1 Introduction

The western Pacific subtropical high (WPSH) and the South Asia High (SAH) are two stable and persistent anticyclonic circulation systems in subtropical region. The former is located in lower- to mid-troposphere over the western Pacific, while the latter is located in the upper-troposphere over the Tibetan Plateau (Hu 1997; Liu and Wu 2004; Liu et al. 2014; Wang et al. 2016). Both of them show remarkable variation on multiple timescales, which significantly influence the climate over East Asia. On seasonal timescale, the WPSH and the SAH advance northward in summer owning to the seasonal variation of solar radiation, and as the result the rain belt moves northward (Qian et al. 2002; Li and Zhu 2008). On interannual-interdecadal timescale, the intensity and the location of these two systems vary due to the variation of thermal conditions of the ocean and the plateau respectively, which greatly affect the drought-flood condition over East Asia (Sui et al. 2007; Jiang et al. 2011; Huang et al. 2017).

As attention is drawn to subseasoanl timescale (10–90 days), it has been found that the climate over East Asia exhibits prominent subseasoanl variability, and it is largely affected by the atmospheric circulation, especially the WPSH and the SAH (Mao and Wu 2006; Mao et al. 2010; Chen and Zhai 2015; Gui et al. 2022). Considering the significant impact of the WPSH and SAH, their subseasonal variabilities have been studied separately. As shown in Fig. 1, the WPSH first extends westward and then retreats eastward, indicating that the WPSH shows notable zonal oscillation on subseasonal timescale (Qian and Yu 1991; Yu and Yang 1995; Qian and Shi 2017; Yang and Li 2020; Zi et al. 2021). During the westward extension of the WPSH, an anomalous anticyclone persists in lower- to mid-troposphere at the west side of the WPSH. When the WPSH stretches to the westernmost position, the anomalous anticyclonic circulation reaches the strongest (Ren et al. 2013; Guan et al. 2018, 2019).

Fig. 1
figure 1

a The standard deviation of subseasonal geopotential height anomaly at 500 hPa (H500, shaded) and the climatological mean 5875 gpm isoline (contour). The red box denotes the core region of the WPSH (110°E–140°E, 10°N–30°N). b The power spectrum of the subseasonal WPSH index (WPSHI, black solid line). The red and black dash lines represent the Markov red noise spectrum and the spectrum of 95% confidence level, respectively. c, e Temporal evolution of the composite 5875 gpm isolines for positive events in 10–25-day period from day − 8 to day + 8 with interval of 2 days. d, f Same as (c, e), but in 25–50-day period from day − 16 to day + 16 with interval of 4 days

Full size image

Investigated in our previous work (Nie et al. 2022), the frequency of the zonal oscillation of the WPSH is mainly concentrated in two bands, a 10–25-day period and 25–50-day periods. These two sub-periods of the zonal oscillation of the WPSH may be related to two sub-periods of tropical convection activities (Li et al. 2023). In 10–25-day period, the zonal oscillation of the WPSH is mainly attributed to the convection-circulation interaction, that is, the atmospheric internal dynamics. The suppressed convection induces the divergence anomaly at the west side of the WPSH, further stimulating the anomalous anticyclonic circulation here. In turn, the anomalous anticyclonic circulation motivates the downward dry advection, which favors the propagation of this combined convection-circulation system and leads to the zonal oscillation of the WPSH. The boundary forcing has no obvious feedback on the zonal oscillation of the WPSH in 10-25-day period. In 25–50-day period, the zonal oscillation of the WPSH is also related to the convection-circulation interaction. In comparation, in 10–25-day period the convection anomaly generates at the latitude near 10°N, while in 25–50-day period the convection anomaly generates northerly compared to the shorter period. When the convection anomaly reaches the maximum, in 10–25-day period the center of the convection anomaly is much stronger with a more westward location than the center in 25–50-day period. Thus, the WPSH could extend more westward in shorter period (Fig. 1). Besides, the air-sea interaction, that is, the boundary forcing, also works in 25–50-day period. When the SST anomaly fully develops, the oceanic effect dominates the turbulent fluxes anomalies by changing the air-sea humidity and temperature difference. These may have impact on the energy balance of the atmosphere, and then, the circulation anomaly. In addition, the Rossby wave train (Enomoto et al. 2003; Guan et al. 2019), through transporting energy and momentum, also partly contribute to the zonal oscillation of the WPSH.

