A connection between the tropical Pacific Ocean ... - Wiley Online Library

Journal of Geophysical Research: Atmospheres RESEARCH ARTICLE 10.1002/2014JD022324 Key Points: • There is a connection between the tropical Pacific and the winter climate • The SVD2 influence the winter climate over a large area of NH • The western tropical Pacific forcing contributes to the SVD2

Correspondence to: H. Lin, [email protected]

Citation: Jia, X., S. Wang, H. Lin, and Q. Bao (2015), A connection between the tropical Pacific Ocean and the winter climate in the Asian-Pacific region, J. Geophys. Res. Atmos., 120, 430–448, doi:10.1002/2014JD022324.

Received 26 JUL 2014 Accepted 14 DEC 2014 Accepted article online 17 DEC 2014 Published online 21 JAN 2015

A connection between the tropical Pacific Ocean and the winter climate in the Asian-Pacific region XiaoJing Jia1,2 , Su Wang1 , Hai Lin3 , and Qing Bao2 1 Department of Earth Sciences, ZheJiang University, HangZhou, China, 2 State Key Laboratory of Numerical Modelling for

Atmospheric Sciences and Geophysical Fluid Dynamics, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China, 3 Atmospheric Numerical Weather Prediction Research, Environment Canada, Dorval, Québec, Canada

Abstract The impact of the tropical Pacific sea surface temperature (SST) anomaly on the winter mean surface air temperature (SAT) in the Asian-Pacific region is investigated during the period from 1948 to 2008 using both observations and a linear baroclinic model (LBM). A singular value decomposition (SVD) analysis is conducted between the 500 hPa geopotential height (Z500) over the Northern Hemisphere and the SST over the tropical Pacific Ocean to obtain the large-scale atmospheric patterns related to tropical Pacific SST. Focus is given to the second pair of SVD mode (SVD2) which bears some similarities in the Z500 field to the Arctic Oscillation over the North Atlantic sector and can impact the SAT over a larger area of Asian-Pacific. In the winter of a positive SVD2 the SAT over the midlatitude to high-latitude Asian continent, the Arctic Ocean, the Indian Ocean, and the western subtropical Pacific Ocean tends to be warmer than normal, while the North Pacific Ocean around the Bering Strait is abnormally cold, and vice versa. Examination of the associated surface general circulation shows that a positive SVD2 tends to shift the Siberian High southward and the Aleutian Low eastward resulting in anomalous weak pressure gradient between the Asian continent the North Pacific. Anomalous positive sea level pressure anomalies around Japan and southerly wind along the east coast of the Asian continent are observed. At the same time, the East Asian trough at midtroposphere becomes weaker than normal and the East Asian westerly jet stream is increased in magnitudes and shifted northward. The analysis of the wave activity flux and result of idealized numerical experiments show a possible influence of the western tropical Pacific SST forcing on the SVD2.

1. Introduction The East Asian winter monsoon, which is mainly driven by land-ocean thermal contrast, is one of the most active components of the global climate system in boreal winter [Lau and Li, 1984; Ding, 1994; Huang et al., 2003; Chen et al., 2005; Wang and Chen, 2010; Yang et al., 2010; Wang and Chen, 2014]. Associated with the East Asian winter monsoon are two large-scale surface circulation systems, namely the Siberian High and the Aleutian Low. The Siberian High is centered over Siberia and controls almost the entire region of the continental Asia in winter. The east-west sea level pressure (SLP) gradient between the Siberian High and the Aleutian Low is often used to indicate the variability of the East Asian winter monsoon [Gong et al., 2001; Wang and Chen, 2010]. The strong pressure gradient between the Siberian High and the Aleutian Low favors strong northerly and northeasterly winds over East Asia. These winds penetrate the whole of East Asia and adjacent oceans and reach the subtropical western and central Pacific and the Indo-China Peninsula and influence the climate of underlying regions directly [Lau and Li, 1984; Wang et al., 2000, 2010; Chen et al., 2005; Huang et al., 2012]. The northerly wind brings intense cold air from the high-latitude inner land to lower latitude coastal regions resulting in dramatic changes in the air temperature over East Asia and the surrounding oceans. The East Asian winter monsoon can exert a large social and economic impact on many East Asian countries like China, Korea, Japan, and the surrounding regions. Severe cold surges and associated heavy snowfall have frequently caused serious damages to crops, daily life, and economic activities. The East Asian winter monsoon experiences year-to-year and interdecadal variability resulting from influence of both external forcing and atmospheric teleconnections which make it very difficult to predict [Ding, 1994; Wang et al., 2000; Clark and Mserreze, 2000; Gong et al., 2001; Bao et al., 2010; Chen et al., 2013a; Yang et al., 2013; Wang and Chen, 2014; Wang et al., 2010]. Numerous efforts have been made to understand its mechanisms and to predict its variation [Liu et al., 2014; Takaya and Nakamura, 2013]. JIA ET AL.

©2014. American Geophysical Union. All Rights Reserved.

430

Journal of Geophysical Research: Atmospheres

10.1002/2014JD022324

Except the Siberian High, some other large-scale atmospheric patterns such as the Eurasian pattern, the Western Pacific pattern, and the North Pacific Oscillation may also play notable roles in the climate variability over the East Asian regions [Wu and Huang, 1999; Wang et al., 2000; Wu et al., 2003; Wang et al., 2007; Huang et al., 2012]. Among the atmospheric factors that influence the winter climate over East Asia, the impact of the Arctic Oscillation (AO) on the East Asian winter monsoon has been widely documented [Gong et al., 2001; Chen et al., 2013b]. The AO represents an out-of-phase relation between the SLP in the Arctic basin and that at the midlatitudes. Both observational and model studies showed that the AO is, to a significant degree, an internal mode of variability of the atmospheric circulation [Limpasuvan and Hartmann, 1999; Rodwell et al., 1999; Lin and Derome, 2004]. The influence of the AO on the climate over the North Atlantic is more significant and direct. Some studies also showed that the AO can significantly influence the East Asian winter monsoon through the impact on the Siberian High. Gong et al. [2001] pointed out that there are significant out-of-phase relationships between the AO and the East Asian winter monsoon. They illustrated that the East Asian winter monsoon tends to be weak during the positive phase of AO. Chen et al. [2005] proposed that the AO can influence the climate of East Asia by influencing the propagation of the quasi-stationary planetary wave activity. They found that during the positive phase of the AO, more quasi-stationary planetary waves propagate from high latitudes to lower latitudes in the troposphere, accompanied by weaker than normal Siberian High and Aleutian Low and therefore a weak East Asian winter monsoon. Although many efforts have been made to understand the mechanisms of the variation of the winter monsoon circulation in East Asia, the question of which one of the proposed mechanisms is more essential and how it works still remains. Some recent studies demonstrated that the forecast skill of the Northern Hemisphere extratropical seasonal climate anomalies mainly comes from the climate models ability to capture the leading modes of the SST-related large-scale atmospheric circulation pattern [Jia et al., 2010; Lee et al., 2011]. In a previous study, using both observations and outputs from four atmospheric general circulation models from the second phase of the Historical Forecasting Project for the period from 1969 to 2001, Jia and Lin [2011] demonstrated that the SAT over China is significantly influenced by the leading large-scale atmospheric patterns associated with the tropical Pacific SST forcing. This relationship could be potentially useful to improve the seasonal forecast skill in China. In a following study [Jia et al., 2014], they demonstrated that the time variability of the tropical Pacific SST-related atmospheric patterns and its relationship to SAT over China could be reasonably captured by the multimodel ensemble seasonal forecasts. The mechanisms accounting for the connection between the SAT over China and the tropical Pacific Ocean, however, were not explained. In this study, we further investigate the relationship between the leading atmospheric modes associated with the tropical Pacific SST anomaly and the SAT in wintertime. Focus will be given to the second pair of SVD mode (SVD2) that influence the SAT over China more significant comparing to other leading modes [Jia and Lin, 2011; Jia et al., 2014]. The investigated domain is extended to the whole Asian-Pacific region, and the differences between the SVD2 and the AO are compared. The mechanism behind the connection between the SVD2 and the SAT in the Asian-Pacific region is investigated using both the observations and a linear baroclinic model. The paper is organized as follows: In section 2 the data and the diagnostic techniques used in this study are described. Section 3 gives the tropical Pacific SST-related large-scale atmospheric patterns in the observations. The association between SAT in the Asian-Pacific region and the SVD2 as well as the atmospheric circulations associated with SVD2 is presented in section 4. The connection between the tropical Pacific forcing and the SVD2 is explored in section 5 by examining the SVD2-associated wave activity flux in the observations and by designing idealized numerical experiments using a linear baroclinic model, followed by the conclusion and discussion in section 6.

