cold–warm period shows an increase in the negative phase (Fig. 4b, leftmost bars), which also implies a decrease in the positive phase. In other words, weaker storm events in the North Atlantic are more frequent in the recent warm period, which is consistent with the conclusion of Sun et al. (2016), who suggested that the risk of extremes in the high latitudes would be reduced as a result of the sea ice loss. By doing the same analysis with the cases of cold and warm Arctic years, the change in frequency is more obvious, especially for the case of local Arctic Ts change (Fig. 4b, center and rightmost bars). Therefore, under the warm Arctic condition, strong storm-track events are less frequent over the Atlantic while the weak storm-track events are more frequent,

VOLUME 30

consistent with the weakening trend in the NAST as we discussed before.

4. The underlying mechanism of the recent changes in storm tracks and their near-future projection In this section, we will investigate the underlying mechanism of the observed storm-track changes and examine the role of Arctic amplification. We will also estimate the near-future projection of storm tracks by using the future scenario from the CESM-LE.

a. Mechanism of the recent changes in storm tracks Under the influence of rapid Arctic warming, the baroclinicity is expected to decrease because of a weakening in the meridional temperature gradient (Frierson 2006; Screen 2014; Sun et al. 2015; Blackport and Kushner 2016). Because the genesis of midlatitude cyclones is largely controlled by baroclinic instability (Eady 1949), changes in the baroclinicity will therefore influence the STA. To test this hypothesis, the climatological baroclinicity for both the North Pacific and the North Atlantic and their changes between the warm and cold periods are displayed in Fig. 5. The baroclinicity is defined as the negative meridional temperature gradient 2dT/dy integrated from the surface to 400 hPa. The climatology of baroclinicity is positive for both regions, indicating canonical negative temperature gradient as expected. During the warm period, a significant decrease in the baroclinicity is found over 408–708N in the North Atlantic, exactly in the same region where the largest weakening in the NAST is observed (Figs. 1b–d). In the North Pacific, there is a tripolar change in the baroclinicity that comprises an increase to the north (408–608N) and a decrease to the south (208–408N) of the climatological NPST. This also matches the observed poleward shift of the NPST (Figs. 1b–d). These results support the hypothesis that the change in the baroclinicity is an important contributor to the changes in both the NPST and the NAST. Moreover, the weakening in the meridional temperature gradient over the NAST region reinforces the possible influence of the Arctic warming on the NAST, while the dipole change in the baroclinicity over the NPST region supports the plausible contribution of La Niña–like SST change to the poleward shift of the NPST over Arctic amplification as suggested by Fig. 4. The variation of STA should be related to the energy conversion between mean flow and eddies (e.g., Chang et al. 2002). To explain the interdecadal change in storm tracks in terms of its energetics, we first analyze barotropic energy conversion (BTEC) between the mean kinetic energy and the eddy kinetic energy using the

15 MAY 2017

WANG ET AL.

3713

FIG. 5. Climatology of longitudinal-mean baroclinicity over the North Atlantic (black line; 908W–08) and North Pacific (gray line; 1208E–1208W), and their corresponding changes (red line with squares for the North Atlantic; blue line with filled squares for the North Pacific) between the warm and cold periods. The baroclinicity is defined as negative temperature gradient integrated from the surface to 400 hPa. The original values of changes are multiplied by 100 while the climatology are maintained as original values.

following formula (Hoskins et al. 1983; Simmons et al. 1983; Lee et al. 2012):    P0 1 0 2 ›u ›y 02 (y 1 u ) 2 BTEC 5 ›x ›y g 2   ›y ›u 1 (2u0 y 0 ) 1 , (3) ›x ›y where g is the acceleration of gravity, P0 is the reference pressure (51000 hPa), prime indicates transients, and the overbar represents asymmetric mean flow. The BTEC is related to the interactions between the deformation of the basic flow and eddy fluxes. The BTEC and zonal wind at 400 hPa, the level where maximum changes in storm tracks are found (Figs. 2b,c, especially for the NAST), are shown in Fig. 6. The distribution of the BTEC in both the cold and warm periods (Figs. 6a,b) is characterized by a BTEC from mean kinetic energy to eddy kinetic energy along the northern flank of the Pacific jet as well as at the entrance of the Atlantic jet (Black and Dole 2000; S.-S. Lee et al. 2011, 2012). The mean flow in those regions is confluent and characterized by cyclonic meridional shear; meanwhile, eddies are tilted upshear across the confluence. Thus, both shearing and stretching induce positive BTEC (Black and Dole 2000). On the other hand, the mean flow obtains kinetic energy from the eddies at the southern flank of the Pacific jet and from the exit region of the Atlantic jet as a result of barotropic decay of the eddies. The northward shift and northeastern extension

FIG. 6. Zonal wind (contour) and BTEC from mean kinetic energy to eddy kinetic energy (shading; W m22) at 400 hPa for the (a) cold period, (b) warm period, and (c) their differences (warm– cold difference). Contour interval in (a) and (b) is 10 m s21, while in (c) is 1.5 m s21. Contour values smaller than 20 m s21 are omitted in (a) and (b), and the zero contour is omitted in (c). Dotted areas in (c) indicate differences of the BTEC exceeding the 95% significance level according to the Student’s t test.

of the Pacific jet, as well as the weakening of the Atlantic jet, are clearly seen in the warm period (Fig. 6c). The variation of the subtropical jet is partly related to the change in the baroclinicity via the thermal wind relationship. Therefore, the northward shift of the Pacific jet is indicative of an enhancement of baroclinicity to the north and attenuation to the south, corresponding to the northward shift of the NPST (Fig. 1); the attenuation in the northern flank of the Atlantic jet is related to a local decrease in the baroclinicity which implies a weakening in the NAST. In this sense, the change in the baroclinicity could also be inferred by changes in the jet location. A closer investigation into the change in the BTEC (Fig. 6c) reflects a negative change in the BTEC to the north of the NPST and a positive change to the south, which should favor a southward shift of the NPST. Moreover, an

3714

JOURNAL OF CLIMATE

VOLUME 30

FIG. 7. As in Fig. 6, but for the vertical integral of BCEC (W m22) from (a)–(c) BCEC1 and (d)-(f) BCEC2 between 900 and 300 hPa, respectively. Contour interval in (a),(b),(d), and (e) is 5 W m22 and contour values smaller than 5 W m22 are omitted. Dotted areas in (c) and (f) indicate differences exceeding the 95% significance level according to the Student’s t test.

increase in the BTEC over the Atlantic should enhance the local STA. Therefore, since changes in the BTEC have opposite tendencies with changes in storm tracks, the observed interdecadal change in storm tracks is likely not the result of the change in the BTEC. The energy for eddy growth is derived primarily from the baroclinic energy conversion (BCEC; Peixoto and Oort 1992), which involves energy conversion from mean available potential energy to eddy available potential energy (BCEC1) and from eddy available potential energy to eddy kinetic energy (BCEC2). The analysis of the BCEC is based on the following formulas (Dole and Black 1990; Cai et al. 2007; Lee et al. 2012):

!