The SAH also oscillates remarkably in zonal direction (Mason and Anderson 1963; Tao and Zhu 1964; Zhang et al. 2002). On subseasonal timescale, it can be found that the east edge of the SAH first moves eastward and then retreats westward (Liu and Lin 1991; Ren et al. 2015; Cheng et al. 2021; Zhu et al. 2023). Accompanied with this oscillation, an anomalous anticyclonic circulation persists in upper-troposphere at the east side of the SAH. It reaches the strongest when the SAH extends to the easternmost position. The zonal oscillation of the SAH is mainly caused by the diabatic heating over the continents (Liu et al. 2007; Yang and Li 2016; Shang et al. 2018). Based on potential vorticity (PV) diagnosis, Ren et al. (2015) emphasizes that the anomalous diabatic heating-induced negative PV dominates the east side of the SAH, and with strengthening and eastward propagation of this negative PV anomaly, the SAH extends more eastward. Moreover, the zonal oscillation of the SAH is also associated with a mid-latitude wave train across Eurasia (Liu et al. 2020; Ren et al. 2015).

Although these two systems are studied separately in most studies, it should be highlighted that their zonal movements are not independent of one another (Cheng et al. 2021; Jiang et al. 2011; Wei et al. 2019a). On subseasoanl timescale, numbers of studies indicated that accompanied with westward extension of the WPSH, the SAH in upper-troposphere also extends eastward (Tao and Zhu 1964; Ren et al. 2007; Chen and Zhai 2015; Zhu et al. 2023). This zonal approach movement between the WPSH and the SAH is essential for climate variabilities over East Asia, as it enhances the moisture transportation by the low-level southwesterly (Guan et al. 2018; Li and Mao 2018; Mao et al. 2010), and meanwhile, maintains the local anomalous ascending motion (Chen and Zhai 2014, 2015; Wei et al. 2019b), favoring the occurrence of the precipitation events. Currently, there are two points of view to understand the cause of the zonal approach movement between the WPSH and the SAH. One emphasizes the dynamical effect of the vorticity advection in upper-troposphere (Ren et al. 2007; Chen et al. 2014). It suggests the negative vorticity advection, related to the eastward extension of the SAH, forces the anomalous divergence anomaly in lower-troposphere, and the WPSH extends westward as the adjustment to this anomalous divergence anomaly. The other emphasizes the thermodynamical effect of anomalous diabatic heating over East Asia (Liu et al. 1999; Chen and Zhai 2015; Cheng et al. 2021). With anomalous heating over continent and cooling over western Pacific, it generates anomalous negative vorticities at the east of the SAH in upper-troposphere, and also the west of the WPSH in lower- to mid-troposphere. As this wave-like pattern of anomalous diabatic heating strengthens, the WPSH and the SAH approach. Thus, an interesting question arises as which process dominates in the zonal approach movement of the WPSH and the SAH. What’s more, as the frequency of the zonal oscillation of the WPSH is concentrated in two bands (Fig. 1), it needs more discussion whether there are differences in the connection of the WPSH and SAH between these two periods.

In this study, our goal is to understand the evolution characteristics of the WPSH and the SAH, and the underlying mechanism of the zonal approach movement between these two systems. The remainder of paper is structed as follow. In Section 2, the data and methods are introduced. In Section 3, we present the evolution characteristics of the zonal oscillation of the WPSH and the SAH in 10–25-day period and 25–50-day period respectively, based on the composite of identified zonal oscillation events of the WPSH. In Sections 4 and 5, we investigate the underlying mechanism of the connection between the WPSH and the SAH and discuss the differences between two periods. The last section is a summary and discussions for further study.

2 Data and methods

In this study, the atmospheric data is obtained from the Japanese 55-year Reanalysis (JRA-55) dataset with spatial resolution of 1.25°×1.25° (Kobayashi et al. 2015). The daily precipitation data is obtained from the US Climate Prediction Center (CPC) with spatial resolution of 0.5°×0.5° (Xie et al. 2007). The daily outgoing longwave radiation (OLR) data, as a proxy of the convection activity, is obtained from the National Oceanic and Atmospheric Administration (NOAA) with spatial resolution of 2.5°×2.5° (Liebmann and Smith 1996). The period of 1979–2020 summer (June–August) is selected in this study. For all the data above a 5-day (13-day) running average is used to remove the components with a period of less than 10 days (25 days). Then the daily climatological mean annual cycle and the yearly summer average are removed. Meanwhile, the butterworth filter is applied to extract anomalies in 10–25-day period and in 25–50-day period.

