A Pan–North Pacific Wintertime Surface Air Temperature Pattern Influenced by the SST Anomalies over the Kuroshio and Oyashio Extension

Zhihui Che aCollege of Life Sciences and Oceanography, Shenzhen University, Shenzhen, China
bCollege of Physics and Optoelectronic Engineering, Shenzhen University, Shenzhen, China

Search for other papers by Zhihui Che in
Current site
Google Scholar
PubMed
Close
,
Lin Mu aCollege of Life Sciences and Oceanography, Shenzhen University, Shenzhen, China

Search for other papers by Lin Mu in
Current site
Google Scholar
PubMed
Close
https://orcid.org/0000-0002-2966-6370
, and
Linhao Zhong cNational Institute of Natural Hazards, Ministry of Emergency Management of the People’s Republic of China, Beijing, China

Search for other papers by Linhao Zhong in
Current site
Google Scholar
PubMed
Close
Free access

Abstract

A new wintertime surface air temperature (SAT) pattern, called the Asia–Kuroshio and Oyashio Extension–North America (AKNA) pattern, is identified over the pan–North Pacific region (85°E–85°W, 25°–65°N) based on NCEP reanalysis data. The AKNA pattern is likely to influence the climate of the extratropical area of Asia and North America via two SAT dipoles and has a significant impact on the wintertime extremely cold weather along the eastern coastal regions of East Asia. Simulations using an atmospheric general circulation model indicate that wintertime sea surface temperature anomalies (SSTa) in the Kuroshio and Oyashio Extension (KOE) region can force an equivalent barotropic atmospheric ridge downstream and weaken the Siberian high and Alaska atmospheric ridge, resulting in the formation of the AKNA pattern. This circulation pattern tends to intensify the midlatitude (40°–60°N) westerlies over East Asia, which inhibits the southward invasion of the cold air into southern East Asia. Further diagnostic analysis indicates that the KOE SSTa can modulate the variation of storm track and westerlies by affecting baroclinic instability and eddy–mean flow interaction. Moreover, the KOE SSTa can provide a favorable environment for the development of the local atmospheric ascending motion and secondary circulation across the KOE SSTa, thereby affecting variability of the free atmosphere.

Significance Statement

This study aims to build a connection between the wintertime extratropical climate and the variation of the Kuroshio and Oyashio Extension. This work isolated a new wintertime surface air temperature (SAT) pattern over mid–high-latitude Asia and North America, which explains a considerable proportion of cold extremes over the eastern regions of East Asia. The reanalysis data and model simulations indicate that the temporal variability of the SAT pattern is influenced by the change of sea surface temperature in the Kuroshio and Oyashio Extension. These findings emphasize the important role of midlatitude air–sea interaction in the modulation of the mid–high-latitude climate.

© 2023 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

This article is included in the Climate Implications of Frontal Scale Air-Sea Interaction Special Collection.

Corresponding author: Lin Mu, mulin@szu.edu.cn

Abstract

A new wintertime surface air temperature (SAT) pattern, called the Asia–Kuroshio and Oyashio Extension–North America (AKNA) pattern, is identified over the pan–North Pacific region (85°E–85°W, 25°–65°N) based on NCEP reanalysis data. The AKNA pattern is likely to influence the climate of the extratropical area of Asia and North America via two SAT dipoles and has a significant impact on the wintertime extremely cold weather along the eastern coastal regions of East Asia. Simulations using an atmospheric general circulation model indicate that wintertime sea surface temperature anomalies (SSTa) in the Kuroshio and Oyashio Extension (KOE) region can force an equivalent barotropic atmospheric ridge downstream and weaken the Siberian high and Alaska atmospheric ridge, resulting in the formation of the AKNA pattern. This circulation pattern tends to intensify the midlatitude (40°–60°N) westerlies over East Asia, which inhibits the southward invasion of the cold air into southern East Asia. Further diagnostic analysis indicates that the KOE SSTa can modulate the variation of storm track and westerlies by affecting baroclinic instability and eddy–mean flow interaction. Moreover, the KOE SSTa can provide a favorable environment for the development of the local atmospheric ascending motion and secondary circulation across the KOE SSTa, thereby affecting variability of the free atmosphere.

Significance Statement

This study aims to build a connection between the wintertime extratropical climate and the variation of the Kuroshio and Oyashio Extension. This work isolated a new wintertime surface air temperature (SAT) pattern over mid–high-latitude Asia and North America, which explains a considerable proportion of cold extremes over the eastern regions of East Asia. The reanalysis data and model simulations indicate that the temporal variability of the SAT pattern is influenced by the change of sea surface temperature in the Kuroshio and Oyashio Extension. These findings emphasize the important role of midlatitude air–sea interaction in the modulation of the mid–high-latitude climate.

© 2023 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

This article is included in the Climate Implications of Frontal Scale Air-Sea Interaction Special Collection.

Corresponding author: Lin Mu, mulin@szu.edu.cn

1. Introduction

In winter, the anomalous surface air temperature (SAT), particularly the extremely cold weather, has an important impact on ecosystem, human health, economic, and social activities (Blunden et al. 2011). The pan–North Pacific region, mainly consisting of extratropical East Asia and North America, is one of the most frequently influenced regions by the cold extremes. Because this region is also a densely populated and economically developed area, understanding the variability of atmospheric circulation and the associated SAT anomaly in the pan–North Pacific region is crucial.

Several previous studies have explored the dominant modes of winter SAT variability over the extratropical Northern Hemisphere (Koide and Kodera 1999; Miyazaki and Yasunari 2008; Wu et al. 2009; Park et al. 2021). For example, the contribution of Arctic sea ice loss to the warm Arctic–cold Eurasian continent mode has been extensively investigated (Cohen et al. 2012; Mori et al. 2014; Woollings et al. 2014; Kug et al. 2015). The reduced westerlies favor the cold air advection (Yao et al. 2017), which leads to a colder Eurasian continent. For North America, many studies emphasize the crucial role of the Pacific–North America (PNA) mode and Arctic Oscillation/North Atlantic Oscillation (AO/NAO) during boreal winter (Rodionov and Assel 2003; Griffiths and Bradley 2007; Ning and Bradley 2016). Yu et al. (2016) and Yu and Lin (2018) proposed the Asian–Bering–North American mode, which describes the coherent changes of wintertime SAT over North Asia and North America. In addition, powerful basin-scale oceanic processes, including El Niño, Pacific decadal oscillation (PDO), Victoria mode (VM), and Atlantic multidecadal oscillation (AMO), among others, are believed to affect nearly global atmospheric circulation and SAT (Trenberth and Hurrell 1994; Wang et al. 2000; Sheppard et al. 2002; Li and Bates 2007; Linkin and Nigam 2008; Kim et al. 2014; Wang 2019).

These studies mainly focus on the forcing role of polar sea ice and basin-scale sea surface temperature anomalies (SSTa). On the other hand, the midlatitude SST frontal zone is another hotspot of air–sea interaction despite the narrow spatial scale (Chelton et al. 2004; Nakamura et al. 2004, 2008; Minobe et al. 2008, 2010; Chelton and Xie 2010; Kwon et al. 2010). The SST fronts could influence the convergence and divergence in the local air–sea boundary layer by causing differences in the release of heat and moisture across the front (Lindzen and Nigam 1987). The SST fronts can also promote (inhibit) the vertical mixing of air to accelerate (decelerate) wind speed over the warmer (cooler) side of the SST front (Wallace et al. 1989). Thus, the midlatitude SST fronts play an important role in shaping the average state of the midlatitude atmospheric circulation, including the anchoring and maintenance of the storm track and the westerly jet (Hoskins and Valdes 1990; Nakamura et al. 2004, 2008; Frankignoul et al. 2011; Taguchi et al. 2009, 2012; Kwon and Deser 2007; Kwon et al. 2010; Ma et al. 2017; Yao et al. 2018; Wang et al. 2019; Zhang et al. 2019; C. Zhang et al. 2020).