P0 R   P g 0 0 ›T 0 0 ›T uT 1y T BCEC1 5 du ›x ›y dp   Cy / Cp P R 0 0 (v T ) , BCEC2 5 2 0 g P

and (4)

(5)

where u denotes potential temperature, R (5287 J kg21 K21) is the gas constant for dry air, and Cy (5717 J kg21 K21) and Cp (51004 J kg21 K21) are the specific heat of dry

air at constant volume and pressure, respectively. Therefore, BCEC1 is related to horizontal eddy heat flux and mean temperature gradient, while BCEC2 is an indication of vertical eddy heat flux. The vertical integrals of BCEC1 and BCEC2 from 900 to 300 hPa in the cold and warm period and their differences are shown in Fig. 7. Overall, the magnitudes of BCEC1 and BCEC2 are comparable with their maximum upstream of the climatological storm tracks, which further allude to the role of baroclinic instability as an energy source. Their differences between the two periods (Figs. 7c,f) suggest that changes in the BCEC contribute to the interdecadal change in the NAST with weakened BCEC from mean flow to eddy over the North Atlantic during the warm period. Meanwhile, the significant increase in the BCEC over the North Pacific is indicative of its role in the northward shift of the NPST.

b. Model simulation of the interdecadal change in storm tracks and their near-future projections In this subsection, the model performance in simulating the observed interdecadal changes in storm tracks will be investigated by using simulations from the CESM-LE. Furthermore, the near-future projection of

15 MAY 2017

WANG ET AL.

3715

FIG. 8. Differences of Ts (shading; interval: 0.5 K) based on the (a) ERA-Interim and (b) 33-member ensemble mean derived from the CESM-LE between 1997/98–2004/05 and 1979/80–96/97. Climatology of storm tracks (hPa2) based on the (c) ERA-Interim and (d) 33-member ensemble mean for the period 1979/80–2004/05. Contour interval in (c) and (d) is 20 hPa2. Dotted areas in (a) indicate differences exceeding the 95% significance level according to the Student’s t test.

storm tracks under different climate change scenarios (RCP4.5 and RCP8.5) will be examined to determine the role of Arctic amplification. To evaluate the model fidelity in simulating the climatology of both Arctic amplification and storm tracks (defined by the pp), the CESM-LE ensemble mean is compared with the reanalysis for the overlapping period of 1979–2005 (Fig. 8), defining the cold period as 1979/80–96/97 and the warm period as 1997/98–2004/05. Although the warm period extends only to 2005, the characteristics of the Ts change in the reanalysis are almost identical to Fig. 1a, suggesting its robustness, albeit with changes in periods. In contrast to the warming centers in Greenland and the northern North Atlantic in the reanalysis (Fig. 8a), a maximum warming in the model is simulated near the New Siberian Islands (Fig. 8b). Meanwhile, a cooling is simulated over the northern North Atlantic (around 558N, 308W), which is not observed in the reanalysis. The observed cooling trends over the eastern Pacific and northern Asia do not materialize in the simulations. All in all, the CESM could, to some extent, simulate Arctic amplification but with regional biases. On the other hand, the model simulation of the climatological structure of storm tracks (Fig. 8d) bears similar characteristics to that found in the reanalysis (Fig. 8c), although the simulated NPST is weaker. The interdecadal changes in storm tracks are compared with the model and reanalysis (Figs. 9a,b). It is noted that the observed uniform weakening trend in the NAST could hardly be found in the ensemble mean. Instead, the model simulates a meridional dipole change

in the NAST with strengthening in the southern flank and weakening in the northern flank (Fig. 9b). The observed interdecadal northward shift of the NPST is replaced by an eastward shift in the model simulation. Possible reasons for these model biases would be the deficient simulation of the Arctic temperature change and SST change in the North Pacific. Since the model simulates a weaker warming in the Arctic than the reanalysis and a cooling over the North Atlantic that is not observed in reanalysis (Fig. 8b), the uniform weakening in the NAST is accordingly less significant and replaced by a southward shift of the NAST. By the same token, under the influence of the illusive warming center over the New Siberian Islands and the missing La Niña–like change of SST (Fig. 8b), the poleward shift found in the NPST is replaced by a zonally oriented dipole change. In light of this, Arctic amplification may play an important role in the interdecadal change in storm tracks as we hypothesized. In addition, a local change in Ts also has a potential to modulate the change in storm tracks. Hence, realistic simulation of the global temperature is required to improve the simulation of storm tracks. Aside from the influence of Ts, the potential role of atmospheric internal variability (noise) versus external forcing (signal) is also compared (Figs. 9c,d). Noise is defined as the standard deviation of storm-track difference between the warm and cold period across 33 individual ensemble members. The maxima of noise are located over the climatological regions of storm tracks (Fig. 9c), which suggests the importance of the internal variability on the interdecadal STA change. However,

3716

JOURNAL OF CLIMATE

VOLUME 30

FIG. 9. (a),(b) As in Figs. 8a,b, but for storm tracks (hPa2). Dotted areas in (a) indicate differences exceeding the 95% significance level according to the Student’s t test. (c) Noise is defined as the standard deviation of difference across the individual ensemble members, while (d) the signal-to-noise ratio is the ratio of absolute value of (b) divided by (c). The ratio larger than 2 is encircled by black contour.