Note that if we continue use the 5880 gpm isolines to depict the WPSH as convention, the coverage of the WPSH in JRA-55 would be much smaller than other datasets (Figure not shown). Thus, in this study the WPSH is measured by 5875 gpm isolines at 500 hPa. Same as our previous work, the subseasonal variability of geopotential height at 500 hPa (H500) get maximum at the western side of the WPSH (Fig. 1), compared to the rest regions covered by the WPSH. Thus, considering the span of the zonal oscillation of the WPSH, we take the region (10°N–30°N, 110°E–140°E) as the core region of the zonal oscillation of the WPSH (hereafter region-WPSH). The WPSH index (WPSHI) is defined by subseasonal H500 anomaly averaged over the region-WPSH, and a positive index here represents a westward-extended WPSH. Based on the WPSHI, the time-lag composition is used in this study. For composition, the positive (negative) event is identified when the filtered WPSHI exceeds (below) the value of 1.5 times the standard deviation. Following this criteria, 84 positive events and 84 negative events are identified in the 10–25-day period, while 41 positive events and 40 negative events in the 25–50-day period. In time-lag composition, day 0 represents the day when the filtered WPSHI reaches the maximum (minimum) value, that is, the day when the WPSH reaches the westernmost position.

To clarify the connection between the zonal oscillation of the WPSH and the SAH, here, we investigate the cause of the anomalous circulation in the upper-troposphere by diagnosing the vorticity equation:

$$\begin{array}{c}\begin{array}{c}\frac{\;\partial\zeta^{\prime}}{\partial t}=-\;\left(\overset\rightharpoonup{{\mathbf V}_h}\cdot\boldsymbol\nabla\zeta\right)^{\prime}-\;\left(\omega\frac{\partial\zeta}{\partial p}\right)^{\prime}-\;\beta v^{\prime}+\;S^{\prime}-\;\left[\left(\zeta+f\right)D\right]^{\prime}\end{array}\\\begin{array}{ccccccccccc}&&&&&&&&&&\begin{array}{ccccc}&(\mathrm a)&&&\begin{array}{cccccc}&&(\mathrm b)&&(\mathrm c)&\begin{array}{cccccc}(\mathrm d)&&&(\mathrm e)&&\begin{array}{cccccc}&&&&&\end{array}\end{array}\end{array}\end{array}\end{array}\end{array}$$
(1)

where \(\zeta\) is the relative vorticity, \(t\) the time, \(\overset\rightharpoonup{{\mathbf V}_h}\) the horizontal flow \(\left(u\overset\rightharpoonup{\boldsymbol i}+v\overset\rightharpoonup{\boldsymbol j}\right)\), \(\nabla\) the horizontal gradient operator, \(\omega\) the vertical velocity, \(p\) the pressure, \(\beta\) the meridional gradient of the Coriolis parameter \(f\), \(S\) the tilting term expressed as \(\left[\left(\partial \omega /\partial y\right)\left(\partial u/\partial p\right)-\left(\partial \omega /\partial x\right)\left(\partial v/\partial p\right)\right]\), and \(D\) the divergence. The five terms on the right-hand side of the Eq. 1 respectively represent (a) the horizontal vorticity advection, (b) the vertical vorticity advection, (c) the \(\beta\) effect, (d) the tilting effect and (e) the horizontal divergence effect. The prime denotes the filtered anomalies.

3 Evolution characteristics of the WPSH and the SAH

We first focus on the evolution characteristics of the WPSH and the SAH in 10–25-day period. Shown in Fig. 1c and d, the west edge of the WPSH is located east of 145°E at the day − 8. It stretches westward stepwise and reaches its westernmost position near the coast at the day 0. After that, the WPSH retreats eastward, accelerates in region east of 120°E, and finally retreats to the beginning location. Accompanied with the zonal movement of the WPSH in 10–25-day period, the SAH also oscillates remarkably. From the time-lag composition, the zonal oscillation of the SAH lags the WPSH about one day (Fig. 2a, d). It demonstrates that in 10–25-day period the zonal oscillation of the WPSH may has certain effect on the zonal oscillation of the SAH. For positive events (Fig. 2b, c), the east edge of the SAH first moves eastward slowly from the location near 100°E. After the day − 3, the eastward movement of the SAH is accelerating, and the SAH reaches the easternmost position near the 112.5°E at around the day + 1. Then, the SAH withdraw westward stepwise to the starting location until the day + 8. The evolution characteristics of negative events (Fig. 2e, f) are nearly opposite to those of positive events. The SAH first move westward and then extend eastward, with a longitude span of 10°.