The sharpest SST frontal zone in the wintertime North Pacific is located at the confluence region of the Kuroshio Extension and the Oyashio Extension, which is called the Kuroshio and Oyashio Extension (KOE) region. A series of modeling studies has strongly suggested the significant response of the tropospheric atmosphere to the SSTa over the western boundary current. Variables such as the atmospheric horizontal divergence, vertical motion, precipitation, and cloud cover all have a good coupling correspondence with the KOE SSTa (Liu and Xie 2008; O’Neill et al. 2010; Taguchi et al. 2012; Masunaga et al. 2016, 2020; Wang and Liu 2015; Z. Zhang et al. 2020). In addition, both the observations (Kwon and Joyce 2013; O’Reilly and Czaja 2015; Révelard et al. 2016, 2018; Wills and Thompson 2018) and modeling studies (Smirnov et al. 2015; Tatebe et al. 2017; Okajima et al. 2018) have explored the possible impact of the KOE SSTa on the large-scale atmospheric circulation. Considering the importance of midlatitude air–sea interaction in climate change and extreme weather prediction, a few studies have begun to investigate the climatic effect of western boundary current (Ma et al. 2015; Yu et al. 2017, 2019; Lei et al. 2020; Che et al. 2021; Long et al. 2021; Siqueira et al. 2021). But there is still a lack of clear understanding of the specific pathway and mechanism of the interaction between the SSTa and extratropical regional climate. Therefore, this study attempts to investigate the possible link between the wintertime pan–North Pacific SAT pattern and the variation of KOE SSTa.

The remainder of the paper is organized as follows. The dataset, model experiments design, and method used are described in section 2. Section 3 shows the principal SAT modes in the pan–North Pacific region and its related oceanic variability. The atmospheric circulation and SAT responses to the KOE SSTa are illustrated in section 4. Section 5 explores the dynamic mechanism of the atmospheric response. Both the results of data reanalysis and model experiments are presented in sections 4 and 5. Finally, conclusions and further discussion are presented in section 6.

2. Data and model

a. Data

The daily, multilevel atmospheric variables are obtained from National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis (Kalnay et al. 1996), which is available with a 2.5° × 2.5° spatial resolution. The daily SAT from NCEP reanalysis used in this paper is defined by air temperature at the lowest sigma level. Observed monthly SAT anomalies are obtained from Goddard Institute for Space Studies Surface Temperature Analysis (GISTEMP), version 4, which is provided by the National Aeronautics and Space Administration with a 2° × 2° spatial resolution (Lenssen et al. 2019; GISTEMP Team 2022). The NCEP reanalysis and GISTEMP data used in this study extend from December 1949 to February 2020. Monthly skin temperature is gained from the fifth major global reanalysis produced by the European Center for Medium-Range Weather Forecasts (ECMWF) (ERA5; Hersbach et al. 2020) with a spatial resolution at 0.25° × 0.25° from December 1959 to February 2020. Monthly SAT is gained from Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2; Gelaro et al. 2017), with a spatial resolution at 0.5° × 0.625°. The MERRA-2 reanalysis data are provided by the National Aeronautics and Space Administration (NASA) from December 1980 to February 2020. Monthly ground temperature is gained from the Japanese 55-year Reanalysis (JRA-55; Kobayashi et al. 2015; Harada et al. 2016) with a spatial resolution at 1.25° × 1.25°. The JRA-55 data are provided by the Japan Meteorological Agency (JMA) from December 1958 to February 2020. The wintertime period means 90 days from 1 December of one year to 28 February of the next year. The extremely cold weather is defined by daily wintertime SAT anomalies grid by grid. If a grid SAT is lower than the threshold for at least three consecutive days, this grid point is experiencing extremely cold weather. The threshold is defined by the lower-tenth percentile of the total daily wintertime SAT anomaly sequence at each grid point.

To cover a sufficiently long period and obtain a clear structure of SSTa, two SST datasets are used in this study: monthly Hadley Centre Global SST (HadISST) (Rayner et al. 2003) with a 1° × 1° spatial resolution and daily Optimum Interpolation Sea Surface Temperature (OISST), version 2, dataset with a 0.25° × 0.25° spatial resolution (Reynolds et al. 2002). The monthly HadISST is obtained for the period between December 1949 and February 2020, while the daily OISST is obtained for the period between December 1981 and February 2020. Daily sensible and latent heat fluxes for the period 1958–2020 are gained from the WHOI objectively analyzed air–sea fluxes (OAFlux) project (Yu et al. 2008) with a 1° × 1° spatial resolution.

The monthly AO, Niño-3, and Niño-4 indices are obtained from the NOAA website (https://psl.noaa.gov/data/climateindices/list/), extending from January 1950 to December 2020. The wintertime PDO and the VM index are calculated by the principal components (PCs) time series of the first and second dominant empirical orthogonal function (EOF) mode of North Pacific (poleward of 20°N) SST variability, using HadISST data. The index of eastern Pacific (EP) El Niño and central Pacific (CP) El Niño are calculated by the method proposed by Yeh et al. (2009). A KOE front (KOEF) index proposed by Che et al. (2021) is calculated to demonstrate the variation of the KOE region. This index is defined by the regional average of the meridional SST gradient over the strongest SST frontal zone (142°–155°E, 38°–43°N) in the KOE region.

The global warming signal is defined by the ensemble mean of annual global SAT from phase5 of the Coupled Model Intercomparison Project (CMIP5) provided by Dai et al. (2015), which is designed to isolate the contribution of external forcing to global warming from the atmospheric internal variability. During the research period of this paper (1949–2019), the global warming trend is nearly linear. All reanalysis and observation data used in this study have had the part removed that linearly regressed on the global warming signal. Anomalous fields in this study are calculated by removing the long-term mean (1949–2019 for most cases) for each day/month at each grid point.

b. Model description and experiment design

The Community Earth System Model, version 2.2.1 (CESM 2.2.1; Danabasoglu et al. 2020), is a global climate model created by NCAR. Different geophysical models, including atmosphere, ocean, and land, among others, are coupled together by the version 7 coupler (Craig et al. 2012). The model component set used in this article is F2000climo, in which the Community Atmosphere Model, version 6.0 (CAM6; Gettelman et al. 2019), is coupled to the Community Land Model, version 5.0 (CLM5; Lawrence et al. 2019), and forced by prescribed SST and sea ice. CAM6 is integrated with 0.9° × 1.25° horizontal resolution, 32 vertical levels, and a time step of 30 min. The component set F2000climo includes a set of monthly climatological SST and sea ice concentration boundary fields, atmospheric initial fields, and external forcing fields. The SST boundary fields in F2000climo contain the climatological averaged SST fields from 1995 to 2005 for 12 months of the year. They are produced from monthly HadISST, version 1, and weekly OISST, version 2 (Hurrell et al. 2008).

To spin up, the model is run for 5 years under the default monthly climatological SST and sea ice concentration boundary conditions, constant atmospheric composition, and other external forcings. The first day of the sixth year is used as the initial time for the following experiments. The control (CTRL) run keeps all default external forcing and boundary conditions. For the forcing (FORC) run, the SST boundary field from March to November also keeps all default external forcing and boundary conditions. But in December–February (DJF), the SST boundary field of the FORC run is the sum of default SST climatology and fixed forcing SSTa signal. The fixed SSTa (as shown in Fig. 6e) added in the FORC run is the composite differences of wintertime SST between strong and weak KOEF winters over the KOE region. Detailed description of the fixed SSTa forcing signal can be found in section 3c. Both experiments are integrated from the same initial field for 30 years. The differences between the two experiments (the result of FORC minus the result of CTRL) represent the atmospheric response to wintertime SSTa in the KOE region.

