Distinct intraseasonal periodicity of Siberian high as modulated by ENSO: dominance of Madden–Julian oscillation

While previous studies have demonstrated that the El Niño-Southern Oscillation (ENSO) can impact global climate systems on the intraseasonal timescale, how ENSO affects the intraseasonal variability of the Siberian high (SH) still remains unclear. Based on spectral analysis, the SH exhibits evident intraseasonal periodicity (ISP) differences, with 25–50 d during El Niño winters and 50–90 d during La Niña winters. The northward propagation of the Rossby wave from the tropics during the phase transition of the SH leads to the Madden–Julian oscillation (MJO) dominating the difference in the ISP of the SH. The faster eastward propagation of the MJO in El Niño winters leads to a quicker evolution of the SH. Accompanied by the eastward propagation of the MJO, when the tropical westerly in the lower troposphere is located over the Indian Ocean, it can deflect to the north and involved the Eurasian cyclonic circulation, which contributes to the negative phase of the SH. When the tropical westerly moves eastward and the easterly is occupied the Indian Ocean, the middle-to-high latitude northerly completely dominates the East Asia region, penetrates southward to the South China Sea, deflects westward, and involved the tropical easterly, acting as the positive peak stage of the SH. The upper-level tropical zonal wind overrides the low-level tropical zonal wind, forming a coupled circulation and air temperature pattern together with the low level. The faster propagation of the MJO in El Niño winters leads to the intraseasonal evolution of the El Niño-related SH to change from lagging behind the La Niña-related SH to exceeding it, thereby contributing to the ISP of the SH being much shorter during El Niño winters.


Introduction
The Siberian high (SH), centered over the northwestern part of the Mongolian Plateau, is the predominant climate system over Eurasia during boreal winter.The variations in the SH have drawn a great attention because of their close connection to the strength of the East Asian winter monsoon (EAWM) and the local surface air temperature (SAT) (Cohen et al 2001, Gong and Ho 2002, Hou et al 2007, Lan and Li 2016, Riaz et al 2018, Liu and Zhu 2020, Zhang and Wang 2020).The intraseasonal variability (ISV) of the SH has been revealed to be extremely critical to the occurrence of cold-air outbreaks and weather extremes (Park et al 2014, Lü et al 2019).For instance, the record-breaking cold wave in the winter of 2016 was primarily attributed to variations in the intraseasonal periodicity (ISP) of the SH of 20-60 d (Zhou et al 2023).The intraseasonal variations in the SH is becoming an essential concern for winter disaster prevention.
The El Niño-Southern Oscillation (ENSO), as the strongest signal of tropical air-sea coupling on the interannual timescale, is commonly believed to have crucial impacts on the EAWM, including the SAT and precipitation anomalies on seasonal-to-interannual timescales (e.g.Chongyin 1990, Wang et al 2000, Ronghui et al 2004, 2008, Chen et al 2013a, 2013b, 2013c, He and Wang 2013, Gong et al 2015, Zhang et al 2015).Less attention has been paid to the modulation of ENSO on the East Asian climate on intraseasonal timescale.Teng and Wang (2003) suggested that ENSO can modulate the intraseasonal variations in the Asia-Pacific region by affecting the easterly vertical shear.Lau and Nath (2006) proposed that the western Pacific monsoon trough is the key to the ENSO's influence on the amplitude of the ISV of the EAWM.ENSO was also considered to impact the propagation of the ISV over the western North Pacific (Liu et al 2016) and the ISP of the Pacific-Japan teleconnection (Li et al 2020).In addition, there is no doubt that the Madden-Julian oscillation (MJO) (Madden andJulian 1971, 1972), which is regarded as the most primary ISV mode in the tropics, can be influenced by the ENSO (Zhang 2005, Hendon et al 2007, Moon et al 2011, Wu and Song 2018, Fernandes and Grimm 2023).The MJO activity over the equatorial Western Pacific generally becomes stronger in the developing phase of El Niño and reduces during the mature and decaying phases of El Niño (Wu and Song 2018).The propagation velocities are also distinct in the two ENSO phases, with the MJO propagating much faster during El Niño winters (Pohl andMatthews 2007, Wei andRen 2019).