Fig. 2
figure 2

a The time-lag composition of the 10–25-day filtered WPSHI and the zonal location of the eastern ridge points of the SAH for positive events. b, c Temporal evolution of the composite 5875 gpm isolines for positive events in 10–25-day period from day − 7 to day + 9 with interval of 2 days. d-f Same as (a-c), but for negative events

Full size image

Figure 3 shows the temporal evolution of 10–25-day filtered OLR, geopotential height at 850 hPa (H850) and 200 hPa (H200), and the vertical velocity at 400 hPa (ω400) of positive events. Accompanied with the westward extension of the WPSH in 10–25-day period, a positive OLR anomaly can be observed over the western Pacific. At the northwest side of the positive OLR anomaly, there is a positive H850 anomaly, as the atmospheric response to the convection-induced diabatic heating (Gill 1980). Meanwhile, along with the positive OLR anomaly, the anomalous descending motion is also evident. These anomalies grow and propagate northwestward from the region east of 140°E, gradually occupying the western Pacific, and get their peak at around day 0 when the WPSH stretches to the westernmost location. Afterwards, these anomalies weaken as continuously propagating northwestward, and finally disappear near the continent at around day + 4. Simultaneously, the new negative H850 anomaly and negative OLR anomaly with anomalous ascending motion move northwestward and gradually occupy the western Pacific. These are corresponding to the retreat of the WPSH. In upper-troposphere, a positive H200 anomaly is located at the south of the Japan when the WPSH extends westward. It moves westward towards the continent, corresponding to the eastward extension of the SAH. At around day + 1 as the center of the positive H200 anomaly moves to the east edge of the SAH, the SAH extend to the easternmost location. After that, this positive H200 anomaly weakens and continues to move westward to the location near 100°E, and at the south of the Japan it is dominant by the new negative H200 anomaly. It is associated with westward retreat of the SAH.

Fig. 3
figure 3

a-e Composites of 10–25-day filtered anomalies of geopotential height at 850 hPa (H850, green contours with interval of 2gpm) and OLR (shaded, unit: W·m−2) for positive events from day − 4 to day + 4 with interval of 2 days. f-j Same as (a-e), but for anomalies of geopotential height at 200 hPa (H200, green contours with interval of 2gpm) and vertical velocity anomaly at 400 hPa (ω400, shaded, unit: 10−2 Pa·s−1) from day − 3 to day + 5 with interval of 2 days. Purple contours in (a-e) denote the 5875 gpm isolines and (f-j) the 12,520 gpm isolines. Dots and shade in (a-e) and (f-j) denote areas with significance at 95% confidence level for anomalies of H850, ω400, H200 and OLR, respectively

Full size image

The zonal oscillation of the WPSH is also prominent in 25–50-day period (Fig. 1), but in comparation with smaller longitude span than it in shorter period. Accompanied with the zonal oscillation of the WPSH, in 25–50-day period the movement of the SAH is inconspicuous (Fig. 4). For positive events, the east edge of the SAH is first located at the location near 104°E for a long time. After the day − 10, it begins to moves eastward with a longitude span of 5°. After reaching the easternmost location, the SAH retreat to the starting location and also maintains there for several days. For negative events, the maintenance of the SAH is more detectable. the SAH first stays around 107.5°E until day + 4 and then it slightly moves eastward, also with a longitude span of 5°. Such characteristics indicates that in 25–50-day period the zonal oscillation of the SAH is weakly connected with the WPSH.

Fig. 4
figure 4

Same as Fig. 2, but for 25–50-day period

Full size image

Same as Fig. 3, the temporal evolution of 25–50-day filtered variables for positive events is shown in Fig. 5. Accompanied with the westward extension of the WPSH in 25–50-day period, the WPSH-related variables, including the positive H850 anomaly, the positive OLR anomaly and the anomalous descending motion, also can be observed at the west side of the WPSH. They grow and propagate westward, reach the strongest at around day 0. Although these anomalies also occupy the western Pacific, in 25–50-day period the centers of these anomalies are relatively weak and unable to propagates farther westward compared to the shorter period (more discussion in Section 5). After the day 0, these anomalies weaken and proceed to move westward. As they disappear near the continent, the subsequent negative OLR anomaly and negative H850 anomalies enhance to occupy the western Pacific, thus the WPSH retreats. In upper-troposphere, there is also a westward moving H200 anomaly from Japan to the continent. At day 0, it moves to the region near the coast. However, the center of this positive H200 anomaly is located near 35°N, more northly than it in 10–25-day period. As location of the positive H200 anomaly is away from the 12,520 gpm isolines, it reduces the longitude span of the zonal oscillation of the SAH.