an examination of the signal (ensemble mean)-to-noise ratio (Fig. 9d) indicates that, although the internal variability has an influence on the NAST, the change in the NAST is still largely controlled by the external forcing, such as the Arctic warming, with most of the North Atlantic having a ratio larger than 1. But the changes in the NPST seem to be dominated by the internal variability with most areas having a ratio smaller than 1. Although the ensemble mean could barely resolve the observed features of Arctic amplification, the La Niña– like change in SST, and the interdecadal changes in storm tracks, some ensemble members may be able to reproduce these signals. To test this, we calculate the spatial correlation coefficient between individual ensemble members and corresponding observations. The areas chosen for the computation of the spatial correlation coefficient are 608–908N, 1008–308W for Ts over northeastern North America; 108–808N, 1208E–1208W for Ts over the North Pacific; 408–808N, 608W–08 for the NAST; and 208–608N, 1208E–1208W for the NPST. The results are shown in Fig. 10. In terms of the changes in Ts over the North Pacific and northeastern North America, ensemble members 22 and 6 have the highest correlations with the observation (0.24 and 0.53, respectively). Ensemble member 22 captures the signal of the La Niña–like change in SST, whereas member 6 simulates well the warming center over northeastern Canada and western Greenland (Figs. 10a,b). When the La Niña– like change in SST is reproduced by the model, a

significant poleward shift of the NPST could materialize, making ensemble member 22 the best simulation of the NPST (correlation coefficient of 0.53; Fig. 10d). This further supports the link between the northward shift of the NPST and the La Niña–like change in SST found in the observational analysis. When an intense warming over northeastern North America is simulated, a significant weakening in the NAST is found (Figs. 10b,e). However, the best simulation still has a cooling bias in the northern North Atlantic Ocean, which causes an enhanced baroclinicity to the south of the cooling and a weakened baroclinicity to the north (figure not shown) with a change opposite to the observation. As a result, the corresponding weakening detected in the NAST shifts northward, and a strengthening is simulated to the southern flank of the NAST. On the other hand, ensemble member 20 is the best representation of the observed weakening trend in the NAST with the spatial correlation coefficient of 0.59 (Fig. 10c). It is interesting to note that, in the corresponding simulation of Ts, the cooling bias found in the northern North Atlantic is weaker and shifts to the far south. As a result of this, the simulated strengthening in the southern flank of the NAST seems to be nearly gone, and a uniform weakening of the NAST is reproduced similar to the observed change. This further implies that the surface temperature bias in the model, especially over the North Atlantic Ocean, may be a factor that worsens the NAST simulation.

15 MAY 2017

WANG ET AL.

3717

FIG. 10. Simulations of changes in (a) Ts over the North Pacific, (b) Ts over the North Atlantic (K), (c) the NAST, and (d) the NPST (hPa2) that have highest spatial correlation coefficients with the observations. (e),(f) Corresponding simulations of the pp and Ts. The regions chosen for the computation of spatial correlation coefficients are given in the text.

Near-future projections of storm tracks and Ts under different emission scenarios are investigated by comparing the projected changes between RCP4.5 and RCP8.5 (Figs. 11a and c, respectively). A larger increase in the Arctic temperature is expected in RCP8.5. Meanwhile, a weakening of storm tracks is suggested in both scenarios (Figs. 11b,d). It is interesting to note that when a warmer temperature over the western Arctic is projected in RCP8.5 compared with RCP4.5, a broader area of the weakening of extratropical storms is found in nearby regions. Meanwhile, similar warming magnitude between RCP4.5 and RCP8.5 over eastern Canada may be a reason for similar projected change in the NAST. The above features reinforce our conjecture that Arctic amplification has a remarkable influence on the change in storm tracks. However, it needs to be mentioned here that some previous investigators have suggested an increase just west of the United Kingdom based on the multimodel mean of models from phase 5 of the Coupled Model Intercomparison Project (CMIP5; Chang et al. 2012; Harvey et al. 2012). This is found in our RCP4.5 projection, but it is absent in the RCP8.5 runs.

Therefore, although the near-future projection of storm tracks based on the CESM-LE supports the relationship between the Arctic temperature and storm tracks well, the plausible bias of the CESM simulation still needs to be considered.

5. The potential feedback of the NAST on Arctic amplification In the previous sections, the potential role of Arctic amplification on the weakening of the NAST has been examined using the reanalysis data and model simulation. In this section, we will discuss the possible feedback of the weakened NAST on Arctic amplification. First, the principal component of the leading EOF mode (PC1) of the NAST (Fig. 3d) is used to represent the daily variation of the NAST. Then, the temporal correlation coefficient between the PC1 and daily Ts is computed (Fig. 12). A significant negative (positive) correlation is found between daily changes in the NAST intensity and Ts over North America and its adjacent ocean (northern Europe and the Barents and

3718

JOURNAL OF CLIMATE

VOLUME 30

FIG. 11. Near-future projections of (a),(c) Ts (shading; interval: 1 K) and (b),(d) storm tracks (shading; interval: 0.5 hPa2) under the RCP4.5 and RCP8.5 scenarios, which is derived as differences between the 15-member ensemble mean of RCP4.5 and RCP8.5 for the period of 2016/17–39/40 and that of the historical runs for 1997/98– 2004/05.

Kara Seas). This indicates that, on a daily time scale, weak NAST is likely related to a dipole pattern of the Ts change with warming over the Arctic adjacent to North America and cooling over northern Europe. Since midlatitude storms mainly influence weather and climate via moisture and heat transport (e.g., Simmonds and Keay 2009; Screen and Simmonds 2010; Zhang et al. 2013), we will investigate first the characteristics of seasonal mean moisture transport in the warm and cold Arctic years defined in section 3. During the warm years, significant enhancement in poleward moisture transport is seen over the Atlantic; while during the cold years, more zonal moisture transport is seen (Fig. 13). To distinguish the relative contributions from low- (30–90 day), intermediate- (9–29 day), and high-frequency (2–8 day) components to the enhancement of the seasonal mean poleward moisture transport during the warm years, a decomposition of anomalous moisture flux is conducted. We define a quantity A as A 5 A 1 AL 1 AI 1 AH , where the overbar indicates climatology and superscripts L, I, and H represent low, intermediate, and high frequency, respectively. A Lanczos filter is applied to anomalies to extract their low-, intermediate-, and high-frequency components. Then, the anomalous moisture flux integrated from the surface to 300 hPa is decomposed as follows: Vq0 1 V 0 q 1 V 0 q0 ffi (VqH 1 V H q 1 V H qH ) 1 (VqI 1 V I q 1 V I qI ) 1 (VqL 1 V L q 1 V L qL ) ,

where V is the horizontal wind and q the specific humidity. The results of decomposition are shown in Fig. 14. The structure of Fig. 14a is roughly similar to Fig. 13a with remarkably enhanced poleward moisture transport in the North Atlantic during warm years. This indicates the applicability of the abovementioned decomposition. By comparing the relative contribution of each component (Figs. 14b–d), an anomalous equatorward moisture transport is found in the North Atlantic for the high-frequency component, but an intensified poleward transport is detected in the low- and intermediate-frequency components. In this sense, a weakening in storm activities (high-frequency component) during the warm Arctic years is suggested because the climatological transport of the moisture and heat by the NAST (high-frequency component) is directed eastward and poleward, which is consistent with results discussed in previous sections.