As the most predominant intraseasonal mode in the tropics, the MJO exhibits distinct propagation and intensity variations under ENSO's modulation.However, there is no strong evidence that the ENSO plays a significant role in the intraseasonal variations in the SH so far.How the SH reacts to distinct ENSO backgrounds on the intraseasonal timescale was the primary focus of this study.Furthermore, the possible role of the MJO in ENSO's modulation to the SH was clarified.The rest of this paper is organized as follows: the data and methods are introduced in section 2. Section 3 presents the characteristics of the ISP of the SH, the associated cause-and-effect relationship between the SH and MJO, and the temporal evolution of atmospheric circulations under ENSO's modulation.Finally, the results are summarized and further discussion is provides in section 4.

Data and methods
In this study, the daily mean atmospheric data were obtained from the National Centers for Environmental Prediction-Department of Energy reanalysis 2 (Kanamitsu et al 2002).The daily mean outgoing longwave radiation (OLR) data were obtained from the National Oceanic and Atmospheric Administration interpolated OLR dataset (Liebmann and Smith 1996).The monthly mean sea surface temperature (SST) data were obtained from the Hadley Centre Sea Ice and Sea Surface Temperature dataset version 1 (HadISST) (Rayner et al 2003).All of the analyses were performed in boreal winter (from November to March, denoted as NDJFM) from 1979 to 2020.The long-term linear trend and the first three harmonics of the climatological annual cycle were removed to focus on the climate variability.To precisely identify the intraseasonal component, the Lanczos bandpass filter was applied (Duchon 1979).The statistical significance was assessed using the two-tailed Student's t-test and the effective degree of freedom (Bretherton et al 1999).
The daily SH index is defined as the area-weighted average of the sea level pressure (SLP) over the main active region of the SH (30 S1) (Gong and Wang 1999).It has been proven that this definition can objectively represent the variations in the intensity and location of the SH on the intraseasonal timescale (Zhou et al 2023).The EAWM index was calculated as the standardized SLP difference between (20 et al 1996).The real-time multivariate MJO index was used to identify the eight phases of the MJO (Wheeler and Hendon 2004).To clarify the cause-and-effect relationship between the SH and MJO, the Rossby wave source (RWS) (Sardeshmukh and Hoskins 1988) and flux (T-N flux) (Takaya and Nakamura 2001) were examined to represent the upper-level vorticity source setting off the Rossby wave train and to assess the large-scale wave activities that are responsible for the remote influence.El Niño (La Niña) events were identified as the 3 month running mean of the Niño 3.4 (5 • S-5 • N, 170 • W-120 • W) SST anomalies that exceeded the threshold of 0.5 • C (−0.5 • C) for at least five consecutive months (Ren et al 2018).Based on the above definition, we eventually identified a total of 14 El Niño winters and 12 La Niña winters (table S1).Details of the methodology are provided in supporting information S1.

Distinct ISP of the SH in El Niño and La Niña winters
Figure 1 illustrates the power spectra of the SH index in both El Niño and La Niña winters.If we connect the NDJFM index of all of the year, the spectra (figure 1(a)) exhibit significant peaks that exceed the red noise test on an extended timescale (10-25 d) and the intraseasonal timescale (25-50 d).Another peak occurs on the timescale of 50-90 d, but fails to pass the red noise test.To focus on the impact of the interannually varying background state related to the ENSO on the ISP of the SH, we recalculated the spectra of the NDJFM index during 14 El Niño winters and 12 La Niña winters.There was no clear difference in the extended timescale between in the El Niño and La Niña winters, but the peak of 25-50 d was significantly intensified in the El Niño winters and quite smooth in the La Niña winters.Nevertheless, the spectra of the 50-90 d were visibly amplified and close to exceeding the red noise test in the La Niña winters.To isolate the intraseasonal signal of the SH index, the effect of the interannual variability was removed by subtracting the running mean of 91 d, and a 5 day running mean was applied to remove the synoptic fluctuations (figure 1(b)).Despite the extended timescale, the spectral peak was prominent during 25-90 d.Two frequency bands were identified during both the El Niño and La Niña winters, which occurred on 25-50 and 50-90 d, respectively.Figures 1(c) and (d) present the average of the power spectra for NDJFM index during each year, and these spectra also exhibit quite distinct peaks at 10-25 and 25-50 d during the El Niño winters, and 10-25 and 50-90 d during the La Niña winters.The periodicity of the SH during the El Niño winters was apparently shorter than that during the La Niña winters.