c. Method

The synoptic eddy activity is obtained by a 2.5–8-day bandpass filter. The storm track is defined by meridional eddy heat flux at 850 hPa and meridional eddy wind variance at 300 hPa (Blackmon et al. 1977; Lau 1978). Baroclinic instability in this article refers to the maximum Eady growth rate (EGR; Lindzen and Farrell 1980) as shown in Eq. (1), in which f, u, g, and θ denote the Coriolis parameter, zonal wind speed, gravitational acceleration, and potential temperature, respectively:
EGR=0.31fuz/gdθθdz.
The horizontal divergence of the two-dimensional E vector [Eq. (2)] is investigated to quantify the synoptic eddy feedback to the zonal wind time tendency (Hoskins and White 1983). The u and υ in Eq. (2) denote the zonal and meridional wind speed. The overbar denotes the time means, and the prime denotes the synoptic-scale component. The divergence (convergence) of the E vector indicates the acceleration (deceleration) of mean zonal wind contributed by synoptic eddy activity:
E=(υ2u2¯,uυ¯),
In addition, barotropic kinetic energy conversion (BTEC) and baroclinic energy conversion (BCEC) are examined to further demonstrate the process of atmospheric eddy–mean flow interaction (Cai et al. 2007). A positive BTEC means the kinetic energy is converted from mean flow to eddy activity, and a positive BCEC means the available potential energy is converted from mean state to eddy activity. The BTEC and BCEC are defined as follows:
BTEC=P0gυ2u22(u¯xυ¯y)P0guυ(u¯y+υ¯x),
BCEC=C0(uTT¯xυTT¯y), and
C0=(PP0)R/CPdθdp(PP0)CV/CPgR,
where u, υ, and T are the zonal wind, meridional wind, and atmospheric temperature, respectively. The terms R, CP, CV, and g are the gas constant, specific heat capacity at constant pressure or constant volume, and acceleration of gravity, respectively. The term P0 is defined as 1000 hPa; P is the air pressure; θ is potential temperature. The overbar and prime indicate the climatological mean and the synoptic-scale disturbance, respectively.
Unless otherwise specified, regression analysis and composite analysis in this study are based on DJF mean data. The two-sided Student’s t test is used to examine the significance of the correlation, regression, and composite analysis. The effective degree of freedom (EDOF) is defined by the following equation (Bretherton et al. 1999), where r1 and r2 are the lag-1 autocorrelation of the time series and n is the sample size:
EDOF=n(1r1r2)(1+r1r2).

3. Wintertime pan–North Pacific SAT patterns and related SST signals

a. The first four pan–North Pacific SAT EOF modes

The principal mode of wintertime pan–North Pacific SAT is investigated by EOF analysis over the region 85°E–85°W and 25°–65°N (referred to as the pan–North Pacific region in this paper, as circled by magenta boxes in Fig. 1), based on NCEP reanalysis data. Because of the large latitudinal span of the pan–North Pacific region, the area represented by the grid points at different latitudes varies greatly. To avoid exaggerating the influence of high-latitude grid points, which only represent a small area of the pan–North Pacific region, the DJF mean SAT anomalies used to calculate the EOF mode are weighted by multiplying the square root of the cosine of latitude. Only the SAT anomalies over land are selected for the EOF analysis. The variance contribution rates of the first six EOF modes are 29.08%, 18.43%, 10.82%, 8.90%, 5.04%, and 4.53%. The method proposed by North et al. (1982) was used to test whether the EOF mode was significant. The specific criterion is as follows: when the eigenvalue γk of the kth EOF mode satisfies Eq. (7), in which n is the sample size, then the kth EOF mode is significant,
γkγk+1γk(2n)1/2.
Only the first four EOF modes pass the significance test. Figure 1 shows the regression coefficients of geopotential height at 500 hPa (Z500, hereafter) and SAT on the normalized PC time series (gray bars in Fig. 2). The first EOF mode (EOF 1; Fig. 1a) depicts the coherent change of SAT over the mid–high-latitude pan–North Pacific region. There are significant positive correlations between PC 1 and SAT in Eurasia and North America to the north of 40°N (Yu et al. 2016; Yu and Lin 2018), while there are negative correlations in the Bering Strait and Greenland. The relevant atmospheric circulation field (Z500) shows significant opposite-sign signals between the polar regions and mid–high latitudes, which indicates that EOF 1 is related to the variability of the polar vortex.
Fig. 1.
Fig. 1.

Regression coefficients of DJF mean SAT (shading; K) and Z500 (contours at 10-m intervals; solid line = positive; dashed line = negative) on the normalized first four PC time series. The magenta box marks the region used to calculate the EOF modes. Dot hatching indicates where the regression coefficient of SAT is significant at a 95% confidence level.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

Fig. 2.
Fig. 2.

(a) Normalized PC 1 time series from NCEP (gray bars), ERA5 (black line), MERRA-2 (blue line), and JRA-55 (red line). (b) As in (a), but for PC 2. (c) As in (a), but for PC 3 time series from NCEP and MERRA-2, and PC 4 time series from ERA5 and JRA-55. (d) As in (a), but for PC 4 time series from NCEP and MERRA-2, and PC 3 time series from ERA5 and JRA-55. The correlation coefficients between the NCEP PC time series and other PC time series are given at the bottom of the panels in corresponding colors.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

The EOF 2 (Fig. 1b) demonstrates an opposite change in SAT over East Asia and North America. The regressed Z500 field against the PC 2 has three local extrema, which are in Siberia, North America, and North Pacific. In the positive (negative) phase of EOF 2, the intensified (weakened) Siberia high pressure and East Asian trough will promote (inhibit) the invasion of cold air to the south to 40°N. EOFs 3 (Fig. 1c) and 4 (Fig. 1d) both depict the regression coefficients of SAT with dipole patterns over extratropical East Asia and North America. EOF 3 depicts an opposite change in SAT from north to south over extratropical East Asia and an opposite change from northwest to southeast over North America. The regressed Z500 field against PC 3 shows a PNA-like pattern, with a deepened Aleutian low pressure and strengthened Alaska ridge. For EOF 4 (Fig. 1d), the regression coefficients of SAT over East Asia (North America) are similar (opposite) to that in Fig. 1c, associated with the change of East Asian trough, Alaska ridge, and North Pacific westerly jet.

EOF analysis results may not be the same when using different datasets. Figure 2 compares the PC time series obtained by different reanalysis datasets. Despite the four datasets’ differences in the temporal scope and spatial resolution, the first two EOF modes obtained are almost identical. The orders of EOFs 3 and 4 in JRA-55 and ERA5 are reversed compared to the NCEP results. Nevertheless, the correlation coefficients of the PC 3 and PC 4 time series between the results calculated from NCEP and the other reanalysis data are above 0.69.

b. The related SST signals

To investigate the relevant climate phenomena in the North Pacific and the possible link between them to these four EOF modes, Fig. 3 shows the regression coefficients of Pacific SST on the normalized PC time series. In the Pacific, only the SST changes over the eastern equatorial Pacific are significantly correlated with EOF 1 (Fig. 3a), and the correlation coefficient between the EP El Niño and PC 1 is 0.27 (Table 1). The variability of EOF 1 is highly related to the change of polar vortex, with a correlation coefficient of 0.60. AO is also significantly related to PCs 2 and 3. Figure 3b shows a band of significant warm SSTa from the Kuroshio to the central North Pacific, surrounded by opposite-sign SSTa to the north of 40°N and the south of 20°N. Those SSTa are similar to the pattern regressed to the VM, and the correlation coefficient between PC 2 and the VM index is −0.41. As for EOF 3, the regressed SST field shows a clear PDO pattern and significant signal over tropical Pacific, and the correlation coefficients between PDO and EP and CP El Niño are significant at 0.62, 0.26, and 0.32, respectively. Figure 3d shows significant warm SSTa on the coast of the northwest Pacific, especially to the east of Japan, that is, the KOE region. It also presents weak VM-like and EP El Niño–like SSTa patterns, and the correlation coefficient analysis shows PC 4 is significantly correlated with the VM index and EP El Niño index. This result is not surprising, considering the impact of PDO, VM, and El Niño on the variability of the Kuroshio Extension (Qiu and Chen 2005, 2010; Taguchi et al. 2007; Ceballos et al. 2009; Di Lorenzo et al. 2010).

Fig. 3.
Fig. 3.

Regression coefficients of DJF mean OISST (K) on the normalized first four PC time series. Dot hatching indicates where the regression coefficient is significant at a 95% confidence level.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

Table 1.

Correlation of multiple climate indices with the PC time series and the AKNA index. The asterisk indicates the correlation coefficient is significant at a 95% confidence level.

Table 1.

c. AKNA pattern and KOE SSTa

As stated in the introduction, the influence of climate signals, including AO, El Niño, PDO, and VM, on the wintertime SAT has been extensively discussed (Trenberth and Hurrell 1994; Wang et al. 2000; Sheppard et al. 2002; Rodionov and Assel 2003; Kim et al. 2014; Wang 2019). However, the correlation between the EOF 4 mode and the SSTa in the KOE region suggests that EOF 4 may play the role of being the “bridge” between extratropical climate and the SST frontal zone in the KOE region. Because of the association of EOF 4 with KOE SSTa, and SAT anomalies in East Asia and North America, this EOF 4–related SAT pattern is called the Asia–Kuroshio and Oyashio Extension–North America (AKNA) pattern herein.