Fig. 5
figure 5

Same as Fig. 3, but for 25–50-day period

Full size image

4 Impact of the WPSH on the zonal oscillation of the SAH in 10–25-day period

From the investigation in Sect. 3, it suggests that in 10–25-day period the WPSH has certain effect on the zonal oscillation of the SAH. In this section, we explore this underlying mechanism based on the vorticity equation (Eq. 1). As shown in Fig. 6a, the variability of subseasaonal H200 anomaly is significant over the region (25°N–32.5°N, 107.5°E–117.5°E) compared to the rest regions covered by the SAH. Meanwhile, the subseasaonal H200 anomaly over this region is highly positive correlated with the variation of the east edge of the SAH. Thus, we take this region as the core region of the zonal oscillation of the SAH (hereafter region-SAH).

Fig. 6
figure 6

a The correlation between the eastern ridge points of the SAH and the subseasonal H200 anomaly (shaded) and the departure from the zonal mean of the standard deviation of subseasonal H200 anomaly (black contour, with interval of 3). Purple contour denotes the climatological mean 12,520 gpm isoline, and the green box denotes the core region of the SAH (107.5°E–117.5°E, 25°N–32.5°N). b, c Temporal evolution of 10–25-day filtered anomalies of the vorticity at 200 hPa (purple line, unit: 10−6s−1), the vorticity tendency (pink line, unit: 10−11 s−2), and the vorticity source terms on the right side of Eq. 1 (unit: 10−11 s−2), including the horizontal vorticity advection term (dark blue line), the vertical vorticity advection term (dark green line), the \(\beta\) effect term (black line), the tilting effect term (red line) and the horizontal divergence effect term (light green line), and the residual of the Eq. 1 (light blue line)

Full size image

The remains in Fig. 6 shows temporal evolution of the 10–25-day filtered anomalies of the vorticity, the vorticity tendency (the left-hand side of the Eq. 1) and the vorticity source terms (the right-hand side of the Eq. 1) at 200 hPa, averaged within the region-SAH. It can be seen that the vorticity tendency precedes the vorticity anomaly by a quarter of cycle, indicating that there is a periodic oscillation in vorticity anomaly. The residual, the difference between the vorticity tendency and the sum of the vorticity source terms, is much smaller than the vorticity tendency itself, which mainly represents the weak effect of the friction on vorticity variation. These confirm the balance framework exist in Eq. 1. Among the vorticity source terms in Eq. 1, the vertical vorticity advection term and the tilting term can be neglected due to their small magnitude. The \(\beta\) effect is the consequence of the variation of the anomalous circulation, due to its positive correlation with local anomalous meridional wind. It leads the vorticity anomaly by about a quarter of cycle, which means that the \(\beta\) effect favors the westward movement of the vorticity anomaly. The horizontal relative vorticity advection is also the consequence of the variation of the anomalous circulation, as it lags the vorticity anomaly. It suggests that the SAH extends eastward with advecting negative vorticity to its east, unfavorable for the westward movement of the vorticity anomaly.

Except the above terms, the transition of the horizontal divergence effect term \(-{\left[\left(\zeta +f\right)D\right]}^{{\prime }}\) from positive to negative leads the vorticity anomaly in temporal. In spatial, the horizontal divergence effect term becomes obvious over the continent at day − 1 and moves nearly in-phase with the H200 anomaly (left panel of the Fig. 7). These suggest that in 10–25-day period the generation and the maintenance of vorticity anomaly in region-SAH are dominantly caused by the horizontal divergence effect. To further explore what contributes to the horizontal divergence effect term, a decomposition is applied as we did in (Nie et al. 2022). The first-order approximation of the decomposition can be written as:

Fig. 7
figure 7

a-e Composites of 10–25-day filtered anomalies of the H200 (black contours with interval of 2gpm) and the horizontal divergence effect term \(-\left[\left(\zeta +f\right)D\right]^{\prime }\) (shaded, unit: 10−11 s−2) at 200 hPa from day − 3 to day + 5 with interval of 2 days. f-j Composites of 10–25-day filtered anomalies of the horizontal divergence effect term (blue bar) and its two components, the anomalous divergence-related term (\(-\stackrel{-}{\left(\zeta +f\right)}{D}^{{\prime }}\), red bar) and the anomalous vorticity-related term (\(-\left(\zeta +f\right)^{\prime }\stackrel{-}{D}\), green bar), averaged within the core region of the SAH (unit: 10−11 s−2). (k-o) Longitude–height sections of 10–25-day filtered anomalies of the diabatic heating (shaded, unit: K·day−1), the divergence in upper-troposphere (pink contour with interval of 0.3×10−6s−1) and the wind (vectors, \(v\), m·s-1, ω, -0.01 Pa·s−1), averaged over 25°N–30°N

