1. Introduction
The East Asian summer monsoon (EASM) exhibits robust northward migration after the onset of the South China Sea summer monsoon in May (Qian et al. 1998; Wang and Ding 2008; Zhu et al. 2005). The rainfall over North China peaks in August when the EASM arrives at its northern edge, while the western North Pacific subtropical high (WNPSH) is near 35°N and a tropical southwesterly wind prevails in the lower troposphere of South and Southeast Asia (Chen et al. 2006; Qian et al. 2002). Rainfall usually decreases over North China when the EASM retreats in late August (Wang and LinHo 2002; Zeng and Lu 2004; Zhang and Wang 2008). In September, the EASM retreats to the south of the Huang River. It terminates the rainy season over North China but starts the West China rainfall in autumn. The autumn rainfall over West China accounts for more than 20% of the in situ annual amount of precipitation (Bai and Dong 2004; Ding and Wang 2008; Gao and Guo 1958; He 1984). Several studies have treated the rainfall in September–November over West China as integral to investigate its variability at the interdecadal (Qin et al. 2018; Wang and Zhou 2019; Yang et al. 2016) or interannual time scale (Luo et al. 2013; Zhou and Wang 2019). It has been suggested that global sea surface temperature anomalies (SSTAs), including El Niño–Southern Oscillation (ENSO) events (Gong and Wang 1999; Hu et al. 2021; B. Wang et al. 2020), the Indian Ocean dipole (IOD) (Li and Zhao 2019; Qiu et al. 2014), and the North Atlantic SSTAs (Xu et al. 2013; Zhu et al. 2020), could regulate West China autumn rainfall by changing either the low-level WNPSH or the upper-tropospheric East Asian jet (EAJ) in the post monsoon stage.
As a summer-to-winter transition period, the circulation regime of the EASM and rainfall in September are distinct from those in the remaining two months in autumn (Ding 2004; Kuang and Zhang 2005; Zhang et al. 2006). Comparing with the situation in August, the meridional position of the upper-level EAJ remains around 40°N but its zonal center shifts evidently eastward from 95° to 135°E in September, along with the low-level WNPSH near 35°N and the southward withdrawal of monsoonal rainfall. In October, the circulation changes abruptly in the Northern Hemisphere (Yeh et al. 1958), including the southward migration of the EAJ and WNPSH as well as the demonstrable shrinking of EASM rainfall. The EAJ further intensifies and maintains at 30°N in November, when the winter monsoon circulation controls the East Asian continent and the WNPSH moves toward the western Pacific. In 2021, the climate anomalies were striking over East Asia in September, when extremely heavy rainfall occurred in central northern China (CNC), outside the domain of West China autumn rainfall (China Meteorological Administration 2021). Serious floods have affected the CNC region since mid-September. They broke the historical records in some tributaries of the Huang River and profoundly impacted agricultural production, economic manufacturing, and residents’ lives and property. For instance, floods affected 1.75 million people and damaged more than 17 000 houses in Shanxi Province, China.
The anomalous autumn rainfall in East China has been attributed to the anomaly of intensity or meridional position of the WNPSH (Bai and Dong 2004; H.-X. Li et al. 2017). But it is still unknown how the WNPSH modulates the CNC rainfall in September at the interannual time scale. On the other hand, previous studies have comprehensively summarized the ENSO impacts on the WNPSH (T. Li et al. 2017; Wang et al. 2013; Xie et al. 2016). However, in September 2021, the La Niña signal was too weak to explain the extreme anomaly of CNC rainfall, implicating the possible contribution of other SSTAs. Therefore, the present study aims to investigate the influences of the WNPSH on CNC rainfall in September and identify the combined effects of SSTAs in different basins on extreme events in 2021 on an interannual time scale. We arrange the remainder of the article as follows: section 2 describes the data, methods, and numerical models; section 3 shows the anomalous climate and circulation around the CNC in September 2021; section 4 quantifies the WNPSH behavior in September; section 5 illustrates the mechanisms of anomalous WNPSH in September by case diagnosis, statistical analysis, and numerical experiments; and finally section 6 provides the conclusions and discussion.