2. Data and Diagnostic Techniques The reanalyses of SLP, Z500, 200 hPa geopotential height (Z200) SAT, 200 hPa, and 925 hPa wind from the National Centers for Environmental Predictions (NCEP) and the National Center for Atmospheric Research (NCAR) Reanalysis I [Kalnay et al., 1996] during the period from 1948 to 2008 are used in this study. The precipitation data are from the NOAAs Precipitation Reconstruction Data Set [Chen et al., 2002]. The satellite-observed outgoing longwave radiation (OLR) data are from the National Oceanic and Atmospheric Administration (NOAA) polar-orbiting series of satellites (Liebmann and Smith 1996). The SST is Hadley Centre Sea Ice and Sea Surface Temperature data set (HadISST) 1.1 which is a unique combination JIA ET AL.


431


10.1002/2014JD022324

of monthly globally complete fields of SST and sea ice on a 1◦ latitude-longitude grid [Rayner et al., 2003]. In this study, we focus on winter seasonal means that are constructed by averaging the monthly means of December-January-February (DJF). Here the winter of 1948 refers to the boreal 1948/1949 winter. It is known that the NCEP-1 data may have systematic errors in the period before 1980 [Wu et al., 2005]. To validate the results presented in this study, another set of data is obtained by combining the ERA-40 and NCEP-1 data. To get a longer time length data set, the ERA-40 data for the period 1957–1979 and the NCEP-1 data from 1980 to 2010 are combined together. To maintain temporal homogeneity, the 1980–2010 NCEP-1 data were adjusted by removing the climatological difference between the ERA-40 and NCEP-1 data. It was found that the results obtained using the combined data are generally consistent with those obtained from the NCEP-1 data. Thus, only the results by use of the NCEP-1 data are presented in this paper. In order to explore the mechanisms behind the connection between the tropical Pacific Ocean and the SAT in the Asian-Pacific region, a phase-independent flux of wave activity, defined for stationary Rossby waves on a zonally asymmetric climatological-mean flow by Takaya and Nakamura [2001], was computed. The wave activity flux is parallel to the local group velocity in the Wentzel-Kramers-Brillouin sense and is a useful diagnostic tool in identifying the sources or sinks of the wave activity. In this study, we only use the horizontal component of this flux. Following Takaya and Nakamura [2001], the horizontal flux vector is given by ( ) ( ′2 ) ′ ′ ′ ⎛ ⎞ + V 𝜓x′ 𝜓y′ − 𝜓 ′ 𝜓xy 1 ⎜ U (𝜓x − 𝜓 𝜓xx ) ( )⎟ W= 2|U| ⎜ U 𝜓 ′ 𝜓 ′ − 𝜓 ′ 𝜓 ′ + V 𝜓 ′2 − 𝜓 ′ 𝜓 ′ ⎟ x y xy y yy ⎠ ⎝ where U = (U, V) is the two-dimensional geostrophic zonal and meridional velocity components of the basic state and 𝜓 is the stream function. The primes represent departures from the climatological mean and the subscripts denote partial derivatives.

3. Tropical Pacific SST-Related Large-Scale Atmospheric Patterns To get the large-scale atmospheric patterns related to the tropical Pacific SST forcing, following previous studies [Lin et al., 2008; Jia et al., 2010; Jia and Lin, 2011; Jia et al., 2014], an SVD analysis is performed between the DJF seasonal averaged NCEP-NCAR Z500 over the Northern Hemisphere north of 20◦ N and the SST of the same season over the tropical Pacific (20◦ N–20◦ S, 120◦ E–90◦ W). The time series of the atmospheric and oceanic components of the first two pair of SVD modes are then regressed onto the global Z500 and SST fields and are illustrated in Figure 1. The atmospheric patterns of the leading two SVDs, depicted in Figures 1a and 1b, represent the dominant atmospheric patterns associated with the SST forcing in the tropical Pacific, which would represent the dominant source of seasonal forecast skill in the Northern Hemisphere, especially in wintertime [Lin et al., 2005, 2008; Jia et al., 2010; Lee et al., 2011; Jia et al., 2014]. The first pair of SVD mode (Figures 1a and 1c) explains 80% of the total covariance between the Z500 and SST fields during the period under examination based on a squared covariance fraction. The atmospheric component of the first SVD mode (Figure 1a) is characterized by weak positive Z500 anomalies in the subtropical Pacific and over the northeastern North American region together with pronounced negative Z500 anomalies over the North Pacific Ocean and southeast America, resembling the traditional Pacific North American (PNA) pattern. The associated oceanic component of the first SVD mode (Figure 1c) ˜ structure with positive SST anomalies dominating in the eastern tropical Pacific represents a typical El Nino Ocean. The temporal correlation coefficient (TCC) between the expansion coefficients of the atmospheric and the oceanic component of the leading SVD is 0.73, exceeding a confidence level of 99% according to a Student’s t test. The TCC between the time series associated with the oceanic component of the first SVD mode and the Nino3.4 index obtained from the NOAA Climate Prediction Center which is the average of the SST anomalies over the eastern central tropical Pacific (5◦ N–5◦ S, 170◦ W–120◦ W) is 0.98, far beyond the confidence level of 99%. The second pair of SVD mode (Figures 1b and 1d) explains 10% of the total covariance between the Z500 and SST fields. The atmospheric component of the SVD2 (ASVD2) (Figure 1b) shows pronounced negative Z500 anomalies centered Greenland and another relatively weak one over the high-latitude North Pacific around the Bering Strait. Positive Z500 anomalies are observed along the middle latitude North Atlantic and North Pacific. The spatial structure of the ASVD2 bears some similarities to the AO over the JIA ET AL.