Dominant role of the MJO in ENSO's modulation to the ISP of the SH
Figure 2 illustrates the temporal evolution (from day −20 to 0, also see figure S2 for the evolution from day 0 to +20) of the SH obtained by regressing the SLP, SAT and OLR anomalies against the 25-90-day bandpass-filtered SH index for both the El Niño and La Niña winters.During the El Niño winters, it exhibited a nearly complete phase transition from a negative Siberian SLP anomaly to a positive Siberian SLP anomaly as the SAT anomaly transformed from a cold Arctic-warm Eurasia pattern to a warm Arctic-cold Eurasia pattern.The associated OLR anomaly exhibited a zonal dipole pattern of propagating from west to east in the tropics, i.e. the typical MJO pattern.The temporal evolution of the SH during the La Niña winters was slower than that during the El Niño winters.For example, on day −20, the negative Siberian SLP anomaly and the associated cold Arctic-warm Eurasia pattern during the El Niño winters were stronger than those during the La Niña winters, indicating that the similar spatial pattern occurred at an earlier time for the SH during the La Niña winters (about day −24, figure not show).If we take East Asia (EA) (red box in figure 2) as a reference region, it is clear that on El Niño-related SH day −8, EA experienced a negative SLP anomaly, while on day −4, the positive SLP anomaly extended southward to reach EA.However, on La Niña-related SH day −8, the southern edge of the positive SLP anomaly has already reached EA.The transition from a warm SAT anomaly to a cold SAT anomaly over EA during the La Niña winters (from day −12 to −8) also occurred earlier than during the El Niño winters (from day −8 to −4).It should be noted that the SH reached its peak on day 0 during both the El Niño and La Niña winters, we suggest that the temporary evolution of the SH is faster during El Niño winters than during La Niña winters.
Accompanied by the distinct intraseasonal evolution features of the SH during the El Niño and La Niña winters, the MJO exhibited a clear difference.The negative OLR anomaly primarily occurred over the tropical eastern Indian Ocean on El Niño-related SH day −20 and occurred over the region from the tropical eastern Indian Ocean to the Maritime Continent on La Niña-related SH day −20.However, on El Niño-related SH day −16, the main body of the negative OLR anomaly had already moved eastward to the Maritime Continent, while on La Niña-related SH day −16, the negative OLR anomaly slowly moved, but the main body was still located over the region from the tropical eastern Indian Ocean to the Maritime Continent.On SH day 0, negative OLR anomaly in the El Niño-related SH had caught up with the negative OLR anomaly in the La Niño-related SH.Overall, during the evolution from El Niño-related SH day −20 to 0, the negative OLR anomaly propagated from the tropical eastern Indian Ocean to across the dateline, while during the evolution from La Niñarelated SH day −20 to 0, the negative OLR anomaly propagated from the tropical western Maritime Continent to west of the dateline.This suggests that there was a clear difference in the speed of the MJO propagation during the evolution of El Niño-related SH and that of the La Niña-related SH.The velocity potential at 200 hPa also supported the clear difference in the propagation velocity (figure S3).