Although the variance contribution of the EOF 4 mode passed the significance test, it is still generated from the residuals of the EOF 1–3 modes. Considering the differences in EOF analysis among the four reanalyses and to further confirm the reliability of the AKNA pattern, an objective AKNA index is defined based on regional mean SAT anomalies. Figure 4a shows the regression coefficients of wintertime SAT on the normalized NCEP PC 4 time series, in which four areas [marked by A1 (70°E–120°W, 60°–75°N), A2 (80°–120°E, 35°–50°N), A3 (100°–120°E, 25°–35°N), and A4 (90°–75°W, 35°–50°N)] are chosen as key areas to compute the AKNA index. The formula of the AKNA index is shown in Eq. (8). The term An represents the SAT anomalies in the region, and the overbar represents the area-weighted average. As shown, the four key areas’ regression coefficients are unequal. Therefore, the weight coefficient of An is approximately derived by the regional average of the regression coefficients in each area. The weight coefficient of A1¯ is fixed to −1.0, and the weight coefficients of other regions are scaled up and down in the same proportion. For simplicity, all weight coefficients are only reserved to one decimal place:
AKNA=0.9A2¯+0.8A3¯+0.8A4¯A1¯.
The time series of A2¯ and A3¯ are highly correlated at 0.78. Aside from this pair, these four regional time series can be regarded as independent of each other. Thus, the contribution of three terms (A1¯,0.9A2¯+0.8A3¯,and0.8A4¯) on the variance of AKNA is approximately estimated by the ratio of the variance of each term to the variance of AKNA index, which are 28.77%, 45.38%, and 23.96%, respectively. The regression field of SAT (Fig. 4b) on the normalized AKNA index is extremely similar to EOF 4 (Fig. 4a). Results from the observational GISTEMP data (Fig. 4c) and other reanalysis datasets (Figs. 4d–f) confirm the validity of the AKNA pattern. The correlation coefficients of the AKNA index from NCEP and from other datasets reach 0.97 (GISTEMP), 0.97 (ERA5), 0.98 (MERRA-2), and 0.98 (JRA-55). For clarity, the newly established NCEP AKNA index will be used to represent the change of the AKNA pattern in subsequent analyses. It is also worth mentioning that the conclusions obtained using the AKNA index or the PC 4 index are consistent.
Fig. 4.
Fig. 4.

Regression coefficients of wintertime SAT (shading; K) on (a) the normalized PC 4 index, (b) the normalized NCEP AKNA index, (c) the normalized GISTEMP AKNA index, (d) the normalized ERA5 AKNA index, (e) the normalized MERRA-2 AKNA index, and (f) the normalized JRA55 AKNA index. Dot hatching indicates where the regression coefficient is significant at a 95% confidence level. (g) Five normalized AKNA indices.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

Figure 5a exhibits the regression coefficients of extremely cold weather on the AKNA index. The variation of extremely cold weather shows two dipole patterns in extratropical Asia and North America, just like the anomalous SAT pattern shown in Fig. 4b. The AKNA pattern exhibits a strong (weak) impact on the extremely cold weather in the eastern coastal regions of East Asia (North America), with the local variance contribution rate (square of correlation coefficient) reaching up to 20% (5%) in the eastern coastal region of East Asia (North America). Figures 5c–e show the regression coefficients of DJF mean SST and turbulent heat flux on the AKNA index. The SST and turbulent heat flux show opposing-sign correlations with the AKNA index in most parts of the Pacific Ocean; the intensified (reduced) heat flux forced by the atmosphere cools (warms) the ocean, which emphasizes the dominant role of the atmosphere (Cayan 1992). In the KOE region, both the SST and heat flux show significant and same-sign correlations with the AKNA index (as shown in Figs. 6b,c); the warmer (colder) seawater promotes (inhibits) the release of heat flux in the positive (negative) phase of AKNA, which indicates a possible forcing role of SSTa in the KOE region (Tanimoto et al. 2003; Chelton and Xie 2010).

Fig. 5.
Fig. 5.

Regression maps of (a) extremely cold weather (days), DJF mean; (c) SST (K; from OISST); (d) latent and (e) sensible heat flux (W m−2; from OAFlux, positive values indicate ocean losing heat) on the normalized AKNA index. Extremely cold weather days are calculated using NCEP reanalysis data. Dot hatching indicates where the regression coefficient is significant at a 95% confidence level. (b) The variance contribution rate (%) of the AKNA pattern to the variation of wintertime extremely cold weather.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

The KOEF index, the time series of the meridional SST gradient area average over the KOE region (142°–155°E, 38°–43°N, marked by the black dashed box in Fig. 6a), is calculated to depict the variation of the KOE region. The variation of KOE SST is related to the dynamic state of Kuroshio Extension flow (Vivier et al. 2002; Qiu and Chen 2005, 2010; Qiu et al. 2014; Che et al. 2021). When the Kuroshio Extension axis remains unstable, the intensified (reduced) meridional (zonal) flow and increased eddy activity will transport more warm water to the KOE region (Qiu 2000; Sugimoto and Hanawa 2009; Sugimoto et al. 2014; Masunaga et al. 2016), causing warming of the KOE region and strengthening of the KOEF. The change of KOEF lags the change of Kuroshio dynamics by about 1 year (Che et al. 2021). Figures 6c and 6d show the composite difference of SST and turbulent heat flux (sensible + latent; positive values indicate the ocean is losing heat) between the positive and negative phases of the KOEF. As the results show, there are positive regression coefficients of SST and turbulent heat flux in the KOE region, which means the KOEF index successfully portrayed the forcing role of KOE SST anomalies (Tanimoto et al. 2003; Chelton and Xie 2010) in local sea–air interaction.

Fig. 6.
Fig. 6.

(a) 1981–2019 climatology of wintertime SST front [K (100 km)−1] from OISST data. The black dashed box indicates the region in which the KOEF index is generated. Composite differences of DJF mean (b) SST meridional gradient [K (100 km)−1], (c) SST (K; derived from OISST), and (d) turbulent heat flux (W m−2; derived from OAFlux, positive values indicate ocean losing heat) between the positive and negative phase of the normalized OISST KOEF index. Dot hatching indicates where the difference is significant at a 95% confidence level. Magenta lines in (b) indicate the two front branches of KOEF. (e) The difference in DJF mean SST between the FORC and CTRL runs. This fixed forcing SSTa pattern is selected from the dashed box in (b). (f) Normalized AKNA index (bar: orange for positive, blue for negative) and KOEF indices calculated by HadISST (dashed black line) and OISST (solid black line) data.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

The contemporaneous correlation between the AKNA index and the OISST (HadISST) KOEF index is significant at 0.32 (0.29). Further lead–lag regression suggests the DJF mean AKNA pattern is significantly cross correlated with the value of the OISST KOEF when leading for 1–4 months or lagging for 1 month. Lead–lag regression also reveals the DJF mean AKNA pattern is significantly cross correlated with the value of the EP El Niño when leading for 1–3 months or lagging for 1–4 months. Considering the complexity of the lead–lag regression and the contamination by ENSO variability, it is difficult to determine the causality from observations alone. This study will focus on exploring whether KOE SSTa has a forcing effect on the formation of AKNA, as implied by Figs. 5a–c. The CAM6 numerical simulations are designed to explore the atmospheric response to the KOE SSTa. Thus, the KOEF-related SSTa is used as a forcing signal of CAM6 numeric simulation. The fixed forcing SSTa field added to the FORC run is shown in Fig. 6e, which is selected from the dashed box (140°–155°E, 32.5°–44°N) in Fig. 6c.