432


10.1002/2014JD022324

Figure 1. (a) Observed DJF mean, (b) 500 hPa height, and (c, d) SST distributions of the leading two SVD modes. The contour interval is 5 m in Figures 1a and 1b, 0.15◦ C in Figure 1c, and 0.1◦ C in Figure 1d. Negative values are shaded in Figure 1d. The magnitudes of Z500 and SST correspond to 1 standard deviation of their respective expansion coefficients.

North Atlantic sector. However, notable difference can be observed between Figure 1b and AO over the North Pacific region. The SST component of the SVD2 is dominated by negative SST anomalies over the tropical Pacific with maximum values appearing over the central equatorial Pacific Ocean, consistent with previous studies [Lin et al., 2005; Jia et al., 2010; Jia and Lin, 2011]. Weak SST anomalies of a triple pattern with alternative signs over the North Atlantic are also observed. In a previous study, Jia and Lin [2011] examined the seasonality of the influence of the tropical Pacific SST-related large-scale atmospheric patterns on the SAT over China. Results showed that the wintertime SAT over China can be significantly influenced by the SVD2. The first SVD mode impacts more on the SAT over the North Pacific North America regions than JIA ET AL.


433


10.1002/2014JD022324

the East Asia. Here in the following, focus will only be given to the SVD2 and we explore its impact to the SAT in the Asian-Pacific region and investigate the mechanisms behind it.

3

2

1

The TCC between the time series associated with the ASVD2 (APC2, 0 hereafter) (thick solid line in Figure 2) and the time series associated with -1 the SST field (dashed line in Figure 2) is 0.73, with a significance exceeding -2 a level of 99%. Figure 2 shows that both time series display interannual and interdecadal fluctuations. There -3 1950 1956 1962 1968 1974 1980 1986 1992 1998 2004 is an obvious transition of these two Year indices at around 1980s with negative Figure 2. The time series associated with the atmospheric component phase dominating before and (thick solid line) and the oceanic component (thin dashed line) of the SVD2. positive phase dominating after 1980s, consistent with many previous studies [e.g., Wang and Chen, 2010]. Since the spatial structure of the ASVD2 is similar to the AO, the TCC between APC2 and the AO index is examined. Following Thompson and Wallace [1998], an empirical orthogonal function (EOF) analysis is conducted for the DJF mean SLP poleward of 20◦ N using the NCEP-NCAR reanalysis. The AO appears to be the first EOF, explaining 29% of the total variance of SLP during this period. The corresponding principal component of the leading EOF is regarded as the AO index (not shown). The AO index obtained here and the AO index obtained from http://www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/ao. shtml are 0.95 during the period from 1950 to 2011, far beyond the 99% confidence level. The APC2 is found highly correlated to the AO index with a correlation of 0.67. The correlation between the APC2 and AO index is 0.58 after the linear trend is removed from both time series, still significant at the 95% confidence level according to a Student’s t test.

4. Climate Anomalies Represented by the ASVD2 4.1. The Association Between the ASVD2 and the SAT in the Asian-Pacific Region To assess the association between the ASVD2 and the wintertime SAT in the Asian-Pacific region, the regression of the SAT onto the APC2 is calculated and depicted in Figure 3a. Figure 3a shows significant positive SAT anomalies over a large area of the midlatitude to high-latitude Asian continent, the Arctic Ocean, the Indian Ocean, and the western subtropical Pacific Ocean. Negative SAT anomalies are observed over the high-latitude North Pacific centered the Bering Strait. The regression of the SAT onto the AO index is also examined and presented in Figure 3b for the purpose of a comparison. The regression maps of Figures 3a and 3b are similar to each other, as expected, considering the high correlation between the APC2 and the AO index (0.67), whereas differences can also be noticed over many areas. For example, the SAT anomalies over the midlatitude to high-latitude East Asia and the east coastal regions of the Asian continent associated with AO are obviously more pronounced than the APC2. However, wider significant areas can be observed on the regression map of the APC2 over oceans, for example, the Indian Ocean, the subtropical western Pacific, and the Polar region. The above results suggest that the AO is mainly related to the SAT variation over the Asian continent, while the ASVD2 can capture the variability of the SAT over a broader area in the Asian-Pacific region. Figure 2 shows that there are both interannual and long-term trends of the APC2. To examine the impact of the long-term trend on the relationship between the ASVD2 and the wintertime SAT in the Asian-Pacific region, the regression of the SAT onto the detrended APC2 is also examined (not shown). It shows that the significant correlation areas and the magnitudes are obviously decreased comparing to Figure 3a. However, there are still quite large areas of significant correlations, such as the high-latitude EA, the Maritime continent, and the subtropical and high-latitude North Pacific, which are independent of the trend. JIA ET AL.


434


10.1002/2014JD022324

Figure 3. The regression of the NCEP-NCAR SAT to (a) APC2 and (b) AO index. Contours with negative values are dashed. The contour interval is 0.2◦ C. The light- and dark-shaded areas represent correlations exceeding a confidence level of 95% and 99% according to a Student’s t test.

As shown in Figure 1d, the SST anomalies associated with the SVD2 over the tropical Pacific are relatively weak with maximum values of about 0.2◦ C; to have an idea of the extent to which the SVD2-related SST forcing can impact the wintertime SAT in the Asian-Pacific region, the regression of the normalized SAT anomalies (the anomalous SAT divided by their corresponding SAT standard deviation) onto the APC2 and the AO index is examined and presented as percentage in Figure 4, where the absolute values larger than 20% are shaded. Figure 4b shows that with 1 standard deviation of the AO index, the associated SAT change is about 20% to 70% of its SAT standard deviation over the midlatitude to high-latitude Asian continent. The SAT change associated with the APC2 is less pronounced than the AO in these regions which is about 20% to 40% of the SAT standard deviation. However, the SAT changes over the Indian Ocean, the subtropica l western Pacific, and the Polar region are more pronounced and related to the APC2 than the AO and account for about 20% to 40% of their local SAT standard deviation. It is known that the Siberian High is the most significant surface system that influences the winter climate in East Asia. Previous studies showed that the Siberian High has a close relationship to the AO, and the AO can influence the winter climate over East Asia through the impact on the Siberian High [e.g., Gong et al., 2001]. Following Gong et al. [2001], the Siberian High index is defined as the regionally averaged winter SLP JIA ET AL.


435


10.1002/2014JD022324

Figure 4. The ratio of the regression of SAT to the standard deviation of the SAT for (a) APC2 and (b) AO index. Areas with absolute values greater than 20% are shaded.

anomaly over the region 40◦ N–60◦ N, 70◦ E–120◦ E. The sign of the Siberian High index is reversed to facilitate the comparison. The TCC between the AO index and the Siberian High index is found to be 0.40, exceeding a confidence level of 95% according to a Student’s t test. As the APC2 and AO index are highly correlated to each other (with a TCC of 0.64), the TCC between the Siberian High index and the APC2 is also examined. Result shows that the TCC between the two indices is almost zero (0.05), suggesting that the ASVD2 and the Siberian High are two independent patterns of atmospheric variability. The above results imply that although the ASVD2 and the AO share many similarities in both spatial structures and their associated time variation, obvious differences between them also exist. 4.2. The General Atmospheric Circulation Anomalies Associated With the ASVD2 To understand how the ASVD2 impact the wintertime SAT in the Asian-Pacific region as illustrated above, we examine the large-scale atmospheric circulation anomalies related to the ASVD2. The anomalous SLP regressed onto the APC2 is presented in Figure 5b along with the SLP climatology shown in Figure 5a. Figure 5a shows that the surface circulation over the Asian continent is dominated by the pronounced atmospheric center of action, namely the Siberian High. This anticyclonic circulation system is centered over JIA ET AL.