Full size image
$$\begin{array}{c}-\left[\left(\zeta+f\right)D\right]{^\prime}\approx-\left(\overline{\zeta+f}\right)D{^\prime}-\left(\zeta+f\right)'\overline D\end{array}$$
(2)

where the terms on the right-hand side represent the anomalous divergence-related term and the anomalous vorticity-related term, respectively. Based on Eq. 2, diagnosis shows that the horizontal divergence effect term is dominated by the anomalous divergence-related term (middle panel of the Fig. 7), which indicates that in 10–25-day period the SAH-related anomalous circulation is highly related to local divergence anomaly. The Coriolis force turns the anomalous divergent winds into anticyclonic shear, finally causes the eastward extension of the SAH.

What causes this divergence anomaly? Shown in right panel of the Fig. 7, the divergence anomaly in upper-troposphere is nearly in-phase with the anomalous diabatic heating (Q) in 10–25-day period. The anomalous diabatic heating first become obvious at the location 117.5°E at the day − 1, with its strongest heating in 300–400 hPa. This configuration of the diabatic heating may expand the air column, induce the ascent and generate local divergence anomaly in upper-troposphere. From day − 1 to day + 1, the anomalous diabatic heating enhances with its center moving to the location beneath the region-SAH, which benefits the strengthening of the local ascent and divergence anomaly in upper-troposphere. As it triggers the anomalous anticyclonic circulation in the region-SAH by the horizontal divergence effect, we observe an eastward-extended SAH. Afterward, the anomalous diabatic heating proceeds to move westward and beneath the region-SAH the anomalous diabatic heating reduces. In this case, the local ascent and the divergence anomaly weaken gradually in region-SAH. As the result, in upper-troposphere the intensity of anomalous anticyclonic circulation decreases here, and the SAH retreats westward. From the analysis above, it is the anomalous diabatic heating that result in divergence anomaly at east side of the SAH, and thus affects the atmospheric circulation here.

The anomalous diabatic heating is dominated by the latent heat of the precipitation over subtropical region of the East Asia. Accompanied with westward extension of the WPSH, in the Yangtze River basin it would experience above-normal precipitation, which supports the anomalous diabatic heating at east side of the SAH, while in southern China the opposite condition (left panel of the Fig. 8). This precipitation pattern is attributed to the anomalous moisture convergence in the Yangtze River basin as well as the anomalous moisture divergence in southern China (right panel of the Fig. 8). The extended WPSH contribute a lot to moisture convergence. Similar to the evolution characteristics of the anomalous diabatic heating, the excessive moisture transportation is detectable in the Yangtze River basin at day − 1, which causes remarkable precipitation here. To the day + 1, with the WPSH extending further westward, the anomalous moisture convergence strengthens over the continent, leading to the intensification of the precipitation and the westward expansion of the precipitation region. As the precipitation center is located beneath the region-SAH, it supports the anomalous diabatic heating here. Afterwards, with the weakness of the WPSH-related anomalous moisture convergence, the precipitation intensity decreases in region-SAH and thus the anomalous diabatic heating reduces.

Fig. 8
figure 8

a–e Composites of 10–25-day filtered precipitation anomaly (shaded, unit: mm) from day − 3 to day + 5 with interval of 2 days. Purple contours denote the 5875 gpm isolines. Dots denote areas with significance at 95% confidence level for precipitation anomaly. f–j Same as (a-e), but for integrated moisture flux anomalies (vectors, unit: kg·m−1·s−1) and its divergence (shaded, unit: 10−5 kg·m−2·s−1)

Full size image

Overall, it seems that in 10–25-day period the WPSH-related precipitation is the key factor for the WPSH to impact on the zonal oscillation of the SAH. When the WPSH extends westward, it enhances the moisture transportation to the east side of the SAH, causing above-normal precipitation here. With the enhancement of the precipitation, the released anomalous diabatic heating induces divergence anomaly in upper-troposphere and further stimulates anomalous anticyclonic circulation by the horizontal divergence effect. This anomalous anticyclonic circulation moves westward to the region-SAH, and finally cause the eastward extension of the SAH.