4. Response of atmospheric circulation and SAT to the KOE SSTa

a. Comparison between the NCEP reanalysis data and CTRL experiment

In this section, whether the AKNA pattern is forced by the KOE SSTa is investigated through CESM experiments. To test the ability of the CAM6 to represent the realistic wintertime atmospheric circulation, Fig. 7 compares the climatological wintertime zonal wind speed at 300 hPa (U300), stationary wave at Z500, and storm track at 300 hPa between NCEP reanalysis data and the CTRL run. The stationary wave is defined as the Z500 anomaly obtained by subtracting the zonal average. The model performed well in catching the distribution of the midlatitude westerly jet (Fig. 7b) and storm track (Fig. 7f), as their position and intensity are consistent with the result in NCEP reanalysis data (Figs. 7a,e). Similarly, Figs. 7c and 7d show a similar pattern of stationary wave (e.g., the East Asian trough and the Alaska ridge) in NCEP reanalysis data and the CTRL run. The comparisons confirm that the model has successfully depicted both the high- and low-frequency atmospheric processes, which increases the confidence in the model to test the atmospheric response to KOE SSTa.

Fig. 7.
Fig. 7.

Climatology of (a),(b) wintertime zonal wind speed (m s−1) at 300 hPa; (c),(d) stationary wave (m) at Z500; and (e),(f) meridional eddy heat flux (m K s−1) at 300 hPa from (left) NCEP reanalysis data and (right) CTRL run.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

Subsequently, the correlations between wintertime mean SST and turbulent heat flux are examined in North Pacific. According to Kalnay et al. (1996), the NCEP reanalysis model is forced by OISST, version 2, for the period 1982–present, and forced by GISST (replaced with the HadISST in this study) for earlier periods (1948–81). Thus, the correlations resolved by NCEP reanalysis data and observation were compared in two periods, 1949–80 (Figs. 8a,b) and 1982–2019 (Figs. 8c,d). The correlations between heat flux and SST depicted by NCEP data are basically consistent with the observation, except the NCEP reanalysis underestimates the correlations during 1982–2019. Over the KOE region, the result of NCEP reanalysis shows significant positive correlations during both periods, which indicates the NCEP reanalysis data can depict the forcing role of the KOE SSTa. The variability of the DJF mean turbulent heat flux resolved in the model (a combination of CTRL and FORC) is compared with that in observations. The variability of turbulent heat flux in model experiments is similar to the observed results but with much smaller magnitude. This is because the SST is prescribed as a monthly climatological mean in CAM6. The variation of turbulent heat flux reaches a peak in the KOE region in the CAM6 experiments, where the fixed SSTa is added.

Fig. 8.
Fig. 8.

(a)–(d) The correlations between DJF mean SST and turbulent heat flux. The dataset used is noted in the upper-left corner, and the time range used is noted in the upper-right corner. The black boxes in (a)–(d) indicate the domain where the KOE SSTa is located. Dot hatching indicates where the correlation coefficient is significant at a 95% confidence level. The standard deviations of DJF mean turbulent heat flux from (e) OAFlux and (f) a combination of CTRL and FORC runs.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

b. Response of atmospheric circulation and SAT in NCEP reanalysis data and model experiments

To facilitate comparison of results in NCEP reanalysis data and model experiments, positive and negative AKNA winters are to be selected at time periods coinciding with positive and negative phases of the normalized NCEP AKNA index. From 1949 to 2019, 34 winters are identified as positive AKNA winters, and 37 winters are identified as negative AKNA winters. Figures 9a and 9b show the composite differences of Z500 and SAT between the two AKNA phrases (Fig. 9a) and between the two experiments (Fig. 9b). On both continents, the climatological high pressure systems (i.e., the Siberian high pressure and the atmospheric ridge of the Alaska Gulf) are weakened, consistent with warming southeast of East Asia and North America. Over the North Pacific, the significant, downstream, positive Z500 anomalies confirm the ridge response of the atmosphere to warm SSTa over the KOE region (Kushnir and Lau 1992; Peng and Whitaker 1999; Liu and Wu 2004; Liu et al. 2007; Frankignoul et al. 2011; Gan and Wu 2012; Taguchi et al. 2012; Okajima et al. 2014; Révelard et al. 2016, 2018). The response of extremely cold weather to the variation of KOE is consistent with the change of SAT in both model experiments and NCEP reanalysis data.

Fig. 9.
Fig. 9.

(a) Composite differences in DJF mean SAT (shading; K) and Z500 (contours at 10-m intervals; solid line = positive; dashed line = negative) between positive and negative phases of AKNA. (b) As in (a), but for the differences between the FORC and CTRL runs (contours at 4-m intervals). (c),(d) As in (a) and (b), but for the extremely cold weather (days). Dot hatching indicates where the difference is significant at a 95% confidence level.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

The differences in temperature advection at 700 hPa are examined by Eq. (9) to explore the cause of the AKNA pattern (Wang et al. 2010). The terms V, T, and ∇ indicate the horizontal wind vector, air temperature, and gradient operator, respectively. The overbar and upper wavy lines indicate the climatological mean and the anomalous part. The temperature advection (−V ⋅ ∇T) was divided into advection by the anomalous wind across mean temperature (V˜T¯), advection by the mean wind across anomalous temperature (V¯T˜), and the nonlinear advection (V˜T˜). It is worth noting that V˜T¯ (Figs. 10a,b) generates the warming over the southeast part of both continents and contributes to the formation of the AKNA pattern. On the contrary, V¯T˜ (Figs. 10c,d) inhibits the formation of the AKNA pattern by dampening the effect of V˜T¯. The difference in the nonlinear part is relatively weak (Figs. 10e,f). The difference in downwelling radiative flux is consistent with that of SAT, which is mainly caused by downwelling longwave radiation:
VT=V˜T¯V¯T˜V˜T˜.
Figure 11 exhibits a barotropic response of the upper (represented by 300 hPa) and lower (represented by 850 hPa) troposphere to the variation of KOE. In the positive phase of AKNA and the FORC run, the midlatitude westerlies (Figs. 11a–d) over East Asia and the North Pacific strengthen to the north of 40°N and weaken to the south. The westerly jet shifts poleward, resulting in a north–south dipole pattern of zonal wind anomalies. The significant southerly wind anomalies (Figs. 11e–h) in the southeast of both continents indicate reduced equatorward transport of cold air. Compared to the results of NCEP data reanalysis, the westerlies dipole moves more northward to around 45°N in the model, and the wind speed anomalies are weak. The basin-scale, dipole-type westerlies response is consistent with previous research (Frankignoul et al. 2011; Taguchi et al. 2012; Nakamura and Miyama 2014; Omrani et al. 2019; Kohyama et al. 2021), which indicates the midlatitude westerly jet shifts meridionally with the warming of the western boundary current.
Fig. 10.
Fig. 10.

Composite differences in DJF mean horizontal temperature advection (K day−1) at 700 hPa between (a),(c),(e) positive and negative phases of AKNA and (b),(d),(f) two model experiments. Advection by anomalous wind across mean temperature is shown in (a) and (b). Advection by mean wind across anomalous temperature is shown in (c) and (d). Advection by anomalous wind across anomalous temperature is shown in (e) and (f). (g),(h) As in (a) and (b), but for downwelling radiative flux (longwave + shortwave) at surface. Dot hatching indicates where the difference is significant at a 95% confidence level.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

Fig. 11.
Fig. 11.

Composite differences in DJF mean (a)–(d) zonal and (e)–(h) meridional wind speed (m s−1) at 300 and 850 hPa between positive and negative phases of (left) AKNA and (right) two model experiments. Dot hatching indicates where the difference is significant at a 95% confidence level.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

Naturally, there are differences between the model results and NCEP reanalysis data that cannot be ignored. Compared to the results in the NCEP data reanalysis, the response of atmospheric circulation and SAT locates more northerly in the model, with a weaker amplitude. In North America, the meridional dipole pattern changes into a northwest–southeast type. The negative Z500 anomaly along the west coast of North America extends eastward and the positive Z500 anomaly in eastern North America diminishes. Kohyama et al. (2021) highlighted the synergistic influences of the Kuroshio and Gulf Stream on the atmospheric jet stream, which may explain the discrepancy between NCEP reanalysis data and model. The low consistency of the AKNA-related patterns (e.g., SAT, extremely cold weather, and westerlies) in eastern North America may be caused by the absence of the synergistic effect of the Gulf Stream in the model. In general, the above comparisons between the NCEP reanalysis data and model results suggest the response of the AKNA pattern to the variation of KOE SSTa. The physical mechanisms are further investigated in the following section.