436


10.1002/2014JD022324

Figure 5. The climatological winter mean of (a) SLP and (c) 500 hPa geopotential height and the regression to APC2 for (b) SLP and (d) 500 hPa geopotential height. The contour interval is 5 hPa for Figure 5a, 100 m for Figure 5c, 0.5 hPa for Figure 5b, and 5 m for Figure 5d. The light- and dark-shaded areas in Figures 5b and 5c represent correlations exceeding a confidence level of 95% and 99% according to a Student’s t test.

the inner territory at around 50◦ N, 100◦ E and controls almost the entire Asian continent. The Aleutian Low is seen centered in the North Pacific Ocean at around 50◦ N, 175◦ E. Over the North Pacific, the most prominent characteristics of the SLP anomalies associated with a positive phase of ASVD2 is a meridional dipole structure with anomaly centers of opposite sign. Pronounced negative SLP anomalies are noticed at high-latitude North Pacific around east of the Bering Strait (Figure 5b). Compared with the climatology, the negative SLP anomalies reflect an eastward shifted Aleutian Low pressure system. Positive SLP anomalies are observed in the middle latitude east Pacific south of the Aleutian Low, extending southwestward from the West Coast of North America toward the middle subtropical North Pacific. Over the Asian continent, negative SLP anomalies are observed north of 50◦ N, centered 70◦ N, 10◦ E, suggesting a northwestward shift of the Siberian High. To better understand the relationship between the ASVD2, the Siberian High and the AO systems, the SLP anomalies associated with the APC2, the Siberian High index, and the AO index in the whole Northern Hemisphere are depicted in Figure 6. Keep in mind that the Siberian High index is reversed in sign to facilitate the comparison. The SLP associated with the negative Siberian High index (Figure 6b) is dominated by pronounced negative anomalies prevailing over the Asian continent. Weak but significant negative and positive SLP anomalies can also be noticed over Greenland and the midlatitude North Atlantic, respectively. The spatial structure of Figure 6b in the North Atlantic is in a form of a positive AO distribution which is not surprising considering the high correlation between the Siberian High (SH) and AO indices (0.40). Obvious differences can be noticed between the APC2-related SLP anomalies (Figure 6a) and Figure 6b. The significant SLP dipole structure over the North Pacific appearing in Figure 6a is not observed in Figure 6b, whereas the SLP anomalies related to the APC2 is much weaker over the East Asia than the SH-related SLP anomalies. For the SLP anomalies associated with the APC2, another pronounced dipole of SLP anomaly pattern prevails in the North Atlantic with a meridional seesaw between the high and middle latitudes JIA ET AL.


437


10.1002/2014JD022324

Figure 6. The regression of SLP onto (a) APC2, (b) SH index, and (c) AO index. The light- and dark-shaded areas represent correlations exceeding a confidence level of 95% and 99% according to a Student’s t test. The contour interval is 0.5 hPa.

(Figure 6a). It is clear that the most significant SLP signals associated with the ASVD2 are over the North Pacific and the North Atlantic sectors, while that over the East Asia is quite weak. The spatial structure of the SLP anomalies in Figure 6a is strikingly similar to the AO-related SLP anomalies as shown in Figure 6c, especially over the North Atlantic. However, differences can also be noticed between Figure 6a and Figure 6c. For example, the dipole structure of the SLP anomalies in the North Pacific related to the ASVD2 is not seen in Figure 6c The negative SLP anomalies in the polar region associated with the AO are more pronounced comparing to the ASVD2 and extend much further eastward to the Asian continent. Over the midlatitude to high-latitude Asian continent, there are some overlap of the SLP anomalies between Figures 6b and 6c that can explain the high correlation between the SH and AO indices. The SLP regression results shown in Figure 6 suggest that although the APC2 and the AO index are highly correlated to each other there are obvious differences between their related large-scale circulations, especially over the North Pacific and midlatitude to high-latitude Asian continent. The surface circulation associated with the APC2 is further examined by calculating the regression of the wind at 925 hPa (shown in Figure 7 as vector) along with the SAT (shaded). The APC2-related SLP is also overlapped in Figure 7 as contour. As we mentioned before, the SLP gradient between the Siberian High JIA ET AL.


438


10.1002/2014JD022324

Figure 7. The regression of wind at 925 hPa (vector), the SLP (contour), and SAT (shaded) onto the APC2. The contour intervals are 0.2◦ C for the SAT and 0.5 hPa for the SLP. Scaling for arrows is given at the bottom.

and the Aleutian Low favors a northerly wind blowing from high latitude to lower latitude along the east coast of the Asian continent and can influence the climate of underlying regions. However, a positive SVD2 tends to shift the Siberian High southward and the Aleutian Low eastward resulting in anomalous weak pressure gradient between the Asian continent the North Pacific. Anomalous positive SLP anomalies around Japan and southerly wind along the east coast of the Asian continent are observed in Figure 7. Over the North Pacific, there is anomalous northerly wind along the west flank of the negative SLP anomalies near the Bering Strait accompanied by increased advection of cold air from the polar region which can decrease the temperature nearby. In the subtropical North Pacific between 10◦ N and 30◦ N there is an anomalies easterly surface wind associated with the positive SLP anomalies of the dipole. This easterly blows all the way across the North Pacific and penetrates deeply into the interior of the Asian continent. This easterly wind transports warm air to the west subtropical Pacific, the Indian Ocean, and the Asian continent and contribute to the warm SAT anomalies over there (Figure 3a). Also, noticed that there is a divergence anomaly along the equatorial Pacific centered 140◦ W. The easterly wind on the west flank of the divergence also contributes to the subtropical North Pacific easterly wind anomaly. A cyclonic system is observed lying over the western tropical Pacific centered 120◦ E. It will be shown later that this cyclonic system is associated with positive precipitation anomalies over the Maritime continent. In the middle troposphere, a climatological East Asian trough centered along the longitudes of Japan is observed in Figure 5c. The linear regression coefficient between the APC2 and Z500 is depicted in Figure 5d. The dipole structure over the North Pacific is even more clearly seen on this level. The positive Z500 anomalies of the dipole over the eastern North Pacific extend westward to the Asian continent, suggesting a weaker than normal East Asian trough in a winter of positive APC2. We then looked at the upper zonal wind near the jet level at 200 hPa (U200) (Figure 8). Figure 8a shows the climatology of zonal wind at 200 hPa where the dominant feature is the East Asia jet stream with its maximum wind speed exceeding 75 m/s located just to the southeast of Japan. This jet stream is closely related to intensive baroclinicity, large vertical wind shear, and advection of cold air [e.g., Lau and Chang, 1987; Ding, 1994]. Tropical upper easterlies are found south of 10◦ N with its maximum around Philippines. To see the change of U200 associated with the APC2, regression of the normalized U200 to the APC2 is presented in Figure 8b. The U200 at each grid point is normalized by its standard deviation and then regressed with respect to the APC2. The westerly anomaly between 30◦ N and 40◦ N and the easterly anomaly between 10◦ N and 25◦ N over the JIA ET AL.