To determine the dominance of the MJO in the ENSO's modulation to the ISP of the SH, figure 3 presents spatiotemporal cross sections of the associated tropical OLR and zonal wind at 925 hPa (U925) for the evolution of both the El Niño-related SH and the La Niña-related SH.Both the OLR and U925 anomalies exhibited typical zonal dipole eastward propagation patterns.The negative (positive) OLR anomaly was accompanied by a positive (negative) U925 anomaly.The eastward propagation of the MJO during the evolution of the El Niño-related SH was faster than that of the evolution of the La Niña-related SH.Taking the negative OLR anomaly as a reference, during the evolution of the El Niñorelated SH, the central position of the negative OLR anomaly was located at about 95 • E on day −20 and 150 • W on day +5.The average eastward propagation speed was about 5.95 m s −1 .During the evolution of the La Niña-related SH, the central position of the negativeSince the ISP of the SH was accompanie OLR anomaly was about 105 • E on day −20 and 175 • E on day +5, so the average eastward propagation velocity was about 3.62 m s −1 .This was consistent with the results of Wei and Ren (2019) that the ENSO modulates the fast (about 6.0 m s −1 ) and slow (4.0 m s −1 ) MJO modes during boreal winter.
The EAWM and EA SAT also exhibited distinct evolution.As shown in figure 3, the phase transition frequencies of the EAWM and EA SAT associated with the El Niño-related SH were faster than those associated with the La Niña-related SH.The peak time of the two indices roughly corresponded to the time when the center of the negative OLR anomaly reached the dateline (i.e.MJO phase 7).Analysis of the spectra of the daily EA SAT index (figure S4) and the daily EAWM index (figure S5), and the autocorrelations of the 25-90-day bandpass-filtered SH index, EAWM index, and EA SAT index (figure S6) further support the conclusion that the ENSO modulated the ISP of the SH.
Since the ISP of the SH was accompanied by eastward propagation of the MJO, we subsequently focused on how the MJO dominated the distinct ISP of the SH during the El Niño and La Niña winters.To clarify the cause-and-effect relationship between the ISP of the SH and the propagation of the MJO, we investigated the upper-level wave activity associated with the evolution of the SH and examined the association between the MJO and the SH by compositing the 25-90-day bandpass-filtered SH index during the eight phases of the MJO and calculating the percentage of the 25-90-day bandpass-filtered SH index around its peak from day −30 to +30 during the eight phases of the MJO (figure 4).The relationship between the MJO phase and SH was found to be locked.When the MJO convection was located over the Indian Ocean (i.e.phases 2-4), the SH index reached its valley, acting as the negative phase of the SH.When the MJO convection reached the Western Pacific as its eastern region (i.e.phases 6-8), the SH evolved to its strongest stage.The transition from the negative SH to the positive SH occurred around the time when the MJO convection was located over the Maritime Continent (i.e.phase 5).These features were strong during both the El Niño and La Niña winters, and the peak time (MJO phase 7 for the El Niño-related SH and MJO phases 6-7 for the La Niña-related SH) and valley time (MJO phases 3-4 for the El Niño-related SH and MJO phases 2-3 for the La Niña-related SH) of the El Niño-related SH occurred slightly earlier than during those of the La Niña-related SH according to the probability of the distribution of the SH index during the MJO phase.
It is clear that the tropical and high latitude regions were the two main RWSs of the SH activity, and they played quite different roles.When the SH researched its peak or valley, the upstream Ural anticyclone had a crucial impact on the intensity of the SH (figure S7).When the SH was in the developing (near MJO phase 4 to phase 6) or decaying (near MJO phase 8 to phase 2) stage (viz., SH phase transition), the northward propagation of the Rossby waves from the tropics played a dominant role, while the highlatitude Rossby waves were relatively weak.The major difference between the El Niño and La Niña winters was the maintenance time of the effect of the MJO, which was primarily related to the propagation speed of the MJO.The faster propagation of the MJO during the El Niño winters led faster evolution of the SH.