5. Diagnosis of physical mechanisms

a. Response of baroclinic instability and storm track

Previous studies have emphasized the important role of midlatitude SST in the variation of the storm track (Nakamura et al. 2004, 2008; Kwon and Deser 2007; Kwon et al. 2010; Taguchi et al. 2009; Sampe et al. 2010; Ogawa et al. 2012; Ma et al. 2017). Figures 12a and 12d depict the response of the DJF mean storm track at 850 and 300 hPa. The storm track is strongly intensified to the north of the storm track axis in both the upper and lower troposphere in the positive phase of AKNA and FORC run. The North Pacific storm track moves poleward in response to the warm SSTa over the KOE region (Taguchi et al. 2009; Ogawa et al. 2012; Yao et al. 2018; Wang et al. 2019). As suggested by Vallis (2006), baroclinic instability is the key source of atmospheric eddy. Figures 12e and 12f demonstrate the differences in EGR between the positive and negative phases of AKNA and the two experiments. At the lower troposphere (850 hPa), the change of EGR not only shows consistent results in the NCEP data reanalysis and model experiments but also corresponds to the change of the storm track. The increased (decreased) EGR to the north (south) of 40°N during the positive (negative) phase of AKNA promotes (inhibits) the development of eddy activity.

Fig. 12.
Fig. 12.

Composite differences in DJF mean (a),(b) meridional eddy heat flux (m K s−1) at 850 hPa, (c),(d) meridional eddy wind variance (m2 s−2) at 300 hPa, and (e),(f) maximum Eady growth rate (10−6 s−1) at 850 hPa between positive and negative phases of (left) AKNA and (right) two model experiments. Dot hatching indicates where the difference is significant at a 95% confidence level.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

b. Response of eddy–mean flow interaction

Because the westerlies show a barotropic response to the KOE SSTa, the horizontal divergence of the two-dimensional E vector (Hoskins and White 1983) at 300 hPa is examined. During winter, the midlatitude North Pacific westerlies (around 40°N) are accelerated by synoptic eddy forcing over the North Pacific in both NCEP reanalysis data and model experiments (Figs. 13a,b). The differences between the two AKNA phases (Figs. 13c,d) indicate an enhanced acceleration of westerlies over the North Pacific and East Asia to the north of 40°N. Similar to the variation of zonal wind, the difference in divergence of the E vector between model experiments moves poleward compared with that in NCEP reanalysis data.

Fig. 13.
Fig. 13.

Climatology of DJF mean divergence of E vector at 300 hPa from (a) NCEP reanalysis data and (b) CTRL run. Composite differences in DJF mean divergence of E vector at 300 hPa between positive and negative phases of (c) AKNA and (d) two model experiments. Dot hatching indicates where the difference is significant at a 95% confidence level.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

The differences in energy conversion at the upper (300 hPa) and lower (850 hPa) troposphere further demonstrate the response of eddy–mean flow interaction. The significant variation of BCEC at the lower troposphere (Figs. 14a,b) explains the distribution of the storm track. Available potential energy is gained from the mean state to promote eddy activity to the north of 40°N. The difference of BTEC is stronger at 300 hPa instead (Figs. 14g,h), showing similar but opposite-sign differences to the 300-hPa BCEC. The eddy intensified by available potential energy loses kinetic energy to the mean flow to intensify the westerly jet aloft the KOE region, while in the downstream region, the eddy gains kinetic energy to dampen the intensification of mean flow (Cai et al. 2007; Gan and Wu 2014).

Fig. 14.
Fig. 14.

Composite differences in DJF mean (a)–(d) BCEC (W m−2) and (e)–(h) BTEC (W m−2) at 850 and 300 hPa between (left) positive and negative phases of AKNA and (right) two model experiments. Dot hatching indicates where the difference is significant at a 95% confidence level.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

c. Different responses of vertical motion in the western and eastern North Pacific

The response of vertical motion is investigated in this subsection. Since the warm KOE SSTa is restricted in KOE region, it is reasonable to separate the local atmosphere (hereafter western North Pacific) from the downstream part (hereafter eastern North Pacific). The longitude of the eastern boundary of the KOE region (155°E in this paper) is chosen to be the dividing line. Figure 15 presents the latitude–pressure plot for the vertical velocity (defined by omega, positive downward). There are significant differences in the vertical velocity anomalies between the eastern and western North Pacific. In the eastern North Pacific, there is enhanced sinking motion between 30° and 50°N during the positive AKNA winters, and that is reproduced in model experiments.

Fig. 15.
Fig. 15.

Vertical profiles of the differences in vertical atmospheric motion (presented by omega; 10−3 Pa s−1; a positive omega indicates downward atmospheric motion) between the positive and negative phases of (a),(c) AKNA and (b),(d) two model experiments. The variable is averaged over the (top) eastern North Pacific (155°E–140°W) or (bottom) western North Pacific (140°–155°E). Cross hatching indicates the region where the difference is significant at a 95% confidence level.

Citation: Journal of Climate 36, 16; 10.1175/JCLI-D-22-0351.1

In the western North Pacific, the enhanced upward motion is concentrated in the range of 20°–40°N during the positive phase of AKNA. The results of model experiments show a stronger and more complicated structure of the omega anomaly in the western North Pacific. Negative vertical velocity anomalies cover the latitude range of 20°–50°N. Among them, there is significantly enhanced ascending motion over the KOE area (centered at 37.5°N), with two corresponding sinking motions on the north and south sides. On the one hand, this result is consistent with previous studies, which suggest that the air ascends above the warm SSTa and sinks at the cold sides (Lindzen and Nigam 1987; Feliks et al. 2004, 2007; Chelton and Xie 2010; Masunaga et al. 2020). The secondary circulation forced by KOE SSTa is believed to reach the middle troposphere (Brachet et al. 2012; Wang and Liu 2015; Z. Zhang et al. 2020). On the other hand, the model results are likely to exaggerate the effect of KOE SSTa on the vertical motion of the atmosphere. This is because the fixed SST anomaly in the model has been forcing the atmosphere as a constant heat source, thus causing the vertical motion anomalies to far exceed the observed values.

6. Discussion and conclusions

The AKNA pattern, a new wintertime pan–North Pacific SAT pattern, is identified using EOF analysis based on NCEP reanalysis data. The EOF analysis is performed in the pan–North Pacific region (85°E–85°W, 25°–65°N), and long-term signals reflective of global warming have been removed. As the fourth EOF mode, the AKNA pattern presents a meridional dipole anomaly over extratropical Asia and a zonal dipole anomaly over North America, accounting for 8.90% of the variance in DJF mean pan–North Pacific SAT. A new AKNA index is defined by the difference of regionally averaged SAT anomalies. Correlation analysis indicates the AKNA pattern is related to the variation of SSTa over the KOE region.

Two model experiments are carried out to explore whether the AKNA pattern arises as a response to the variation of KOE SSTa, using the CAM6 model forced by monthly climatological prescribed SST field. The differences in atmospheric circulation and SAT between the FORC and CTRL runs are compared to the differences between positive and negative AKNA winters produced by NCEP reanalysis data. After adding warm SSTa in the wintertime KOE region (positive AKNA winters in NCEP), the downstream tropospheric atmosphere presents a ridge response, accompanied by a reduced Siberia high pressure, a reduced Alaska ridge, and an anticyclonic circulation over eastern North America. Consequently, the cold air from the polar region is restricted to northern Asia and northwestern North America, which makes southern East Asia and southeastern North America experience warmer winters and less extremely cold weather days.