FIG. 12. Correlation coefficient map between the PC of the leading EOF mode of the daily EKE anomalies over 108–808N, 908W–508E (region of the NAST) for the period 1979/80–2014/15 and Ts (shading; interval: 0.02). Dotted areas denote correlation exceeding the 95% significance level.

15 MAY 2017

WANG ET AL.

3719

FIG. 13. The vertical integral of moisture flux from the surface to 300 hPa (vector; 10 kg m21 s21; only values exceeding the 95% significance level are shown) for the (a) warm and (b) cold years selected based on the entire Arctic warming trend. Shading represents meridional transport (10 kg m21 s21).

To examine the potential feedback of the weakened North Atlantic storm activities on Arctic amplification, a composite analysis is conducted on daily moisture flux, heat flux, downward IR, and Ts anomalies associated with intense negative phases of the NAST (events selected based on the PC1 of the NAST with values less than standard deviation of 21) (Fig. 15). During the

negative phases of the NAST (weakened storm activities over the North Atlantic), equatorward moisture and heat transport anomalies are found over the North Atlantic (Figs. 15a,b). Since the orientation of the anomalous moisture and heat transport is from northern Europe toward the Arctic near North America, which is opposite in direction to the climatological transport

FIG. 14. Composite maps of (a) Vq0 1 V 0 q 1 V 0 q0 , (b) VqH 1 V H q 1 V H qH , (c) VqI 1 V I q 1 V I qI , and (d) VqL 1 V L q 1 V L qL (vector; 10 kg m21 s21; only values exceeding the 95% significance level are shown) for the warm Arctic years. Shading indicates the meridional transport (10 kg m21 s21).

3720

JOURNAL OF CLIMATE

VOLUME 30

FIG. 15. Composite maps of the anomalous (a) moisture flux (vector; 10 kg m21 s21), (b) heat flux (vector; 10 kg K m21 s21), (c) downward infrared radiation (shading ;interval: 3 W m22), and (d) Ts (shading; interval: 0.2 K) associated with the negative phase of the intense NAST. For the moisture and heat fluxes, only fluxes exceeding the 95% significance level according to the Student’s t test are shown. Shading in (a) and (b) denotes the meridional transport (10 kg m21 s21), whereas contours indicate anomalous convergence (green) and divergence (brown) of integrated moisture and heat flux.

(figure not shown), it takes less humidity and heat away from the latter region (the Arctic near North America) and hence brings less heat and humidity to the former region (northern Europe). As a result, an increase in humidity and heat over northeastern North America is expected. The expected change in humidity could be seen in the composite of anomalous convergence of moisture to some extent (contours in Fig. 15a). The convergence of moisture flux is located over Baffin Island and the Labrador Peninsula, while divergence is found to the west of Scandinavian Peninsula. Anomalous moisture convergence is expected to lead to an increase in humidity, which would enhance cloud liquid water. Consistent with this, an increase in the downward IR is found over northern North America (Fig. 15c; Chen et al. 2006; Park et al. 2015). In light of the fact that an increase in the downward IR has a large contribution to the Arctic warming especially during the cool season (Fig. 15d; Graversen 2006; S. Lee et al. 2011; Yoo et al. 2012; Woods et al. 2013), the weakened NAST can provide a positive feedback to Arctic amplification through the anomalous moisture transport. However, the anomalous convergence of heat flux (contours in

Fig. 15b) is inconsistent with the pattern of the Ts change (Fig. 15d), suggesting a negligible role of heat transport by North Atlantic storm activities to the warming over the Arctic near North America. To explain the feedback more comprehensively, a detailed process study including a consideration of the variability at different time scales (from high frequency to seasonal mean) is warranted.

6. Summary and discussion a. Summary Northern Hemisphere storm tracks experienced a salient interdecadal change (i.e., a weakening in the NAST and a northward shift of the NPST). The interdecadal weakening of the NAST in recent decades is shown to be a result of the decreased baroclinicity associated with recent Arctic amplification, while the poleward shift of the NPST is found to be influenced by the La Niña–like change in SST. The investigation of the barotropic and baroclinic energy conversion demonstrates the dominant role of changes in the BCEC on the

15 MAY 2017

3721

WANG ET AL.

interdecadal change of storm tracks. A CESM-LE simulation captures the general features of Arctic amplification and storm tracks, but with some regional biases. The near-future projections of Ts and storm tracks provided by the CESM-LE suggest that Arctic amplification will still be dominant in the future and storm tracks will continue to weaken, which alludes to fewer storm events over the midlatitudes. Fewer storm events in the North Atlantic will further strengthen the warming over the Arctic near North America by increasing the downward IR as a positive feedback but will hinder the warming over northern Europe.

b. Discussion This study points out the role of the decreased baroclinicity under the background of Arctic amplification in the recent weakening trend in the NAST. Although our current study provides new findings, the results are limited to a relatively short period (1979–2015). Because some previous studies showed an increasing trend in the North Atlantic STA during a longer period (e.g., Chang and Yau 2016), which is opposite to the observed recent downward trend, we wanted to test whether the relationship between the Arctic temperature and the NAST on the interdecadal time scale discussed in this study will still hold when including the period of 1958–79. Comparing the differences of storm tracks and Ts between the 1979/80–96/97 mean and the 1958/59–78/79 mean using ERA-40 (figure not shown), an upward trend of the NAST is found during 1958–97 (consistent with Chang and Yau 2016). Adjacent to the intensification of the NAST, a downward trend in Ts near the North Atlantic is shown, which could result in an increase in the local baroclinicity and hence lead to the observed upward trend of the NAST. Moreover, the correlation coefficient between anomalous STA in the North Atlantic and Ts over northeastern North America reaches up to 0.406 for the period of 1958–2002, exceeding the 99% significance level. Therefore, the relationship between the Arctic temperature and NAST is valid even over a longer period. This study discussed the possible impact of Arctic amplification on the changes in the NAST. However, other modes of climate variability, such as the NAO, can also influence the STA over the North Atlantic. For instance, the south–north displacement of the NAST is found to be correlated with the NAO, with a poleward shift in the positive phase of the NAO, and vice versa in the negative phase (Lau 1988; Hurrell et al. 2003; Bader et al. 2011). Because the negative phase of the NAO is more frequently observed recently (Strong et al. 2009; Overland and Wang 2010; Liu et al. 2012; Woollings and Blackburn 2012; Nakamura et al. 2015), investigating