Since the SH is characterized by a cold shallow anticyclonic high-pressure system confined to the low-level troposphere, we further investigated atmospheric circulations of the upper and lower troposphere to determine the evolution of the SH during the El Niño and La Niña winters (figure 5) (also see figures S8-S10).On El Niño-related SH day −17, the low-level tropical westerly was located over the tropical Indian Ocean, deflected northward and involved the middle-to-high latitudes to form a wide Eurasian cyclonic circulation with a cold Arcticwarm Eurasia air temperature pattern.The upperlevel tropical easterly override the low-level tropical westerly, merged into the middle-to-high latitudes, and formed a cyclonic circulation from the Ural Mountains to far eastern Siberia and an anticyclonic circulation over Lake Baikal with an air temperature pattern almost opposite to that in the lower level, exhibiting a typical pattern of the negative phase of the SH.In contrast, on La Niña-related SH day −17, it resembled the low-level cyclonic circulation over Eurasia decayed after El Niño-related SH day −17 with a colder Arctic and weakened warm Eurasia air temperature pattern.The upper-level tropical easterly in the El Niño-related SH was west of that in the La Niña-related SH, indicating that the evolution of the El Niño-related SH lagged behind the evolution of the La Niña-related SH.
On SH day −8, the low-level tropical westerlies in the El Niño-related SH and La Niña-related SH both moved to the Maritime Continent, which disrupted the pattern of the Eurasian cyclonic circulation and transformed into the Eurasian anticyclone.The air temperature anomaly also gradually evolved from a cold Arctic-warm Eurasia pattern to a warm Arctic-cold Eurasia pattern.The low-level transition was coupled with the upper troposphere.The upperlevel circulation pattern was shifted southeastward overall, which was accompanied by eastward movement of the tropical easterly that override the lowlevel westerly to the Maritime Continent.A new anticyclone was formed over the Barents-Kara Sea.Its location in the El Niño-related SH was slightly west of that in the La Niña-related SH, indicating that the evolution of the El Niño-related SH on day −8 still lagged behind that of the La Niña-related SH.
As the time proceeded to day +2, the low-level tropical easterly was occupied over the tropical Indian Ocean in the El Niño-related SH, but was not yet prevailed in the La Niña-related SH, suggesting that the evolution of the El Niño-related SH had taken the lead.The air temperature anomalies in Eurasia entirely evolved into a cold Arctic-warm Eurasian pattern, which was accompanied by a large anticyclonic circulation in the middle-to-high latitudes.The northerly completely dominated in East Asia, penetrated deep into the South China Sea, deflected westward and involved the tropical easterly.Thus, with the eastward movement of the tropical zonal wind (viz., the MJO eastward propagation), the low-level circulation over Eurasia completed the transition from a cyclonic anomaly and cold Arctic-warm Eurasia pattern to an anticyclonic anomaly and a warm Arcticcold Eurasia pattern.In the upper level, the tropical westerly that override the low-level easterly in the El Niño-related SH also outstripped that in the La Niñarelated SH.The tropical zonal winds involved the middle-to-high latitudes and formed an anticyclone from southern China to the Northwestern Pacific, a cyclone over the southern region of Lake Baikal, and an anticyclonic circulation from the Ural Mountains to far eastern Siberia, acting as the further overall southeastward evolution of the circulation pattern on day −8.Thus, it can be seen that the faster eastward propagation of the MJO during the El Niño winters accelerated this process, which allowed the SH to develop faster on the intraseasonal timescale during the El Niño winters than that during the La Niña winters.
According to previous research, the mechanism behind the MJO's faster propagation during El Niño winters is primarily summarized as dynamic wave feedback and moisture-convection feedback (Sobel and Maloney 2013, Liu and Wang 2017, Wang and Chen 2017, Wei and Ren 2019).Typically, the favorable environment in which convection occurs ahead of the direction of movement is seen as a critical condition that influences the velocity and the extent of the propagation of the MJO.By comparing the background winter-mean SST and low-level specific humidity anomalies during the El Niño and La Niña winters (figure S11), it is found that the patterns associated with the two phases of the ENSO events exhibited diametrically opposed characteristics.The Indian Ocean and equatorial east-central Pacific exhibited positive (negative) SST and specific humidity anomalies during the El Niño (La Niña) winters, with negative (positive) SST and specific humidity anomalies over the tropical Pacific to the east of the Philippines.The background air-sea environment during the El Niño winters was ideal for the generation of deep convection and supported the rapid propagation of the MJO to a far eastward.