5 Difference in relationship of the WPSH and the SAH between two periods

As mentioned above, compared to the shorter period, in 25–50-day period the zonal movement of the SAH is much weaker as the positive H200 anomaly center is located away from the 12,520 gpm isolines. It remains questionable what causes this difference between two periods. Given that in 10–25-day period the precipitation over East Asia plays a significant role for the WPSH to affect high-level atmospheric circulation, it is worthwhile to investigate the difference in precipitation pattern between two periods.

Figure 9 shows the composites of precipitation, diabatic heating and high-level divergence anomalies in two periods when the SAH extends to the easternmost position. In 10–25-day period the border of the drought region and the flood region is located south of the Yangtze River. The above-normal precipitation can be observed in the Yangtze River basin, while the below-normal precipitation in the southern China. Accompanied with this precipitation pattern, the anomalous diabatic heating occupies the region north of 25°N over the East Asia. Therefore, in upper-troposphere it is dominated by the divergence anomaly at the east side of the SAH, which further stimulates anomalous anticyclonic circulation. In 25–50-day period, though the precipitation also shows north-south diploe pattern, the border of the drought region and the flood region is located along the Yangtze River. The above-normal precipitation only occurs north of the Yangtze River, and the region south of the Yangtze River is occupied by below-normal precipitation. As the consequence, the anomalous diabatic heating and associated divergence anomaly are restricted to the north of 30°N, and the anomalous diabatic cooling is weak at the east side of the SAH. It is unfavorable for maintenance of the anomalous anticyclonic circulation at the east side of the SAH, and the eastward extension of the SAH is limited.

Fig. 9
figure 9

a Composites of 10–25-day filtered precipitation anomaly when the SAH extends to the easternmost position. b Latitude–height sections of 10–25-day filtered diabatic heating anomaly (shaded, unit: K·day−1) and the divergence anomaly in upper-troposphere (pink contour with interval of 0.3×10−6s−1), averaged over 110°E–117.5°E. c, d Same as (a, b), but for 25–50-day period. Purple contours denote the 5875 gpm isolines

Full size image

To further explore the reasons for the difference in precipitation patterns, the anomalous moisture transportation in two periods are shown in Fig. 10. Accompanied with the westward extension of the WPSH, the related anomalous anticyclonic circulation transports moisture to the continent in both two periods. However, in 10–25-day period more moisture can be transported to the continent due to the farther westward extension of the WPSH, compared to it in longer period. As the result, in 10–25-day period the anomalous moisture convergence occupies the Yangtze River basin, while in 25–50-day period it only occurs in the north of Yangtze River, which is consistent with the anomalous precipitation pattern. Then, why the WPSH can extend farther westward in shorter period? As shown in Fig. 10c, when the WPSH reaches the westernmost location, in 10–25-day period the center of the positive OLR anomaly is much stronger with a more westward location than the center in 25–50-day period. This favors the strengthening of the anomalous anticyclonic circulation in shorter period, and thus leads to more moisture transportation.

Fig. 10
figure 10

a, b Composites of integrated moisture flux anomalies (vectors, unit: kg·m−1·s−1) and its divergence (shaded, unit: 10−5 kg·m−2·s−1) in (a) 10–25-day period and (b) 25–50-day period when the SAH extends to the easternmost position. The amount of the moisture transported into region-SAH are shown in the upper right corner of the figure. Dark blue contours denote the 5875 gpm isolines. c Composites of OLR anomaly in 10–25-day period (shaded, unit: W·m−2) and 25–50-day period (contour with interval of 3 W·m−2). The values of the OLR anomaly centers in both periods are shown inside the figure

Full size image

The results above suggest that the difference in the zonal oscillation of the SAH between two periods is mainly due to distinct precipitation patterns. In 10–25-day period, with a stronger and more westward OLR anomaly, the WPSH extends farther westward compared to the longer period. It favors more moisture converge over the Yangtze River basin, which increases the precipitation at the east side of the SAH. As the result, in 10–25-day period the stimulated anomalous anticyclonic circulation is located close to the SAH, leading to more obvious zonal oscillation of the SAH. While in 25–50-day period, the moisture convergence is restricted to the north of Yangtze River, away from the SAH, and so is the precipitation region. At the east side of the SAH it is dominated by the anomalous cooling with weak convergence anomaly, unfavorable for the maintenance of the anomalous anticyclonic circulation. Therefore, the zonal movement of the SAH is weaker in 25–50-day period.