the role of the NAO in the interdecadal change in the NAST will be of great significance. Thus, in addition to Arctic amplification, the role of various low-frequency climate variabilities and their combined impact on the storm-track change needs to be considered. This study suggests a decreasing trend in frequency and/or intensity of strong storm events due to global warming (Figs. 4b and 11). However, Geng and Sugi (2003) found that, although weak- and medium-strength extratropical cyclones would decrease in the future, intense cyclones would increase, which is consistent with Min et al. (2011) and Westra et al. (2013). On the contrary, Catto et al. (2011) and Chang et al. (2012) suggested that the number and intensity of extratropical cyclone will decrease with global warming, which is consistent with our study. The research on the link between the Arctic and midlatitudes is at an early stage. Our study suggests, however, that improved simulation of the Arctic climate is necessary for better understanding of the physical link between the Arctic and midlatitude weather. Given that significant biases exist in the current models and that the variability of the midlatitude climate system is highly modulated by the internal variability (e.g., Zappa et al. 2014), improvement in the climate models and large multimodel ensembles will be needed for significant progress to be made in this area. Acknowledgments. Constructive and valuable comments from three anonymous reviewers are greatly appreciated. We thank the CESM-LE Community Project for providing the model data analyzed in this study. This study is supported by the KMA R&D Program under Grant KMIPA 2016-6010. EC is supported by NSF Grant AGS1261311. REFERENCES Bader, J., M. D. Mesquita, K. I. Hodges, N. Keenlyside, S. Østerhus, and M. Miles, 2011: A review on Northern Hemisphere sea-ice, storminess and the North Atlantic Oscillation: Observations and projected changes. Atmos. Res., 101, 809–834, doi:10.1016/j.atmosres.2011.04.007. Barnes, E. A., and J. A. Screen, 2015: The impact of Arctic warming on the midlatitude jet-stream: Can it? Has it? Will it? Wiley Interdiscip. Rev.: Climate Change, 6, 277–286, doi:10.1002/wcc.337. Bender, F. A., V. Ramanathan, and G. Tselioudis, 2012: Changes in extratropical storm track cloudiness 1983–2008: Observational support for a poleward shift. Climate Dyn., 38, 2037–2053, doi:10.1007/s00382-011-1065-6. Bengtsson, L., K. I. Hodges, and E. Roeckner, 2006: Storm tracks and climate change. J. Climate, 19, 3518–3543, doi:10.1175/ JCLI3815.1. Black, R. X., and R. M. Dole, 2000: Storm tracks and barotropic deformation in climate models. J. Climate, 13, 2712–2728, doi:10.1175/1520-0442(2000)013,2712:STABDI.2.0.CO;2.

3722

JOURNAL OF CLIMATE

Blackmon, M. L., 1976: A climatological spectral study of the 500 mb geopotential height of the Northern Hemisphere. J. Atmos. Sci., 33, 1607–1623, doi:10.1175/1520-0469(1976)033,1607: ACSSOT.2.0.CO;2. Blackport, R., and P. J. Kushner, 2016: The transient and equilibrium climate response to rapid summertime sea ice loss in CCSM4. J. Climate, 29, 401–417, doi:10.1175/ JCLI-D-15-0284.1. Burkhardt, U., and I. N. James, 2006: The effect of Doppler correction on measures of storm track intensity. Climate Dyn., 27, 515–530, doi:10.1007/s00382-006-0146-4. Cai, M., S. Yang, H. Van den Dool, and V. Kousky, 2007: Dynamical implications of the orientation of atmospheric eddies: A local energetics perspective. Tellus, 59A, 127–140, doi:10.1111/j.1600-0870.2006.00213.x. Catto, J. L., L. C. Shaffrey, and K. I. Hodges, 2011: Northern Hemisphere extratropical cyclones in a warming climate in the HiGEM high-resolution climate model. J. Climate, 24, 5336– 5352, doi:10.1175/2011JCLI4181.1. Chang, E. K. M., 2009: Are band-pass variance statistics useful measures of storm track activity? Re-examining storm track variability associated with the NAO using multiple storm track measures. Climate Dyn., 33, 277–296, doi:10.1007/ s00382-009-0532-9. ——, and Y. Fu, 2002: Interdecadal variations in Northern Hemisphere winter storm track intensity. J. Climate, 15, 642–658, doi:10.1175/1520-0442(2002)015,0642:IVINHW.2.0.CO;2. ——, and A. M. W. Yau, 2016: Northern Hemisphere winter storm track trends since 1959 derived from multiple reanalysis datasets. Climate Dyn., 47, 1435–1454, doi:10.1007/ s00382-015-2911-8. ——, S. Lee, and K. L. Swanson, 2002: Storm track dynamics. J. Climate, 15, 2163–2183, doi:10.1175/1520-0442(2002)015,02163: STD.2.0.CO;2. ——, Y. Guo, and X. Xia, 2012: CMIP5 multimodel ensemble projection of storm track change under global warming. J. Geophys. Res., 117, D23118, doi:10.1029/2012JD018578. ——, ——, C. G. Ma, C. Zheng, and A. M. W. Yau, 2016: Observed and projected decrease in Northern Hemisphere extratropical cyclone activity in summer and its impacts on maximum temperature. Geophys. Res. Lett., 43, 2200–2208, doi:10.1002/ 2016GL068172. Chen, Y., F. Aires, J. A. Francis, and J. R. Miller, 2006: Observed relationships between Arctic longwave cloud forcing and cloud parameters using a neural network. J. Climate, 19, 4087– 4104, doi:10.1175/JCLI3839.1. Christoph, M., U. Ulbrich, and P. Speth, 1997: Midwinter suppression of Northern Hemisphere storm track activity in the real atmosphere and in GCM experiments. J. Atmos. Sci., 54, 1589–1599, doi:10.1175/1520-0469(1997)054,1589: MSONHS.2.0.CO;2. Chylek, P., C. K. Folland, G. Lesins, M. K. Dubey, and M. Wang, 2009: Arctic air temperature change amplification and the Atlantic multidecadal oscillation. Geophys. Res. Lett., 36, L14801, doi:10.1029/2009GL038777. Cohen, J., and Coauthors, 2014: Recent Arctic amplification and extreme mid-latitude weather. Nat. Geosci., 7, 627–637, doi:10.1038/ngeo2234. Colle, B. A., and Coauthors, 2008: New York City’s vulnerability to coastal flooding. Bull. Amer. Meteor. Soc., 89, 829–841, doi:10.1175/2007BAMS2401.1. ——, J. F. Booth, and E. K. M. Chang, 2015: A review of historical and future changes of extratropical cyclones and associated