Conclusions and discussion
In this study, El Niño and La Niña winters were identified to explore the modulation of the ENSO on the intraseasonal SH.Through power spectrum analysis, it is found that the SH exhibited distinct ISP during the two phases of the ENSO events, with 25-50 d during the El Niño winters and 50-90 d during the La Niña winters.The northward propagation of the Rossby wave from the tropics during the phase transition of the SH indicates that the difference in the ISP of the SH was primarily dominated by the MJO.The eastward propagation of the MJO was faster during the El Niño winters, leading to a quicker evolution of the SH.The evolution of the atmospheric circulation revealed that in the low troposphere, when the tropical westerly was located over the Indian Ocean, it deflected to the north and involved the Eurasian cyclonic circulation, which contributed to the negative phase of the SH.When the tropical westerly moved eastward and the easterly was occupied the Indian Ocean, the middle-to-high latitude northerly completely dominated East Asia, penetrated southward to the South China Sea, deflected westward and involved the tropical easterly, acting as the positive peak stage of the SH.The upper-level tropical zonal wind override the low-level tropical zonal wind, forming a coupled circulation and air temperature pattern together with the low level.The faster propagation of the MJO during the El Niño winters caused the intraseasonal evolution of the El Niñorelated SH to change from remaining behind the La Niña-related SH to exceeding it, thereby contributing to the much shorter ISP of the SH during the El Niño winters.
Our findings highlight the dominant role of the tropical intraseasonal oscillation in the effect of the modulation of the interannually varying background state related to the ENSO on the ISP of the SH, and reveal the pathway by which the MJO influenced the SH.However, some questions still need to be further clarified.For example, it should be noted that the intensity of the intraseasonal variation of the SH, as well as the EAWM and EA SAT, were stronger during the La Niña winters than during the El Niño winters.The Rossby wave activity during the peak time of the SH indicates that the intensity of the intraseasonal variation in the SH may partly influenced by the upstream Ural anticyclone, but this requires further research.In addition to the MJO, previous studies have emphasized the feedback of the upstream Rossby wave activity and downstream synoptic eddy to the SH on the intraseasonal timescale associated with the North Atlantic Oscillation (NAO) (Takaya and Nakamura 2005a, 2005b, Zhou et al 2023).Since the NAO has been revealed to have significant ISV (Ren et al 2022), it may be another key factor that enables the ENSO to modify the ISV of the SH.Moreover, recent studies have demonstrated that the intensity and spread of the MJO exhibit a clear difference during Central-Pacific and East-Pacific El Niños (Pang et al 2016) and the developing and decaying stages of the ENSO (Wu and Song 2018).The diversity of the ENSO-MJO connection, as well as the different stages of the ENSO, may contribute to the individual yearly ISP of the SH, which also requires further exploration.

Figure 1 .
Figure 1.(a) Power spectra of the NDJFM SH index.The red, blue, and dark gray dashed curves are the red noise spectra corresponding to the SH index spectra during the El Niño winters, La Niña winters, and all of the winters, respectively.(b) Similar to (a) but for the SH index obtained by removing the interannual variability and synoptic fluctuations.(c) Mean power spectrum for the SH index during NDJFM in each year.The red, blue, and dark gray dashed curves are the red noise spectra corresponding to the SH index spectra during the El Niño winters, La Niña winters, and all of the winters, respectively.(d) Similar to (c) but for the SH index obtained by removing the interannual variability and synoptic fluctuations.

Figure 2 .
Figure 2. Evolution patterns of the SLP anomalies (contours, interval of 50 Pa; the zero contours are omitted), SAT anomalies (blue-red shading, unit: • C), and OLR anomalies (green-yellow shading, unit: W m 2 ) regressed on the 25-90-day bandpass-filtered SH index during the El Niño (left panel) and La Niña (right panel) winters (from day −20 to 0).Only the regressed SAT anomalies and OLR anomalies confident at the 95% level based on the Student's t-test are plotted.