6 Summary and discussion

The zonal oscillation of the WPSH and the SAH are connected on subseasoanl timescale. In this study, we investigate the evolution characteristics of the WPSH and the SAH, and the underlying mechanism of the zonal approach movement between these two systems by the composite analysis based on the WPSHI. Given that the frequency of zonal oscillation of the WPSH is mainly concentrated in two periods, a 10–25-day period and a 25–50-day period, in this study we also discuss the difference in the relationship of the zonal oscillation of the WPSH and the SAH between these two periods.

Accompanied with the westward extension of the WPSH in 10–25-day period, over the pacific a positive OLR anomaly propagates northwestward with a positive H850 anomaly at its northwest side. Both the positive OLR anomaly and positive H850 anomaly get their peak at around day 0 when the WPSH stretches to the westernmost location. Afterwards, these anomalies weaken and finally disappear near the continent, corresponding to the eastward retreat of the WPSH. In upper-troposphere, the SAH shows conspicuous eastward extension as the WPSH extends westward. It is highly related with a westward moving positive H200 anomaly from the south of the Japan. At around day + 1 as this positive H200 anomaly moves to the east edge of the SAH, the SAH extend to the easternmost location. Subsequently, the positive H200 anomaly weakens as moving westward, corresponding to the westward retreat of the SAH.

It can be seen in 10–25-day period that the zonal oscillation of the SAH lags the variation of the WPSH by about one day, indicating that the WPSH may has certain effect on the zonal oscillation of the SAH. Diagnosis based on the vorticity equation (Eq. 1) indicates that the vorticity variation related to the SAH is mainly caused by the horizontal divergence effect term \(\left[-\left(\zeta+f\right)D\right]'\). The transition of the horizontal divergence effect term \(-{\left[\left(\zeta +f\right)D\right]}^{{\prime }}\) from positive to negative leads the vorticity anomaly in temporal. This suggests the horizontal divergence effect can be conducive to the development of anomalous circulation in upper-troposphere. Further analysis shows that the WPSH-induced precipitation is a key factor for the cause of the related divergence anomaly. Concluded in Fig. 11, as the WPSH extends westward, the anomalous anticyclonic circulation over the western Pacific enhances the moisture transportation to the east side of the SAH, causing above-normal precipitation here. With the enhancement of the precipitation, the anomalous diabatic heating induces divergence anomaly in upper-troposphere, and further stimulates anomalous anticyclonic circulation by the horizontal divergence effect. This anomalous anticyclonic circulation moves westward close to the SAH, and as the result, we can observe a more eastward-extended SAH.

Fig. 11
figure 11

Schematic diagram of the impact of the WPSH on the zonal oscillation of the SAH in 10–25-day period

Full size image

In 25–50-day period, the zonal oscillation of the WPSH is also prominent, but in comparation with smaller longitude span than it in shorter period, as the centers of the OLR anomaly and the H anomaly are weaker and unable to propagate farther westward. In upper-troposphere, our findings suggest the zonal movement of the SAH is limited in 25–50-day period due to the H200 anomaly center located away from the SAH. This difference results from the distinct precipitation patterns in two periods. In 10–25-day period, with a stronger and more westward OLR anomaly, the WPSH could extend farther westward. It causes more moisture converge over the Yangtze River basin, generating above-normal precipitation with anomalous diabatic heating at the east side of the SAH. As the result, the stimulated anomalous anticyclonic circulation is close to the border of the SAH, leading to a much stronger zonal oscillation of the SAH. While in 25–50-day period, due to the weaker H anomaly induced by relatively weak OLR anomaly, less moisture is transported to the continent. The anomalous moisture convergence is restricted to the north of Yangtze River, and so is the precipitation region. At the east side of the SAH, it is dominated by the anomalous cooling, unfavorable for maintenance of the anomalous anticyclonic circulation, and thus, the zonal oscillation of the SAH.

Currently, our work reveals the relationship between the zonal oscillation of the WPSH and the SAH, and highlights the significant role of the WPSH-related precipitation and latent heat release in 10–25-day period. It can be found that accompanied with the westward extension of the WPSH, in upper-troposphere there is a westward moving H200 anomaly and thus leads to the eastward extension of the SAH. As proposed by Ren et al. (2015), the SAH’s eastward extension is contributed from an eastward moving high from the western flank of the Tibetan Plateau (TP) to the east of TP due to the variation of the above-normal diabatic heating. This difference in conclusion can be attributed to the distinct events we concern in works, as Ren et al. (2015) is based on the zonal oscillation events of the SAH, while in this work it is based on the zonal oscillation events of the WPSH. It also implies that the diabatic heating related with the zonal oscillation of the SAH is not only affected by the WPSH. The relative contribution of the impact of the WPSH and other factors needs further analyses.