2. Data and methods
a. Data description
b. Numerical experiments
We used two sets of numerical experiments to verify the observed and statistical results. One was an ideal linear baroclinic model (LBM) forced by anomalous diabatic heating. This model had a T42 horizontal resolution and 20 vertical sigma levels (Watanabe and Kimoto 2000). The other was an atmospheric general circulation model of the Community Atmosphere Model version 5.3 (CAM5.3) forced by the observed SSTAs in the specific basin. The horizontal and vertical resolutions of CAM5.3 were 1.9° × 2.5° finite-volume grids and 27 hybrid sigma–pressure levels, respectively (Worley and Drake 2005). The design of numerical experiments will be demonstrated in section 5.
3. Anomalous rainfall and circulation over CNC in September 2021
In September 2021, the centers of anomalous rainfall in China emerged in the lower valley of the Huang River, accompanied by the extremely westward extension of the WNPSH into the East Asian continent (Fig. 1a). The anomaly of accumulative rainfall in CNC reached 85.34 mm, breaking the historical record since 1961. The monthly mean anomaly of rainfall amount is determined by the interannual variability with an explained variance of 83%, whereas the explained variance of the interdecadal component is 17%, playing a secondary role (Fig. 1b). Although the westward extension of the WNPSH in 2015, 2017, and 2020 is comparable to that in 2021 (dashed lines in Fig. 1a), the CNC rainfall amount is normal in 2015 but even below normal in 2020 (Fig. 1b). This is probably because the WNPSH splits over the western Pacific and exhibits bogus westward extension in these years. The rainy season is thus prolonged to mid-October 2021 in CNC and resulted in severe floods (Fig. 1c). The above-normal rainfall over the CNC does not exhibit extreme anomalies in either August or October, which indicates the particularity of climate anomalies in September 2021 (figure not shown).
(a) Cumulative rainfall anomaly over China (shading; mm) and 500-hPa western North Pacific subtropical high (WNPSH) represented by the 5880-gpm geopotential height contours (blue line) in September 2021. Black and gray lines indicate the 5880-gpm geopotential height contours in climatology and each year during 1981–2020, respectively. Dashed pink lines are for the cases in 2015, 2017, and 2020, when the westward extension of the WNPSH resemble that in 2021. The dashed box denotes the central northern China (CNC) domain (31°–41°N, 104°–120°E). (b) Time series of anomalous accumulative precipitation (mm) averaged over the CNC in September from 1961 to 2021. The blue and black lines represent the interdecadal and interannual components of the index, respectively. (c) Temporal evolution of the pentad-mean precipitation rate averaged over the CNC. The red (black) line indicates the case in 2021 (climatology), while the light blue shading represents the variation in the precipitation rate in each pentad from 1981 to 2020.
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
As the WNPSH extends markedly westward in September 2021, more moisture converges over the CNC and the northern Indian Peninsula (NIP) due to the stronger southerly to the west and easterly to the south of the anomalous WNPSH, respectively (Fig. 2a). Another anomaly center of moisture convergence settles around the MC. In the meantime, the water vapor tends to decrease over South China, the South China Sea, and the western Pacific beneath the main body of the WNPSH. The anomalies of moisture flux depend primarily on the anomalous circulation, in contrast to little influences due to the anomalies of specific humidity (figure not shown). The precipitation increases (decreases) in the wetter (drier) region, corresponding to the anomalies of diabatic heating (cooling) in situ (Fig. 2b). Afterward, the ascending motion is enhanced over the MC. The airflow diverges and turns poleward in the upper troposphere. It sinks to form an anomalous baroclinic circulation with upper-level cyclonic and low-level anticyclonic anomalies over South China (Fig. 2b). This anomalous sinking could intensify the westward extension of the WNPSH and meteorological drought over South China. When it reaches the ground, the airflow turns northward to transport more water vapor to the CNC. The extratropical ascending flow develops over the CNC and maintains the anomalous anticyclone in the upper troposphere. The meridional structure of anomalous circulation suggests tropical influences on the WNPSH and CNC rainfall in September 2021.