VOLUME 30

impacts along the US East Coast. Curr. Climate Change Rep., 1, 125–143, doi:10.1007/s40641-015-0013-7. Dee, D., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553–597, doi:10.1002/qj.828. Deng, Y., and T. Jiang, 2011: Intraseasonal modulation of the North Pacific storm track by tropical convection in boreal winter. J. Climate, 24, 1122–1137, doi:10.1175/ 2010JCLI3676.1. Dole, R. M., and R. X. Black, 1990: Life cycles of persistent anomalies. Part II: The development of persistent negative height anomalies over the North Pacific Ocean. Mon. Wea. Rev., 118, 824–846, doi:10.1175/1520-0493(1990)118,0824: LCOPAP.2.0.CO;2. Duchon, C. E., 1979: Lanczos filtering in one and two dimensions. J. Appl. Meteor., 18, 1016–1022, doi:10.1175/ 1520-0450(1979)018,1016:LFIOAT.2.0.CO;2. Eady, E. T., 1949: Long waves and cyclone waves. Tellus, 1, 33–52, doi:10.3402/tellusa.v1i3.8507. Fischer-Bruns, I., H. von Storch, J. González-Rouco, and E. Zorita, 2005: Modelling the variability of midlatitude storm activity on decadal to century time scales. Climate Dyn., 25, 461–476, doi:10.1007/s00382-005-0036-1. Francis, J. A., and S. J. Vavrus, 2012: Evidence linking Arctic amplification to extreme weather in mid-latitudes. Geophys. Res. Lett., 39, L06801, doi:10.1029/2012GL051000. ——, and ——, 2015: Evidence for a wavier jet stream in response to rapid Arctic warming. Environ. Res. Lett., 10, 014005, doi:10.1088/1748-9326/10/1/014005. Frierson, D. M., 2006: Robust increases in midlatitude static stability in simulations of global warming. Geophys. Res. Lett., 33, L24816, doi:10.1029/2006GL027504. Geng, Q., and M. Sugi, 2001: Variability of the North Atlantic cyclone activity in winter analyzed from NCEP–NCAR reanalysis data. J. Climate, 14, 3863–3873, doi:10.1175/1520-0442(2001)014,3863: VOTNAC.2.0.CO;2. ——, and ——, 2003: Possible change of extratropical cyclone activity due to enhanced greenhouse gases and sulfate aerosols—study with a high-resolution AGCM. J. Climate, 16, 2262–2274, doi:10.1175/1520-0442(2003)16,2262: PCOECA.2.0.CO;2. Graham, N. E., and H. F. Diaz, 2001: Evidence for intensification of North Pacific winter cyclones since 1948. Bull. Amer. Meteor. Soc., 82, 1869–1893, doi:10.1175/1520-0477(2001)082,1869: EFIONP.2.3.CO;2. Graversen, R. G., 2006: Do changes in the midlatitude circulation have any impact on the Arctic surface air temperature trend? J. Climate, 19, 5422–5438, doi:10.1175/JCLI3906.1. Harnik, N., and E. K. M. Chang, 2004: The effects of variations in jet width on the growth of baroclinic waves: Implications for midwinter Pacific storm track variability. J. Atmos. Sci., 61, 23– 40, doi:10.1175/1520-0469(2004)061,0023:TEOVIJ.2.0.CO;2. Harvey, B. J., L. C. Shaffrey, T. J. Woollings, G. Zappa, and K. I. Hodges, 2012: How large are projected 21st century storm track changes? Geophys. Res. Lett., 39, L18707, doi:10.1029/ 2012GL052873. ——, ——, and ——, 2015: Deconstructing the climate change response of the Northern Hemisphere wintertime storm tracks. Climate Dyn., 45, 2847–2860, doi:10.1007/ s00382-015-2510-8. Hoerling, M., and Coauthors, 2014: Causes and predictability of the 2012 Great Plains drought. Bull. Amer. Meteor. Soc., 95, 269– 282, doi:10.1175/BAMS-D-13-00055.1.

15 MAY 2017

WANG ET AL.