439


10.1002/2014JD022324

Figure 8. (a) The climatological winter mean of 200 hPa zonal wind and (b) the regression of the normalized 200 hPa zonal wind onto APC2. The shaded areas represent negative value in Figure 8a. The light- and dark-shaded areas in Figure 8b represent correlations exceeding a confidence level of 95% and 99% according to a Student’s t test. The contour interval is 10 m for Figure 8a and 0.1 for Figure 8b.

coast of Asian continent indicate a northward shift of the East Asia jet stream. With 1 standard deviation of the APC2 change, the westerly jet strength increased by more than 40% of its local standard deviation around the center of the jet stream. Figure 8 shows that the East Asia jet stream is significantly increased in magnitudes and shifts northward in a winter of a positive SVD2. To see how the storm track reacts to a positive SVD2, the synoptic eddy activity anomalies related to the SVD2 is examined. The synoptic eddy activity is defined as the variance of band-pass-filtered daily mean meridional wind at 300 hPa, calculated for every winter season. The data have been subjected to a band-pass time filtering, and then only periods from 2 to 7 days are retained in the data. Figure 9 shows that corresponding to a positive SVD2, the synoptic-scale eddy activity over the Asian continent and surrounding seas between 30◦ N and 50◦ N are significantly increased, consistent with Figure 8b, suggesting a northward shift and eastward extension of the storm track. JIA ET AL.


440


10.1002/2014JD022324

The atmospheric circulation anomalies related to the detrended ASVD2 are also examined (not shown). It shows that they are generally consistent with those shown in Figures 5, 7, and 8. The results suggest that the relationship between the ASVD2 and the winter time SAT in the Asian-Pacific region is not random but geophysically meaningful.

Figure 9. The regression of the synoptic eddy activities defined as the variance of band-pass-filtered daily mean meridional wind at 300 hPa and calculated for every winter season onto the APC2. The contour interval is 2 m2 /s2 . The light- and dark-shaded areas represent correlations exceeding a confidence level of 95% and 99% according to a Student’s t test. The dashed lines represent the winter mean 200 hPa zonal wind exceeding 60 m/s and 70 m/s, respectively.

5. The Connection Between the Tropical Pacific Forcing and the SVD2

5.1. The SVD2 Associated Precipitation and Wave Activity Anomalies In this study, the SVD analysis is applied to the Z500 in the Northern Hemisphere and the SST in the tropical Pacific Ocean. The leading SVD modes obtained represent the most pronounced atmospheric patterns that are coupled with the tropical Pacific SST. Above results show that the ASVD2 can significantly influence the SAT in the Asian-Pacific region by changing the large-scale atmospheric circulations over the Asian-Pacific sector. The atmospheric component of the SVD2, i.e., ASVD2, is represented by a AO-like pattern, while its associated SST anomalies are dominated by negative SST anomalies over the tropical Pacific. To further understand the connection between the tropical Pacific forcing and the atmospheric response over the Northern Hemisphere, the regression of the observed precipitation onto the APC2 is presented in Figure 10a. Although the SST anomalies associated with the ASVD2 are global in scale, with signals also appearing over the extratropical North Pacific and the North Atlantic sectors (Figure 1d), the precipitation anomalies associated with the ASVD2 are confined to the tropics (Figure 10a). The precipitation anomalies associated with a positive APC2 is characterized by a pronounced east-west dipole with a dry anomaly centered the dateline and a positive precipitation anomaly located over the Maritime continent. The positive precipitation anomaly is consistent with the cyclonic system around the Maritime continent, while the negative precipitation anomaly is consistent with the divergence anomaly along the equatorial Pacific over the dateline as shown in Figure 8. The regression of the OLR onto the APC2 is also examined and presented in Figure 10b for a purpose of confirmation. The OLR data used here are only for the period from 1979 to 2008. Although the data period between the precipitation and the OLR is different, the main OLR features are very similar to the precipitation regression map just with an opposite sign. To explore how the tropical Pacific forcing (Figure 10) contribute to the variability of the ASVD2 over the midlatitude to high-latitude Northern Hemisphere, the wave activity flux as discussed in section 2 is calculated. It is known that the wave activity flux is a good diagnostic tool to understand the source/sink of the atmospheric motion and variability. In the wave activity flux expression, 𝜓 ′ is the stream function of the regression at 200 hPa of the geopotential height against APC2, and U, V is the velocity component of the climatological mean. Figure 11 shows the horizontal stationary wave activity at 200 hPa. Also shown on the map as contour is the regression of Z200 onto the APC2. The spatial structure of Z200 is consistent with that of SLP (Figure 5b) and Z500 (Figure 5d), indicating an equivalent barotropic vertical structure of the ASVD2. Along the coast of Asian continent, a branch of wave activity flux is clearly seen, originating from the western subtropical Pacific region between 120◦ E and 150◦ E. It points northward along the coast, penetrates northward till 70◦ N, and then turns eastward. This wave activity flux split into two branches with one branch turns southeastward toward the equatorial Pacific between 180◦ E and 150◦ W and the other branch of flux penetrates farther eastward, across the North American continent, and reaches North Atlantic Ocean. The distribution of the wave activity flux suggests a possible energy source of the ASVD2 from the western tropical Pacific region. JIA ET AL.


441


10.1002/2014JD022324

Figure 10. Regression of (a) precipitation and (b) OLR onto the APC2. The light- and dark-shaded areas represent correlations exceeding a confidence level of 95% and 99% according to a Student’s t test. The unit is mm/d and W/m2 for Figures 10a and 10b, respectively.

5.2. Linear Atmospheric Response Experiments Figure 10 shows that the SVD2-related precipitation anomalies are concentrated over the western tropical Pacific. Also, the distribution of the wave activity flux indicates that the ASVD2-related wave activity originates from the western subtropical Pacific region (Figure 11). Inspired by above results, idealized thermal forcings are designed over the tropical Pacific and a numerical model is used to examine how the midlatitude to high-latitude atmosphere response to the remote tropical Pacific forcings. To diagnose the atmospheric response to a specified heating, we use a spectral baroclinic model based on primitive equations linearized about the observed winter climatology derived from NCEP/NCAR reanalysis. The linear baroclinic model (LBM) is described in Watanabe and Kimoto [2000]. It has 20 sigma levels

Figure 11. The 200 hPa wave activity associated with APC2. The arrows are the horizontal (W vector), and the regression of the geopotential height at 200 hPa to normalized time series associated with APC2 is overlaid. Scaling for arrows is given at the bottom (unit: m2 s−2 ). The light- and dark-shaded areas represent correlations exceeding a confidence level of 95% and 99% according to a Student’s t test.

JIA ET AL.