Figure 3 .
Figure 3. (a) Longitude-time cross section of the 25-90-day bandpass-filtered SH index regressed OLR anomalies (green-yellow shading, unit: W m 2 ) and U925 anomalies (contours, interval of 0.1 m s −1 ; the zero contours are omitted) averaged over 15 • S-10 • N from the 30-day lead to the 30-day lag during the El Niño winters.The blue and red dotted lines represent the associated EAWM index and EA SAT index averaged over the red box in figure 2, respectively.(b) Similar to (a) but for the La Niña winters.

Figure 4 .
Figure 4. (a) Latitude-time cross section of the 25-90-day bandpass-filtered SH index regressed 200 hPa RWS anomalies (shading, unit: 10 −10 s −2 ) and T-N flux anomalies (vectors, unit: m 2 s −2 ) averaged over 60-120 • E from the 30-day lead to the 30-day lag during the El Niño winters.The octagon represents the composite of the 25-90-day bandpass-filtered SH index in the eight phases of the MJO.The solid dots denote significant SH index values above the 95% confidence level based on the two-tailed Student's t test.The raster fill represents the percentage of the positive 25-90-day bandpass-filtered SH index values around its peak from day −30 to day +30 in the eight phases of the MJO.The blue line denotes the autocorrelation coefficient of the 25-90-day bandpass-filtered SH index.(b) Similar to (a) but for the La Niña winters.

Figure 5 .
Figure 5. Evolution patterns of the air temperature anomaly (shading, unit: • C) and horizontal wind anomaly (vector, unit: m s −1 ) of each layer (925 hPa and 200 hPa) regressed on the 25-90-day bandpass-filtered SH index, respectively, during the El Niño (left panel) and La Niña (right panel) winters on day −17, day −8, and day +2.Only the regressed anomalies confident at the 95% level based on the Student's t-test are plotted.
migratory quasigeostrophic eddies on a zonally varying basic flow J.Atmos.Sci.58 608-27  Takaya K and Nakamura H 2005aMechanisms of intraseasonal amplification of the cold Siberian high J. Atmos.Sci.62 4423-40 Takaya K and Nakamura H 2005b Geographical dependence of upper-level blocking formation associated with intraseasonal amplification of the Siberian high J. Atmos.Sci.62 4441-9 Teng H and Wang B 2003 Interannual variations of the boreal summer intraseasonal oscillation in the Asian-Pacific region J. Clim.16 3572-84 Wang B and Chen G 2017 A general theoretical framework for understanding essential dynamics of Madden-Julian oscillation Clim.Dyn.49 2309-28 Wang B, Wu R and Fu X 2000 Pacific-East Asian teleconnection: how does ENSO affect East Asian climate?J. Clim.13 1517-36 Wang L, Chen W and Huang R 2008 Interdecadal modulation of PDO on the impact of ENSO on the East Asian winter monsoon Geophys.Res.Lett.35 L20702 Wei Y and Ren H L 2019 Modulation of ENSO on fast and slow MJO modes during boreal winter J. Clim.32 7483-506 Wheeler M and Hendon H 2004 An all-season real-time multivariate MJO index: development of an index for monitoring and prediction Mon.Weather Rev. 132 1917-32 Wu R and Song L 2018 Spatiotemporal change of intraseasonal oscillation intensity over the tropical Indo-Pacific Ocean associated with El Niño and La Niña events Clim.Dyn.50 1221-42 Zhang C 2005 Madden-Julian oscillation Rev. Geophys.43 RG2003 Zhang N and Wang G 2020 Detecting the causal interaction between Siberian high and winter surface air temperature over northeast Asia Atmos.Res.245 105066 Zhang R, Li T, Wen M and Liu L 2015 Role of intraseasonal oscillation in asymmetric impacts of El Niño and La Niña on the rainfall over southern China in boreal winter Clim.Dyn. 45 559-67 Zhou F, Shi J, Liu M H and Ren H C 2023 Linkage between the NAO and Siberian high events on the intraseasonal timescale Atmos.Res.281 106478