Honda, M., J. Inoue, and S. Yamane, 2009: Influence of low Arctic sea-ice minima on anomalously cold Eurasian winters. Geophys. Res. Lett., 36, L08707, doi:10.1029/ 2008GL037079. Horton, D. E., N. C. Johnson, D. Singh, D. L. Swain, B. Rajaratnam, and N. S. Diffenbaugh, 2015: Contribution of changes in atmospheric circulation patterns to extreme temperature trends. Nature, 522, 465–469, doi:10.1038/ nature14550. Hoskins, B. J., I. N. James, and G. H. White, 1983: The shape, propagation and mean-flow interaction of large-scale weather systems. J. Atmos. Sci., 40, 1595–1612, doi:10.1175/ 1520-0469(1983)040,1595:TSPAMF.2.0.CO;2. Huang, B., and Coauthors, 2015: Extended Reconstructed Sea Surface Temperature version 4 (ERSST.v4). Part I: Upgrades and intercomparisons. J. Climate, 28, 911–930, doi:10.1175/ JCLI-D-14-00006.1. Hurrell, J. W., Y. Kushnir, G. Ottersen, and M. Visbeck, 2003: An overview of the North Atlantic Oscillation. The North Atlantic Oscillation: Climatic Significance and Environmental Impact, Geophys. Monogr., Vol. 134, Amer. Geophys. Union, 1–35. Kay, J., and Coauthors, 2015: 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, doi:10.1175/BAMS-D-13-00255.1. Kocin, P. J., A. D. Weiss, and J. J. Wagner, 1988: The great Arctic outbreak and east coast blizzard of February 1899. Wea. Forecasting, 3, 305–318, doi:10.1175/1520-0434(1988)003,0305: TGAOAE.2.0.CO;2. Kosaka, Y., and S.-P. Xie, 2013: Recent global-warming hiatus tied to equatorial Pacific surface cooling. Nature, 501, 403–407, doi:10.1038/nature12534. Kunkel, K. E., D. R. Easterling, D. A. Kristovich, B. Gleason, L. Stoecker, and R. Smith, 2012: Meteorological causes of the secular variations in observed extreme precipitation events for the conterminous United States. J. Hydrometeor., 13, 1131– 1141, doi:10.1175/JHM-D-11-0108.1. Lau, N.-C., 1988: Variability of the observed midlatitude storm tracks in relation to low-frequency changes in the circulation pattern. J. Atmos. Sci., 45, 2718–2743, doi:10.1175/ 1520-0469(1988)045,2718:VOTOMS.2.0.CO;2. Lee, S., T. Gong, N. Johnson, S. B. Feldstein, and D. Pollard, 2011: On the possible link between tropical convection and the Northern Hemisphere Arctic surface air temperature change between 1958 and 2001. J. Climate, 24, 4350–4367, doi:10.1175/ 2011JCLI4003.1. Lee, S.-S., J.-Y. Lee, B. Wang, F.-F. Jin, W.-J. Lee, and K.-J. Ha, 2011: A comparison of climatological subseasonal variations in the wintertime storm track activity between the North Pacific and Atlantic: Local energetics and moisture effect. Climate Dyn., 37, 2455–2469, doi:10.1007/s00382-011-1027-z. ——, ——, ——, K.-J. Ha, K.-Y. Heo, F.-F. Jin, D. M. Straus, and J. Shukla, 2012: Interdecadal changes in the storm track activity over the North Pacific and North Atlantic. Climate Dyn., 39, 313–327, doi:10.1007/s00382-011-1188-9. Lee, Y. Y., and G. H. Lim, 2012: Dependency of the North Pacific winter storm tracks on the zonal distribution of MJO convection. J. Geophys. Res., 117, D14101, doi:10.1029/ 2011JD016417. ——, ——, and J. S. Kug, 2010: Influence of the East Asian winter monsoon on the storm track activity over the North Pacific. J. Geophys. Res., 115, D09102, doi:10.1029/2009JD012813.

3723

Lehmann, J., D. Coumou, K. Frieler, A. V. Eliseev, and A. Levermann, 2014: Future changes in extratropical storm tracks and baroclinicity under climate change. Environ. Res. Lett., 9, 084002, doi:10.1088/1748-9326/9/8/084002. Liu, J., J. A. Curry, H. Wang, M. Song, and R. M. Horton, 2012: Impact of declining Arctic sea ice on winter snowfall. Proc. Natl. Acad. Sci. USA, 109, 4074–4079, doi:10.1073/ pnas.1114910109. Lorenz, D. J., and E. T. DeWeaver, 2007: Tropopause height and zonal wind response to global warming in the IPCC scenario integrations. J. Geophys. Res., 112, D10119, doi:10.1029/ 2006JD008087. Luo, J.-J., W. Sasaki, and Y. Masumoto, 2012: Indian Ocean warming modulates Pacific climate change. Proc. Natl. Acad. Sci. USA, 109, 18 701–18 706, doi:10.1073/pnas.1210239109. Madden, R. A., and P. R. Julian, 1971: Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28, 702–708, doi:10.1175/1520-0469(1971)028,0702: DOADOI.2.0.CO;2. ——, and ——, 1972: Description of global-scale circulation cells in the tropics with a 40–50 day period. J. Atmos. Sci., 29, 1109–1123, doi:10.1175/1520-0469(1972)029,1109:DOGSCC.2.0.CO;2. ——, and ——, 1994: Observations of the 40–50-day tropical oscillation—a review. Mon. Wea. Rev., 122, 814–837, doi:10.1175/1520-0493(1994)122,0814:OOTDTO.2.0.CO;2. Matthews, A. J., and G. N. Kiladis, 1999: The tropical–extratropical interaction between high-frequency transients and the Madden–Julian oscillation. Mon. Wea. Rev., 127, 661–677, doi:10.1175/1520-0493(1999)127,0661:TTEIBH.2.0.CO;2. Min, S.-K., X. Zhang, F. W. Zwiers, and G. C. Hegerl, 2011: Human contribution to more-intense precipitation extremes. Nature, 470, 378–381, doi:10.1038/nature09763. Nakamura, H., 1992: Midwinter suppression of baroclinic wave activity in the Pacific. J. Atmos. Sci., 49, 1629–1642, doi:10.1175/ 1520-0469(1992)049,1629:MSOBWA.2.0.CO;2. ——, T. Izumi, and T. Sampe, 2002: Interannual and decadal modulations recently observed in the Pacific storm track activity and East Asian winter monsoon. J. Climate, 15, 1855–1874, doi:10.1175/1520-0442(2002)015,1855:IADMRO.2.0.CO;2. Nakamura, T., K. Yamazaki, K. Iwamoto, M. Honda, Y. Miyoshi, Y. Ogawa, and J. Ukita, 2015: A negative phase shift of the winter AO/NAO due to the recent Arctic sea-ice reduction in late autumn. J. Geophys. Res. Atmos., 120, 3209–3227, doi:10.1002/2014JD022848. O’Gorman, P. A., 2010: Understanding the varied response of the extratropical storm tracks to climate change. Proc. Natl. Acad. Sci. USA, 107, 19 176–19 180, doi:10.1073/ pnas.1011547107. Overland, J. E., and M. Wang, 2010: Large-scale atmospheric circulation changes are associated with the recent loss of Arctic sea ice. Tellus, 62A, 1–9, doi:10.1111/j.1600-0870.2009.00421.x. ——, K. R. Wood, and M. Wang, 2011: Warm Arctic—cold continents: Climate impacts of the newly open Arctic Sea. Polar Res., 30, 15787, doi:10.3402/polar.v30i0.15787. Park, H.-S., S. Lee, S.-W. Son, S. B. Feldstein, and Y. Kosaka, 2015: The impact of poleward moisture and sensible heat flux on Arctic winter sea ice variability. J. Climate, 28, 5030–5040, doi:10.1175/JCLI-D-15-0074.1. Peixoto, J. P., and A. H. Oort, 1992: Physics of Climate. American Institute of Physics, 520 pp. Perlwitz, J., M. Hoerling, and R. Dole, 2015: Arctic tropospheric warming: Causes and linkages to lower latitudes. J. Climate, 28, 2154–2167, doi:10.1175/JCLI-D-14-00095.1.