442


10.1002/2014JD022324

Figure 12. The (a)vertical and (b) horizontal structure of the idealized thermal forcings.

with horizontal resolution of T42. The model employs Del-forth horizontal diffusion, Raleigh friction, and Newtonian thermal damping with an e-folding scale at 1 day in the lower boundary layer and the uppermost two levels, and at 30 days elsewhere. The LBM is forced with externally imposed heating and integrated toward a steady state. As a guide in developing idealized thermal forcing pattern that is associated with the ASVD2, we use the regression of the observed precipitation anomalies onto the APC2 (Figure 10a). Two-idealized thermal forcings are designed and set over the western tropical Pacific, and the atmospheric responses to these two forcings are examined separately. The horizontal spatial structures of the forcings are shown in Figure 12b. A negative forcing is seen centered the dateline along the equatorial Pacific and a positive one over the Maritime continent centered at 10◦ N, 120◦ E. The forcing has an elliptical horizontal shape with semimajor and semiminor axes of 40◦ longitude and 12◦ latitude for the negative forcing and of 30◦ longitude and 15◦ latitude for the positive forcing, respectively. The magnitude of the forcing is a function of the squared cosine of the distance from the center. The vertical structure of the forcing is shown in Figure 12a. The forcing peaks at 𝜎 = 0.45, and the vertical average at the center is 3.5◦ C/d, corresponding to a precipitation anomaly of about 2 cm/d. The positive and the negative idealized thermal forcings have the same vertical structure while with opposite signs. The idealized thermal forcing anomaly is added to the climatological forcing to drive the model. The forcing is switched on at the beginning and persisted throughout the whole integration period. The atmospheric response of Z200 to the idealized tropical Pacific thermal forcing is shown in Figure 13 from day 2 to day 14 with a 4 day interval for the positive forcing (Figures 13a–13d) and the negative forcing (Figures 13e–13h), respectively. The pattern of the response was found to be quite stable after 2 weeks, with a gradual increase in amplitude. For the positive forcing around the Maritime continent, JIA ET AL.


443


10.1002/2014JD022324

Figure 13. The atmospheric linear response of Z200 to the (a–d) positive thermal forcing and the (e–h) negative thermal forcing from day 2 to day 14 displayed at 4 day interval.

it shows that a positive Z200 anomaly response first appears at day 2 over the western subtropical Pacific Ocean. The magnitude of the positive Z200 anomaly is seen increased on day 6 (Figure 13b), and a negative Z200 anomaly appears around the Ocean of east of Japan followed by another positive Z200 anomaly downstream of it. Another negative Z200 anomaly center develops over the northeast North America at day 10. Energy dispersion is implied by the development of the wave train atmospheric response along the coast of Asian continent, cross the North Pacific, and reaches North America. A comparison between Figures 11 and 13d shows many similarities between them. For example, the observed western subtropical Pacific high and the low-pressure anomaly northward of it, the positive Z200 anomaly over the eastern Pacific and the negative Z200 anomaly downstream of it can also be seen in the model atmospheric response (Figure 13d), suggesting the importance of the linear dynamics. Obvious differences can also be found in the midlatitude and high-latitude North Pacific. The negative Z200 anomaly centered the Bering Strait is much weaker and located more southward in Figure 13d comparing to its counterpart in the observations. The pronounced positive Z200 anomalies in the model atmosphere over the high-latitude North Pacific is not seen in the observations, indicating the roles of nonlinear processes in the generation of the ASVD2 over there. The model atmospheric response to the negative forcing over the dateline along the equatorial Pacific shows a negative Z200 anomaly response at day 2 around the dateline. Energy dispersion is seen through the propagation of the wave train pattern across the North Pacific and North American region in the JIA ET AL.


444


10.1002/2014JD022324

following days. A clear negative PNA pattern response is noticed on day 14 (Figure 13g). The result of the time evolution of the model atmospheric direct response to the idealized thermal forcing indicates that the positive forcing around the Maritime continent is more important for the generation of the ASVD2 comparing to the negative forcing along the equatorial Pacific. The atmospheric response to the combination of the above two-thermal forcings is also examined, and a wave train atmospheric response along the coast of Asian continent is also clearly observed (not shown). A possible link between the tropical Pacific Ocean and the AO has been suggested in many previous studies [Zhou and Miller, 2005; Li et al., 2006; Jia et al., 2009; Gong et al., 2011, 2014]. Li et al. [2006] regressed the NCEP/NCAR precipitation rate against the AO time series and found that the negative polarity of the AO is associated with a positive western tropical Pacific rainfall anomaly centered near 160◦ E, while the rainfall over the tropical Indian Ocean and the western Atlantic Ocean is suppressed. Zhou and Miller [2005] also showed the composite of the pentad-averaged OLR associated with the AO. A high phase of the AO was found to be related to a negative OLR anomaly at the equatorial Indo-Pacific Ocean and a positive OLR anomaly over the western tropical Pacific Ocean, consistent with Gong et al. [2014]. In a previous study, using a simple atmospheric model, Jia et al. [2009] examined the atmospheric response to the thermal forcing over different locations of equatorial Pacific. Results indicate that the negative thermal forcing over the western tropical Pacific Ocean is effective in generating an AO-like atmospheric response while those over the central tropical Pacific is favorable for a negative PNA-like atmospheric response. Here in this study, we showed that, the ASVD2, a tropical Pacific SST-related atmospheric pattern, which is an AO-like mode, is also partly related to some energy source from the western tropical Pacific Ocean. The western tropical Pacific SST forcing can cause midlatitude to high-latitude AO-like atmospheric anomalies, which exert significant impact on the winter SAT in the Asian-Pacific region.

6. Summary and Discussion A previous study demonstrated that the winter time SAT over China can be significantly influenced by the leading atmospheric patterns associated with tropical Pacific SST anomalies [Jia and Lin, 2011]. This relationship could be potentially used to improve the seasonal forecast skill in China [Jia et al., 2014]. In these studies focus was given to the SST anomalies in the tropical Pacific, which is known to be a major forcing area of the atmospheric variability on a seasonal time scale. The mechanisms, however, accounting for the connection between the winter SAT and the tropical Pacific SST anomalies are not clear. In this study the influence of the tropical Pacific SST forcing on the winter time SAT in the Asian-Pacific region is further examined and the mechanism behind the connection between the SVD2 and the SAT in the Asian-Pacific region is investigated using both observations and a linear baroclinic model. Following the previous studies, the tropical Pacific SST-related large-scale atmospheric patterns were obtained using an SVD analysis conducted between the Z500 over the Northern Hemisphere and the SST in the tropical Pacific Ocean. In this study, focus is given to the second pair of SVD mode which has a close relationship to the SAT in the Asian-Pacific region in winter. The spatial structure of the atmospheric component of the SVD2 is characterized by an AO-like pattern, and its related oceanic component is represented by negative SST anomalies over the tropical Pacific. In the winter of a positive SVD2, the SAT over a large areas of midlatitude to high-latitude Asian continent, the Arctic Ocean, the Indian Ocean, and the western subtropical Pacific Ocean tends to be warmer than normal, while the SAT over the North Pacific Ocean around the Bering Strait is abnormally cold, and vice versa. Comparing to the AO, the ASVD2 can impact the SAT over a much wider region in the Asia-Pacific regions, especially over oceans. To understand the mechanisms of how the ASVD2 influence the SAT in the Asian-Pacific region, the large-scale atmospheric circulation associated with the APC2 is examined. At surface, the positive SVD2 tends to shift the Siberian High southward and the Aleutian low eastward resulting in a weaker-than-normal pressure gradient between the Asian continent and the North Pacific. Anomalous positive SLP anomalies around Japan and southerly wind along the east coast of the Asian continent are observed in a positive SVD2 year. Over the North Pacific there is a pronounced barotropic meridional dipole extending from the surface to the tropopause level. The significant cyclone near the Bering Strait is associated with anomalous northerly surface wind along its west flank and accompanied by increased advection of cold air from the polar region which contributes to the cold temperature anomalies over there. Meanwhile, the anomalous easterly surface wind along the south flank of the high-pressure system over the middle east Pacific transports warm air to the western subtropical Pacific, the Indian Ocean, and the interior of the Asian JIA ET AL.