Anomalies of circulation over East Asia in September 2021. (a) Horizontal distribution of anomalous vertically integrated moisture flux from 1000 to 100 hPa (vectors; kg m−1 s−1) and its divergence (shading; 10−5 kg m−2 s−1). (b) 105°–130°E-averaged pressure–latitude cross section of anomalous meridional circulation (vectors; m s−1), diabatic heating (shading; K day−1), and streamfunction (contours; 106 m2 s−1).
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
The convection–wind relationship over East Asia in September 2021 intuitively reflects the horizontal distribution of anomalous circulation at different levels. As a Gill response to the deeper MC convection and suppressed western Pacific rainfall (Gill 1980), the anomalous circulation is vertically baroclinic between 20°S and 30°N, presenting an upper-tropospheric anticyclone (cyclone) and a low-level cyclone (anticyclone) over central Asia (western Pacific) (Figs. 3a,c). A barotropic wave train is observed in the middle latitudes over East Asia. It starts from an anticyclonic anomaly over the CNC and propagates eastward to enhance the East Asian trough near Japan and the anticyclone over the North Pacific (Fig. 4a). The linkage between this wave train and the upstream one originating from the North Atlantic is weak in September when the fluctuations residing in the East Asian jet (EAJ) are weak in climatology. Subsequently, the two wave trains bridge together as the EAJ accelerates in October 2021 (Fig. 4b), consistent with the results of Zhu et al. (2020).
Anomalous winds associated with CNC rainfall in September. (left) Anomalous streamfunction (shading; 106 m2 s−1) and winds (vectors; m s−1) in September 2021 at (a) 150, (c) 500, and (e) 850 hPa. (right) As in the left column, but for the regressed streamfunction (shading; 106 m2 s−1) and winds (vectors; m s−1) against the CNC rainfall in September during 1981–2020. Black vectors and stippled regions denote the 95% confidence level values. The purple dashed box is the domain for the WEI definition. Gray shading indicates the topography.
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
Horizontal distribution of anomalous streamfunction (shading; 106 m2 s−1) and wave activity flux (vectors; 1012 m2 s−2, values > 0.15 × 1012 are in black) at 200 hPa in (a) September and (b) October 2021. Black contours are for the climatological 200-hPa westerly wind starting from 20 with an interval of 10 m s−1. The wave activity flux is calculated following Takaya and Nakamura (2001).
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
The anomalous circulation regime in 2021 resembles the pattern regressed against the CNC rainfall in September from 1981 to 2020 (Figs. 3b,d,f). At the interannual time scale, the 500-hPa anticyclone anomaly over the East China Sea is most significant in the regression field, indicating the universal association of the WNPSH with East Asian climate in early autumn (Fig. 3d), while the Mongolian cyclone in 2021 does not show an evident anomaly as suggested by the regression result. Both case study and statistical analysis confirm the robust relationship between the anomalous WNPSH and CNC rainfall in September.
4. Definition of westward extension index of WNPSH
To quantify the westward extension of WNPSH, we define an index as the normalized area-averaged 500-hPa streamfunction in the domain 20°–40°N, 110°–140°E, which is significantly associated with CNC rainfall in September (Fig. 5a). The WNPSH expands westward to the East Asian continent in the positive index years but retreats eastward to the western Pacific in the negative index cases (Fig. 5b). Thus, this index represents the zonal extension of the WNPSH in September and is termed as the westward extension index (WEI). The temporal correlation coefficient (TCC) of interannual variability between the WEI and CNC rainfall in September is +0.37, exceeding the 95% confidence level during 1981–2020. The WEI is negative in 2015 and normal in 2020, suggesting that the index could partly exclude the bogus extremely westward extension of the WNPSH. Note that the WEI is the third largest, but the CNC rainfall is unprecedented in 2021 since 1981 (Fig. 5a). This is partly ascribed to the two typhoons Chanth and Mindulle that moved northward in the western North Pacific and brought more abundant water vapor into the CNC. In general, the WEI can depict the year-by-year difference in the westward extension of the WNPSH and its association with the CNC rainfall in September.