Further analysis shows similar variations in the westerlies and storm tracks from the NCEP reanalysis and model experiments. For the FORC run/positive AKNA winters, the zonal wind and eddy heat flux shift poleward under the forcing of warm KOE SSTa. The variation of baroclinic instability and eddy feedback is consistent with the variation of storm tracks and zonal wind, which indicates that the KOE SSTa can modulate atmospheric eddy activity and westerlies through changing baroclinic instability. At 850 hPa, the increased (decreased) baroclinic energy conversion promotes (inhibits) atmospheric eddy activity to the north (south) of the westerly jet, causing the storm track to shift poleward. At 300 hPa, diagnostic analysis reveals an opposite-sign response of energy conversion over the western and eastern North Pacific. The intensified eddy activity converts energy to the mean flow to maintain the intensified zonal wind aloft the KOE region and absorbs energy from the mean flow to dampen the intensified zonal wind downstream in the jet exit region. In addition, the atmospheric vertical motion over the western and eastern North Pacific was investigated separately. For the FORC run/positive AKNA winters, the atmospheric vertical motion shows intensified upward motion over the KOE region and downward motion over the eastern North Pacific. In particular, the model experiments show, but may exaggerate, the response of the local atmosphere to the KOE SSTa. The atmosphere rises over warmer SST and sinks on both sides, and this vertical motion can extend from the boundary layer to Z500 troposphere.

Overall, this study reveals a new SAT pattern influenced by the variation of KOE SSTa and contributes to the understanding of their influence on climate. Results of the model experiments suggest the contribution of KOE SSTa to the formation of the AKNA pattern. However, the stabler KOE SSTa is not the only factor affecting the AKNA pattern. EP El Niño partly contributes to the interannual variability of AKNA, and the impact of atmospheric internal physical processes on the AKNA index remains to be explored. In addition, as mentioned by Kohyama et al. (2021), the synergistic effect of the Gulf Stream cannot be ignored. Although the most significant atmospheric response is over the North Pacific, the physical mechanism of atmospheric circulation changes over the two continents still needs further investigation. In addition, the conclusions need to be confirmed in a fully coupled and high-resolution model in future work.

Acknowledgments.

The authors acknowledge the support from the Shenzhen Fundamental Research Program (Grant JCYJ20200109110220482) and National Natural Science Foundation of China (Grant U2006210).

Data availability statement.

Daily NCEP–NCAR reanalysis data and climate indices are obtained from the NOAA PSL website (https://www.esrl.noaa.gov/psd/data/gridded/). Monthly GISTEMP data are obtained from the National Aeronautics and Space Administration’s Goddard Institute for Space Studies (https://data.giss.nasa.gov/gistemp/). Monthly ERA5 reanalysis data are obtained from the Copernicus Climate Change Service (C3S) Climate Date Store (https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-pressure-levels-monthly-means?tab=form). Monthly MERRA-2 reanalysis data are obtained from NASA website (https://disc.gsfc.nasa.gov/datasets/M2IMNXASM_5.12.4/summary). Monthly JRA-55 reanalysis data are obtained from the JMA website (https://search.diasjp.net/en/dataset/JRA55). HadISST data are obtained from Met Office Hadley Centre website (https://www.metoffice.gov.uk/hadobs/hadisst/data/download.html). OISSTv2 data are obtained from the NOAA PSL website (https://psl.noaa.gov/data/gridded/data.noaa.oisst.v2.highres.html). OAFlux data are obtained from the WHOI Objectively Analyzed Air-Sea Fluxes (OAFlux) project through an FTP server (ftp://ftp.whoi.edu/pub/science/oaflux/data_v3/). The numerical model simulations upon which this study is based are too large to archive or transfer. Instead, we provide all the information needed to replicate the simulations. We used CESM 2.2.1. The model code and the default initial field have not been modified in any way. The compilation scripts and boundary condition files are available at https://zenodo.org/record/7513953#.Y7roeP5Bz2c.

REFERENCES

  • Blackmon, M. L., J. M. Wallace, N. C. Lau, and S. T. Mullen, 1977: An observational study of the Northern Hemisphere wintertime circulation. J. Atmos. Sci., 34, 10401053, https://doi.org/10.1175/1520-0469(1977)034<1040:AOSOTN>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Blunden, J., D. S. Arndt, and M. O. Baringer, 2011: State of the Climate in 2010. Bull. Amer. Meteor. Soc., 92 (6), S1S236, https://doi.org/10.1175/1520-0477-92.6.S1.

    • Search Google Scholar
    • Export Citation
  • Brachet, S., F. Codron, Y. Feliks, M. Ghil, H. Le Treut, and E. Simonnet, 2012: Atmospheric circulations induced by a midlatitude SST front: A GCM study. J. Climate, 25, 18471853, https://doi.org/10.1175/JCLI-D-11-00329.1.

    • Search Google Scholar
    • Export Citation
  • Bretherton, C. S., M. Widmann, V. P. Dymnikov, J. M. Wallace, and I. Bladé, 1999: The effective number of spatial degrees of freedom of a time-varying field. J. Climate, 12, 19902009, https://doi.org/10.1175/1520-0442(1999)012<1990:TENOSD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Cai, M., S. Yang, H. M. Van den Dool, and V. E. Kousky, 2007: Dynamical implications of the orientation of atmospheric eddies: A local energetics perspective. Tellus, 59A, 127140, https://doi.org/10.1111/j.1600-0870.2006.00213.x.

    • Search Google Scholar
    • Export Citation
  • Cayan, D. R., 1992: Latent and sensible heat-flux anomalies over the northern oceans—The connection to monthly atmospheric circulation. J. Climate, 5, 354369, https://doi.org/10.1175/1520-0442(1992)005<0354:LASHFA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ceballos, L. I., E. D. Lorenzo, C. D. Hoyos, N. Schneider, and B. Taguchi, 2009: North Pacific gyre oscillation synchronizes climate fluctuations in the eastern and western boundary systems. J. Climate, 22, 51635174, https://doi.org/10.1175/2009JCLI2848.1.

    • Search Google Scholar
    • Export Citation
  • Che, Z., Y. Yao, and L. Zhong, 2021: Wintertime sea surface temperature front around the Kuroshio Extension and its impact on North Pacific blocking. Int. J. Climatol., 41, 19051926, https://doi.org/10.1002/joc.6937.

    • Search Google Scholar
    • Export Citation
  • Chelton, D. B., and S. P. Xie, 2010: Coupled ocean-atmosphere interaction at oceanic mesoscales. Oceanography, 23, 5269, https://doi.org/10.5670/oceanog.2010.05.

    • Search Google Scholar
    • Export Citation
  • Chelton, D. B., M. G. Schlax, M. H. Freilich, and R. F. Milliff, 2004: Satellite measurements reveal persistent small-scale features in ocean winds. Science, 303, 978983, https://doi.org/10.1126/science.1091901.

    • Search Google Scholar
    • Export Citation
  • Cohen, J. L., J. C. Furtado, M. A. Barlow, V. A. Alexeev, and J. E. Cherry, 2012: Arctic warming, increasing snow cover and widespread boreal winter cooling. Environ. Res. Lett., 7, 014007, https://doi.org/10.1088/1748-9326/7/1/014007.

    • Search Google Scholar
    • Export Citation
  • Craig, A. P., M. Vertenstein, and R. Jacob, 2012: A new flexible coupler for earth system modeling developed for CCSM4 and CESM1. Int. J. High Perform. Comput. Appl., 26, 3142, https://doi.org/10.1177/1094342011428141.

    • Search Google Scholar
    • Export Citation
  • Dai, A., J. C. Fyfe, S. P. Xie, and X. Dai, 2015: Decadal modulation of global surface temperature by internal climate variability. Nat. Climate Change, 5, 555559, https://doi.org/10.1038/nclimate2605.

    • Search Google Scholar
    • Export Citation
  • Danabasoglu, G., and Coauthors, 2020: The Community Earth System Model version 2 (CESM2). J. Adv. Model. Earth Syst., 12, e2019MS001916, https://doi.org/10.1029/2019MS001916.

    • Search Google Scholar
    • Export Citation
  • Di Lorenzo, E., K. M. Cobb, J. Furtado, N. Schneider, B. T. Anderson, A. Bracco, M. Alexander, and D. Vimont, 2010: Central Pacific warming El Niño and decadal climate change in the North Pacific. Nat. Geosci., 3, 762765, https://doi.org/10.1038/ngeo984.

    • Search Google Scholar
    • Export Citation
  • Feliks, Y., M. Ghil, and E. Simonnet, 2004: Low-frequency variability in the midlatitude atmosphere induced by an oceanic thermal front. J. Atmos. Sci., 61, 961981, https://doi.org/10.1175/1520-0469(2004)061<0961:LVITMA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Feliks, Y., M. Ghil, and E. Simonnet, 2007: Low-frequency variability in the midlatitude baroclinic atmosphere induced by an oceanic thermal front. J. Atmos. Sci., 64, 97116, https://doi.org/10.1175/JAS3780.1.