445


10.1002/2014JD022324

continent, and makes the temperature abnormally warmer. In the middle troposphere, the positive Z500 anomalies over the extratropical Pacific can extend westward to the Asian continent, indicating a weaker-than-normal East Asian Trough. At the same time, the zonal wind near the jet level is characterized by a significant northward shifted East Asia jet stream and accompanied by increased transient activities between 30◦ N and 50◦ N. Figure 14. The regression of the NCEP-NCAR SAT to the APC2 which has been subjected to a band-pass time filtering with a period of 2 to 9 years APC2. Contours with negative values are dashed. The light- and dark-shaded areas represent correlations exceeding a confidence level of 95% and 99% according to a Student’s t test.

Examination of the wave activity flux related to the ASVD2 shows that a branch of wave activity flux originates from the western subtropical Pacific Ocean. It flows northward along the coast of the Asian continent ◦ and then turns eastward at around 70 N. This wave activity flux split into two branches with one branch turns southeastward toward the equatorial Pacific, while the other branch of flux penetrates further eastward across the North American continent and reaches North Atlantic Ocean. The result of the wave activity flux indicates a possible source of energy of the ASVD2 coming from the western tropical Pacific Ocean and a possible influence of the North Pacific to the North Atlantic Ocean. The regression of the precipitation onto the APC2 indicates that a positive ASVD2 is associated with an east-west dipole structure of precipitation anomalies over the tropical Pacific with negative precipitation anomalies centered the dateline and positive precipitation anomalies around the Maritime continent. The LBM is further used to help understand the complicated sequence of feedback in the dynamical atmosphere, by removing nonlinearity in their processes. Two-idealized thermal forcings are designed based on the regression of the precipitation onto the APC2. The model atmospheric response to the positive forcing over the Maritime continent shows clear wave train pattern in the North Pacific sector, consistent with the observations, indicating the importance of the linear dynamics. Obvious differences are found over the midlatitude and high-latitude North Pacific between the atmospheric response and the observations, suggesting the roles evolved by nonlinear processes in the generation of ASVD2 over there. Figure 2 shows that the APC2 displays both interannual and interdecadal variations during the period under examination. To examine the relative importance of the interannual and interdecadal components of the variation to the relationship between the ASVD2 and the wintertime SAT in the Asian-Pacific region, the APC2 has been subjected to a band-pass time filtering with a period of 2–9 years to exclude the interdecadal variations. This reduces the interference of the interdecadal component that may include uncertainty with the results on interannual time scales. The regression of the SAT in Asia-Pacific regions onto the band-filtered APC2 is presented in Figure 14. The magnitude of the regression is found substantially decreased over many regions. However, significant correlation areas can still be observed over the high-latitude Asian continent, the Maritime continent, and the subtropical western Pacific, indicating that the ASVD2 can impact the winter SAT in the Asian-Pacific region on both the interannual and longer time scales. The regression of the SAT in Asia-Pacific regions onto the APC2 with the interdecadal component is also examined (not shown). Generally, the results suggest that more contribution to the SAT in midlatitude and high-latitude Asia-Pacific regions comes from the interannual component of the ASVD2 while the interdecadal component of the ASVD2 plays more roles for the SAT variability over lower latitudes such as Indian Ocean and the western tropical Pacific. The regression map of the SAT in Asia-Pacific regions onto the band-filtered AO index is also examined and compared with Figure 14 (not shown). It shows that the significant TCC areas of the SAT associated with ASVD2 in the Asian-Pacific region is no longer larger than those related to the AO index. JIA ET AL.


446


10.1002/2014JD022324

Also, as discussed in previous studies [Lin et al., 2005; Jia et al., 2009], it is should be kept in mind that in this study the SVD analysis is applied between the tropical Pacific SST and the Z500 over the Northern Hemisphere. The large-scale atmospheric patterns obtained here only represent the dominant atmospheric signal that is associated with tropical Pacific SST forcing. There is evidence that other external forcings, for example, the SST anomalies in other ocean basins [Gollan et al., 2012; Zhang et al., 2013; Gong et al., 2014; Chen et al., 2014] or the snow cover over the Eurasian continent [e.g., Wang et al. 2010] can also play a role and may not be well represented by these atmospheric modes. Further studies are needed to understand the relative importance of these factors and how they contribute to the variability of the SAT in the Asian-Pacific region. Acknowledgments The reanalysis data used in this paper are available at http://www.esrl.noaa. gov/psd/data/gridded/data.ncep. reanalysis.html. The SST data are available at http://www.metoffice.gov. uk/hadobs/hadisst. This research was jointly funded by National Natural Sciences Foundation of China (grants 41475065 and 91337216) and by the Natural Science and Engineering Research Council of Canada (NSERC). Bao Qin is funded by 973 Project (grant 2012CB417203). We are thankful to National Oceanic and Atmospheric Administration of United States for providing AO index. We are grateful to the reviewers for their helpful suggestions on improving our paper.

JIA ET AL.