Time series of WEI and composed WNPSH during 1981–2021. (a) WEI (bars) and its interdecadal (blue line) and interannual components (black line). Green line denotes the standardized anomaly of CNC rainfall in September. Dashed horizontal lines indicate the standardized deviation (STD). Red (blue) bars are for the composite years with WEI above (below) +0.8 (−0.8) STD. (b) Horizontal distribution of the composite 5880-gpm geopotential height in the selected positive (bold red contour with thin pink lines for each case) and negative (bold blue contour with thin light blue lines for each case) WEI years. The black line denotes the climatological 5880-gpm contour.
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
The variability of WNPSH has many properties, including its intensity, area, meridional position of ridgeline, and western edge. Previous studies have measured these features using original (Liu et al. 2019) or eddy geopotential height at 500 hPa (He et al. 2018; Liu et al. 2019). In the original set, the WNPSH outline is the 5880-gpm contour of geopotential height. By contrast, the 20-gpm contour of eddy geopotential height, which is the deviation from the area-averaged geopotential height in the domain 10°–50°N, 80°E–170°W, acts as the WNPSH outline in the eddy set. The intensity, area, meridional position of ridgeline, and western edge of the WNPSH are the area-averaged geopotential height, the total grid number, the mean latitude of ridgeline, and the westernmost grid point within the WNPSH outline, respectively. The WEI is most significantly correlated with the area and west edge index of the WNPSH in both original and eddy set, indicating that positive WEI corresponds to the larger and more westward extension of the WNPSH. Meanwhile, the CNC rainfall is not only significantly associated with the WNPSH area in both the original and eddy set but is remarkably correlated with the west edge index in the eddy set (Table 1). In September 2021, the evident negative west edge index confirms the remarkable westward extension of the WNPSH. Therefore, the WEI can well describe the relationship between CNC rainfall and westward extension of the WNPSH in the warming world (He et al. 2015, 2018).
Temporal correlation coefficients of WEI and CNC rainfall in September with WNPSH indices in 1981–2021 at the interannual time scale. Values exceeding the 95% and 99% confidence level are marked by one and two asterisks, respectively.
5. Role of tropical SSTAs in westward extension of WNPSH in September
a. Diagnosis results
The anomalous convection and SSTAs in the tropics show effective influences on westward extension of the WNPSH in September. In 2021, the anomalies of the atmospheric heat source (AHS) present two positive centers over the MC and NIP with deeper tropical convection. Negative AHS anomalies exist over the central tropical Pacific and eastern coast of North America, where precipitation is suppressed (Fig. 6a). The positive AHS anomalies over the MC and NIP are above warm SSTAs in the Indo-Pacific warm pool and the northern Indian Ocean. By contrast, cold SSTAs in the central and eastern Pacific are beneath the negative AHS anomalies (Fig. 6c), suggesting oceanic forcing to the atmosphere. Accordingly, the large-scale ascending flow intensifies over the warmer Indo-Pacific warm pool and sinks above the colder tropical Pacific. Meanwhile, the SST warms up in the tropical Atlantic with the local enhanced AHS (Figs. 6a,c).
Anomalies of large-scale circulation and SSTA in September associated with WEI. (left) Horizontal distribution of (a) anomalous 150-hPa velocity potential (contours; 105 m2 s−1), atmospheric heat source (AHS; shading; W m−2), and divergent winds (vectors; m s−1) and (c) SSTA (K) in September 2021. (right) As in the left column, but for the regression field against the WEI. The dashed boxes in (a) and (b) indicate the northern Indian Peninsula and Maritime Continent, respectively. The dashed boxes in (c) are for the critical basins. The black vectors in (b) and stippled area in (d) represent the values exceeding the 95% confidence level. The AHS is defined as vertically integrated diabatic heating from 1000 to 100 hPa.