    • Search Google Scholar
    • Export Citation
  • Frankignoul, C., N. Sennechael, Y. O. Kwon, and M. A. Alexander, 2011: Influence of the meridional shifts of the Kuroshio and the Oyashio Extensions on the atmospheric circulation. J. Climate, 24, 762777, https://doi.org/10.1175/2010JCLI3731.1.

    • Search Google Scholar
    • Export Citation
  • Gan, B., and L. Wu, 2012: Modulation of atmospheric response to North Pacific SST anomalies under global warming: A statistical assessment. J. Climate, 25, 65546566, https://doi.org/10.1175/JCLI-D-11-00493.1.

    • Search Google Scholar
    • Export Citation
  • Gan, B., and L. Wu, 2014: Centennial trends in Northern Hemisphere winter storm tracks over the twentieth century. Quart. J. Roy. Meteor. Soc., 140, 19451957, https://doi.org/10.1002/qj.2263.

    • Search Google Scholar
    • Export Citation
  • Gelaro, R., and Coauthors, 2017: The Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2). J. Climate, 30, 54195454, https://doi.org/10.1175/JCLI-D-16-0758.1.

    • Search Google Scholar
    • Export Citation
  • Gettelman, A., and Coauthors, 2019: The Whole Atmosphere Community Climate Model version 6 (WACCM6). J. Geophys. Res. Atmos., 124, 12 38012 403, https://doi.org/10.1029/2019JD030943.

    • Search Google Scholar
    • Export Citation
  • GISTEMP Team, 2022: GISS Surface Temperature Analysis (GISTEMP), version 4. NASA Goddard Institute for Space Studies, accessed 30 April 2022, https://data.giss.nasa.gov/gistemp/.

  • Griffiths, M. L., and R. S. Bradley, 2007: Variations of twentieth-century temperature and precipitation extreme indicators in the northeast United States. J. Climate, 20, 54015417, https://doi.org/10.1175/2007JCLI1594.1.

    • Search Google Scholar
    • Export Citation
  • Harada, Y., and Coauthors, 2016: The JRA-55 Reanalysis: Representation of atmospheric circulation and climate variability. J. Meteor. Soc. Japan, 94, 269302, https://doi.org/10.2151/jmsj.2016-015.

    • Search Google Scholar
    • Export Citation
  • Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 19992049, https://doi.org/10.1002/qj.3803.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B. J., and G. H. White, 1983: The shape, propagation and mean-flow interaction of large-scale weather systems. J. Atmos. Sci., 40, 15951612, https://doi.org/10.1175/1520-0469(1983)040<1595:TSPAMF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B. J., and P. J. Valdes, 1990: On the existence of storm-tracks. J. Atmos. Sci., 47, 18541864, https://doi.org/10.1175/1520-0469(1990)047<1854:OTEOST>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., J. J. Hack, D. Shea, J. M. Caron, and J. Rosinski, 2008: A new sea surface temperature and sea ice boundary dataset for the Community Atmosphere Model. J. Climate, 21, 51455153, https://doi.org/10.1175/2008JCLI2292.1.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471, https://doi.org/10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kim, J. W., S. W. Yeh, and E. C. Chang, 2014: Combined effect of El Niño-Southern Oscillation and Pacific decadal oscillation on the East Asian winter monsoon. Climate Dyn., 42, 957971, https://doi.org/10.1007/s00382-013-1730-z.

    • Search Google Scholar
    • Export Citation
  • Kobayashi, S., and Coauthors, 2015: The JRA-55 Reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 548, https://doi.org/10.2151/jmsj.2015-001.

    • Search Google Scholar
    • Export Citation
  • Kohyama, T., Y. Yamagami, H. Miura, S. Kido, H. Tatebe, and M. Watanabe, 2021: The Gulf Stream and Kuroshio Current are synchronized. Science, 374, 341346, https://doi.org/10.1126/science.abh3295.

    • Search Google Scholar
    • Export Citation
  • Koide, H., and K. Kodera, 1999: A SVD analysis between the winter NH 500-hPa height and surface temperature fields. J. Meteor. Soc. Japan, 77, 4761, https://doi.org/10.2151/jmsj1965.77.1_47.

    • Search Google Scholar
    • Export Citation
  • Kug, J. S., J. H. Jeong, Y. S. Jang, B. M. Kim, C. K. Folland, S. K. Min, and S. W. Son, 2015: Two distinct influences of Arctic warming on cold winters over North America and East Asia. Nat. Geosci., 8, 759762, https://doi.org/10.1038/ngeo2517.

    • Search Google Scholar
    • Export Citation
  • Kushnir, Y., and N. C. Lau, 1992: The general circulation model response to a North Pacific SST anomaly: Dependence on time scale and pattern polarity. J. Climate, 5, 271283, https://doi.org/10.1175/1520-0442(1992)005<0271:TGCMRT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kwon, Y. O., and C. Deser, 2007: North Pacific decadal variability in the Community Climate System Model version 2. J. Climate, 20, 24162433, https://doi.org/10.1175/JCLI4103.1.

    • Search Google Scholar
    • Export Citation
  • Kwon, Y. O., and T. M. Joyce, 2013: Northern Hemisphere winter atmospheric transient eddy heat fluxes and the Gulf Stream and Kuroshio–Oyashio Extension variability. J. Climate, 26, 98399859, https://doi.org/10.1175/JCLI-D-12-00647.1.

    • Search Google Scholar
    • Export Citation
  • Kwon, Y. O., M. A. Alexander, N. A. Bond, C. Frankignoul, H. Nakamura, B. Qiu, and L. A. Thompson, 2010: Role of the Gulf Stream and Kuroshio–Oyashio systems in large-scale atmosphere–ocean interaction: A review. J. Climate, 23, 32493281, https://doi.org/10.1175/2010JCLI3343.1.

    • Search Google Scholar
    • Export Citation
  • Lau, N. C., 1978: On the three-dimensional structure of the observed transient eddy statistics of the Northern Hemisphere wintertime circulation. J. Atmos. Sci., 35, 19001923, https://doi.org/10.1175/1520-0469(1978)035<1900:OTTDSO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lawrence, D. M., and Coauthors, 2019: The Community Land Model version 5: Description of new features, benchmarking, and impact of forcing uncertainty. J. Adv. Model. Earth Syst., 11, 42454287, https://doi.org/10.1029/2018MS001583.

    • Search Google Scholar
    • Export Citation
  • Lei, T., S. Li, F. Luo, and N. Liu, 2020: Two dominant factors governing the decadal cooling anomalies in winter in East China during the global hiatus period. Int. J. Climatol., 40, 750768, https://doi.org/10.1002/joc.6236.

    • Search Google Scholar
    • Export Citation
  • Lenssen, N., G. Schmidt, J. Hansen, M. Menne, A. Persin, R. Ruedy, and D. Zyss, 2019: Improvements in the GISTEMP uncertainty model. J. Geophys. Res. Atmos., 124, 63076326, https://doi.org/10.1029/2018JD029522.

    • Search Google Scholar
    • Export Citation
  • Li, S., and G. T. Bates, 2007: Influence of the Atlantic multidecadal oscillation on the winter climate of East China. Adv. Atmos. Sci., 24, 126135, https://doi.org/10.1007/s00376-007-0126-6.

    • Search Google Scholar
    • Export Citation
  • Lindzen, R. S., and B. Farrell, 1980: A simple approximate result for the maximum growth rate of baroclinic instabilities. J. Atmos. Sci., 37, 16481654, https://doi.org/10.1175/1520-0469(1980)037<1648:ASARFT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lindzen, R. S., and S. Nigam, 1987: On the role of sea surface temperature gradients in forcing low-level winds and convergence in the tropics. J. Atmos. Sci., 44, 24182436, https://doi.org/10.1175/1520-0469(1987)044<2418:OTROSS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Linkin, M. E., and S. Nigam, 2008: The North Pacific Oscillation–West Pacific teleconnection pattern: Mature-phase structure and winter impacts. J. Climate, 21, 19791997, https://doi.org/10.1175/2007JCLI2048.1.