References Bao, Q., J. Yang, G. Wu, Y. Liu, and B. Wang (2010), Roles of anomalous Tibetan plateau warming on the severe 2008 winter storm in central-southern China, Mon. Weather Rev., 138, 2375–2384, doi:10.1175/2009MWR2950.1. Chen, M., P. Xie, J. Janowiak, and P. Arkin (2002), Global land precipitation: A 50-yr monthly analysis based on gauge observations, J. Hydrometeorol., 3, 249–266. Chen, W., J. Feng, and R. Wu (2013a), Roles of ENSO and PDO in the link of the East Asian winter monsoon to the following summer monsoon, J. Clim., 26, 622–635. Chen, W., X. Land, L. Wang, and Y. Ma (2013b), The combined effects of the ENSO and the Arctic Oscillation on the winter climate anomalies over East Asia, Chin. Sci. Bull., 58, 1355–1362. Chen, W., S. Yang, and R.-H. Huang (2005), Relationship between stationary planetary wave activity and the East Asian winter monsoon, J. Geophys. Res., 110, D14110, doi:10.1029/2004JD005669. Chen, Z., R. Wu, and W. Chen (2014), Distinguishing interannual variation of the northern and southern modes of the East Asian winter monsoon, J. Clim., 15, 835–851. Clark, M., and M. Mserreze (2000), Effects of variations in East Asian snow cover on modulating atmospheric circulation over the North Pacific Ocean, J. Clim., 13, 3700–3710. Ding, Y. (1994), Monsoon Over China, 550 pp., Kluwer Acad. Gollan, G., R. Greatbatch, and T. Jung (2012), Tropical impact on the East Asian winter monsoon, Geophys. Res. Lett., 39, L17801, doi:10.1029/2012GL052978. Gong, D., J. Yang, S. Kim, Y. Gao, D. Guo, T. Zhou, and M. Hu (2011), Spring Arctic Oscillation-East Asian summer monsoon connection through circulation changes over the western North Pacific, Clim. Dyn., 37, 2199–2216. Gong, D., Y. Gao, D. Guo, R. Mao, J. Yang, M. Hu, and M. Gao (2014), Interannual linkage between Arctic/North Atlantic Oscillation and tropical Indian Ocean precipitation during boreal winter, Clim. Dyn., 42, 1007–1027. Gong, D. Y., S. Wang, and J. Zhu (2001), East Asian winter monsoon and Arctic Oscillation, Geophys. Res. Lett., 28, 2073–2076. Huang, R., J. Chen, L. Wang, and Z. Lin (2012), Characteristics, processes and causes of the spatio-temporal variabilities of the East Asian monsoon system, Adv. Atmos. Sci., 29, 910–942. Huang, R. H., L. Zhou, and W. Chen (2003), The prograsses of recent studies on the variability of the East Asian monsoon and their causes, Adv. Atmos. Sci., 20, 55–69. Jia, X., and H. Lin (2011), Influence of forced large-scale atmospheric patterns on surface air temperature in China, Mon. Weather Rev., 139, 830–852. Jia, X., H. Lin, and J. Derome (2009), The influence of tropical pacific forcing on the Arctic Oscillation, Clim. Dyn., 32, 495–509. Jia, X., H. Lin, and J. Derome (2010), Improving seasonal forecast skill of North American surface air temperature in fall using a post-processing method, Mon. Weather Rev., 138, 1843–1857. Jia, X., H. Lin, and X. Yao (2014), The influence of tropical pacific SST anomaly on surface air temperature in China, J. Clim., 27, 1425–1444. Kalnay, E., et al. (1996), The NCEP/NCAR 40-year reanalysis project, Bull. Am. Meteorol. Soc., 77, 437–471. Lau, K., and T. Li (1984), The monsoon of East Asia and its global associations—A survey, Bull. Am. Meteorol. Soc., 65, 114–125. Lau, K.-M., and C.-P. Chang (1987), Planetary scale aspects of the winter monsoon and atmospheric teleconnection, in Monsoon Meteorology, edited by C.-P. Chang and T. N. Krishnamurti, pp. 161–202, Oxford Univ. Press, Oxford, U. K. Lee, J.-Y., B. Wang, Q. Ding, K.-J. Ha, J.-B. Ahn, A. Kumar, B. Stern, and O. Alves (2011), How predictable is the NH summer upper-tropospheric circulation?, Clim. Dyn, 37, 1189–1203. Li, S., M. P. Hoerling, S. Peng, and K. M. Weickmann (2006), The annular response to tropical Pacific SST forcing, J. Clim., 19, 1802–1819. Limpasuvan, V., and D. L. Hartmann (1999), Eddies and the annular modes of climate variability, Geophys. Res. Lett., 26, 3133–3136. Lin, H., and J. Derome (2004), Nonlinearity of the extratropical response to tropical forcing, J. Clim., 17, 2597–2608. Lin, H., J. Derome, and G. Brunet (2005), Correction of atmospheric dynamical seasonal forecasts using the leading ocean-forced spatial patterns, Geophys. Res. Lett., 32, L14804, doi:10.1029/2005GL023060. Lin, H., G. Brunet, and J. Derome (2008), Seasonal forecasts of Canadian winter precipitation by post-processing GCM integrations, Mon. Weather Rev., 136, 769–783. Liu, Y., L. Wang, W. Zhou, and W. Chen (2014), Three Eurasian teleconnection patterns: Spatial structures, temporal variability, and associated winter climate anomalies, Clim. Dyn., 42, 2817–2839. Rayner, N., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan (2003), Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century, J. Geophys. Res., 108(D14), 4407, doi:10.1029/2002JD002670. Rodwell, M., D. P. Rowell, and C. K. Folland (1999), Forcing of the wintertime North Atlantic Oscillation and European climate, Nature, 398, 320–323. Takaya, K., and H. Nakamura (2001), A formulation of a phase-independent wave activity flux for stationary and migratory quasigeostrophic eddies on a zonally varying basic flow, J. Atmos. Sci., 58, 608–627. Takaya, K., and H. Nakamura (2013), Interannual variability of the East Asian winter monsoon and related modulations of the planetary waves, J. Clim., 26, 9445–9461. Thompson, D., and J. M. Wallace (1998), The Arctic Oscillation signature in wintertime geopotential height and temperature fields, Geophys. Res. Lett., 25, 1297–1300.


447


10.1002/2014JD022324

Wang, B., R. Wu, and X. Fu (2000), Pacific-east asian teleconnection: How does ENSO affect East Asian climate?, J. Clim., 13, 1517–1936. Wang, L., and W. Chen (2010), How well do existing indices measure the strength of the East Asian Winter Monsoon, Adv. Atmos. Sci., 4, 855–870. Wang, L., and W. Chen (2014), An intensity index for the East Asian winter monsoon, J. Clim., 27, 2361–2374. Wang, L., W. Chen, and R. Huang (2007), Changes in the variability of North Pacific Oscillation around 1975/1976 and its relationship with East Asian winter climate, J. Geophys. Res., 112, D11110, doi:10.1029/2006JD008054. Wang, L., W. Chen, W. Zhou, J. C. L. Chan, D. Barriopedro, and R. H. Huang (2010), Effect of the climate shift around mid 1970s on the relationship between wintertime Ural blocking circulation and East Asian climate, Int. J. Climatol., 30(1), 153–158. Watanabe, M., and M. Kimoto (2000), Atmosphere-ocean thermal coupling in the North Atlantic: A positive feedback, Q. J. R. Meteorol. Soc., 126, 3343–3369. Wu, B., and R. Huang (1999), Effect of the extremes in the North Atlantic Oscillation on East Asia winter monsoon, Chin. J. Atmos. Sci., 23, 641–651. Wu, R., Z. Hu, and B. Kirtman (2003), Evolution of ENSO-related rainfall anomalies in East Asia, J. Clim., 16, 3742–3758. Wu, R., J. L. Kinter III, and B. Kirtman (2005), Discrepancy of interdecadal change in the Asian region between the NCEP-NCAR reanalysis and obvervations, J. Clim., 18, 3048–3067. Yang, J., B. Wang, B. Wang, and Q. Bao (2010), Biweekly and 21-30 day variabilities of the subtropical East Asian monsoon over the lower reach of Yangtze River basin, J. Clim., 23, 1146–1159. Yang, J., Q. Bao, D. Y. Gong, and B. Wang (2013), Distinct quasi-biweekly variations of the subtropical East Asian monsoon during early and late summers, Clim. Dyn., 42, 1469–1486, doi:10.1007/s00382-013-1728-6. Zhang, J., Q. Liu, C. Wang, and X. Zheng (2013), Impact of heating anomalies associated with rainfall variations over the Indo-western Pacific on Asian atmospheric circulation in winter, Clim. Dyn., 40, 2023–2033. Zhou, S., and A. J. Miller (2005), The interaction of the Madden-Julian Oscillation and the Arctic Oscillation, J. Clim., 18, 143–159.

JIA ET AL.


448