3724

JOURNAL OF CLIMATE

Riahi, K., and Coauthors, 2011: RCP 8.5—a scenario of comparatively high greenhouse gas emissions. Climatic Change, 109, 33–57, doi:10.1007/s10584-011-0149-y. Screen, J. A., 2014: Arctic amplification decreases temperature variance in northern mid- to high-latitudes. Nat. Climate Change, 4, 577–582, doi:10.1038/nclimate2268. ——, and I. Simmonds, 2010: The central role of diminishing sea ice in recent Arctic temperature amplification. Nature, 464, 1334– 1337, doi:10.1038/nature09051. ——, ——, C. Deser, and R. Tomas, 2013: The atmospheric response to three decades of observed Arctic sea ice loss. J. Climate, 26, 1230–1248, doi:10.1175/JCLI-D-12-00063.1. Simmonds, I., and K. Keay, 2009: Extraordinary September Arctic sea ice reductions and their relationships with storm behavior over 1979–2008. Geophys. Res. Lett., 36, L19715, doi:10.1029/ 2009GL039810. Simmons, A., J. Wallace, and G. Branstator, 1983: Barotropic wave propagation and instability, and atmospheric teleconnection patterns. J. Atmos. Sci., 40, 1363–1392, doi:10.1175/ 1520-0469(1983)040,1363:BWPAIA.2.0.CO;2. Song, L., L. Wang, W. Chen, and Y. Zhang, 2016: Intraseasonal variation of the strength of the East Asian trough and its climatic impacts in boreal winter. J. Climate, 29, 2557–2577, doi:10.1175/JCLI-D-14-00834.1. Straus, D. M., and J. Shukla, 1997: Variations of midlatitude transient dynamics associated with ENSO. J. Atmos. Sci., 54, 777–790, doi:10.1175/1520-0469(1997)054,0777:VOMTDA.2.0.CO;2. Strong, C., G. Magnusdottir, and H. Stern, 2009: Observed feedback between winter sea ice and the North Atlantic Oscillation. J. Climate, 22, 6021–6032, doi:10.1175/ 2009JCLI3100.1. Sun, L., C. Deser, and R. A. Tomas, 2015: Mechanisms of stratospheric and tropospheric circulation response to projected Arctic sea ice loss. J. Climate, 28, 7824–7845, doi:10.1175/ JCLI-D-15-0169.1. ——, J. Perlwitz, and M. Hoerling, 2016: What caused the recent ‘‘warm Arctic, cold continents’’ trend pattern in winter temperatures? Geophys. Res. Lett., 43, 5345–5352, doi:10.1002/ 2016GL069024. Takahashi, C., and R. Shirooka, 2014: Storm track activity over the North Pacific associated with the Madden-Julian oscillation under ENSO conditions during boreal winter. J. Geophys. Res. Atmos., 119, 10 663–10 683, doi:10.1002/2014JD021973. Tang, Q., X. Zhang, X. Yang, and J. A. Francis, 2013: Cold winter extremes in northern continents linked to Arctic sea ice loss. Environ. Res. Lett., 8, 014036, doi:10.1088/1748-9326/8/1/ 014036. Trenberth, K. E., and J. W. Hurrell, 1994: Decadal atmosphereocean variations in the Pacific. Climate Dyn., 9, 303–319, doi:10.1007/BF00204745. Walker, G. T., 1924: Correlations in seasonal variations of weather, IX: A further study of world weather. Mem. Indian Meteor. Dep., 24, 275–332.

VOLUME 30

Wallace, J. M., G.-H. Lim, and M. L. Blackmon, 1988: Relationship between cyclone tracks, anticyclone tracks and baroclinic waveguides. J. Atmos. Sci., 45, 439–462, doi:10.1175/ 1520-0469(1988)045,0439:RBCTAT.2.0.CO;2. Walsh, J. E., 2014: Intensified warming of the Arctic: Causes and impacts on middle latitudes. Global Planet. Change, 117, 52– 63, doi:10.1016/j.gloplacha.2014.03.003. Wang, L., and W. Chen, 2014: The East Asian winter monsoon: Reamplification in the mid-2000s. Chin. Sci. Bull., 59, 430–436, doi:10.1007/s11434-013-0029-0. Wang, W., M. Chen, and A. Kumar, 2013: Seasonal prediction of Arctic sea ice extent from a coupled dynamical forecast system. Mon. Wea. Rev., 141, 1375–1394, doi:10.1175/ MWR-D-12-00057.1. Wang, X. L., V. R. Swail, and F. W. Zwiers, 2006: Climatology and changes of extratropical cyclone activity: Comparison of ERA-40 with NCEP–NCAR reanalysis for 1958–2001. J. Climate, 19, 3145–3166, doi:10.1175/JCLI3781.1. Westra, S., L. V. Alexander, and F. W. Zwiers, 2013: Global increasing trends in annual maximum daily precipitation. J. Climate, 26, 3904–3918, doi:10.1175/JCLI-D-12-00502.1. Woods, C., R. Caballero, and G. Svensson, 2013: Large-scale circulation associated with moisture intrusions into the Arctic during winter. Geophys. Res. Lett., 40, 4717–4721, doi:10.1002/ grl.50912. 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. ——, J. M. Gregory, J. G. Pinto, M. Reyers, and D. J. Brayshaw, 2012: Response of the North Atlantic storm track to climate change shaped by ocean–atmosphere coupling. Nat. Geosci., 5, 313–317, doi:10.1038/ngeo1438. 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. Yoo, C., S. Lee, and S. B. Feldstein, 2012: Mechanisms of arctic surface air temperature change in response to the Madden– Julian oscillation. J. Climate, 25, 5777–5790, doi:10.1175/ JCLI-D-11-00566.1. Zappa, G., L. C. Shaffrey, K. I. Hodges, P. G. Sansom, and D. B. Stephenson, 2013: A multimodel assessment of future projections of North Atlantic and European extratropical cyclones in the CMIP5 climate models. J. Climate, 26, 5846–5862, doi:10.1175/JCLI-D-12-00573.1. ——, G. Masato, L. Shaffrey, T. Woollings, and K. Hodges, 2014: Linking Northern Hemisphere blocking and storm track biases in the CMIP5 climate models. Geophys. Res. Lett., 41, 135– 139, doi:10.1002/2013GL058480. Zhang, X., J. He, J. Zhang, I. Polyakov, R. Gerdes, J. Inoue, and P. Wu, 2013: Enhanced poleward moisture transport and amplified northern high-latitude wetting trend. Nat. Climate Change, 3, 47–51, doi:10.1038/nclimate1631.