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
The intrinsic modes of tropical SSTAs obtained by the empirical orthogonal function (EOF) analysis can represent the SSTAs in September 2021 (Fig. 7a). The explained variance of the first EOF mode is 36.25% and its principal component (PC1) shows remarkable interannual variability (Fig. 7c). By contrast, the second EOF mode and the principal component (PC2) with the explained variance of 15.70% exhibits significant global warming trend (Figs. 7b,d). The positive PC1 and peak of PC2 in September 2021 suggest that both the interannual variation and global warming background contribute to the warmer Indo-Pacific warm pool and tropical Atlantic in this case.
EOF analysis on the tropical SSTAs (15°S–15°N) in September from 1981 to 2021. (a),(b) The first and second EOF modes (°C), respectively. The explained variance of each mode is at the top-right corner of each panel. The dashed boxes are for the critical basins. (c),(d) The corresponding normalized primary components (PCs). Red dashed lines indicate ±1.0 standard deviation of each PC.
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
b. Statistical analysis
In general, the remarkable anomalies of AHS and large-scale circulation regressed against the WEI from 1981 to 2020 resemble their observed configuration in 2021 (Figs. 6a,b), which coincides with the significant positive correlation between WEI and AHS over the MC and NIP (Figs. 8a,b). The WEI also shows a prominent positive correlation with PC1 of the tropical SSTAs, the SSTAs in the Indo-Pacific warm pool and tropical Atlantic (Figs. 8c,e,f), along with less significant correlation between WEI and PC2 of the tropical SSTAs (Fig. 8d). In 2021, the WEI reached its third largest value since 1981, accompanied by the strongest AHS over the MC and the second strongest AHS over the NIP (Figs. 8a,b). The warm SSTA is ranked second in the Indo-Pacific warm pool but sets a record in the tropical Atlantic (Figs. 8e,f). The tropical Atlantic becomes the second warmest after removing the long-term warming trend. Since the SSTAs in the tropical eastern Pacific and Indian Ocean are weak, the positive PC1 is moderate in history (Figs. 6c and 8c). The statistical analysis suggests that the striking inland intrusion of the WNPSH in September 2021 satisfies the universal interannual relationship between circulation and tropical SSTAs. However, the individual roles of SSTAs in the Indo-Pacific warm pool, equatorial eastern Pacific, and tropical Atlantic should be distinct in the westward extension of the WNPSH and the CNC rainfall. We will conduct some numerical experiments to identify the influences of AHS and the individual and combined effects of the SSTAs in September 2021.
Scatterplot of WEI and other variables during 1981–2021. (a) MC_AHS: normalized atmospheric heat source (AHS) over the Maritime Continent (10°S–10°N, 100°–160°E). (b) NIP_AHS: normalized AHS over the northern Indian Peninsula (10°–30°N, 60°–85°E). (c),(d) PC1 and PC2 of tropical SSTAs in the EOF analysis. (e) WP_SST: normalized SSTA in the Indo-Pacific warm pool (15°S–10°N, 90°–160°E). (f) TA_SST: normalized SSTA in the tropical Atlantic (15°S–15°N, 70°W–10°E). The linear regression is calculated using the records from 1981 to 2020, and the value on the top right of each panel indicates the correlation coefficient between the elements (one and two asterisks denote values exceeding the 95% and 99% confidence levels, respectively).
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
c. Numerical experiments
1) Influences of AHS
Three LBM experiments were designed to analyze the collaboration of AHS over the MC and NIP in September 2021. The mean flow was the climatological wind in September. In the MC and NIP AHS experiments, ideal diabatic heating anomalies with reasonable vertical profiles were added over the MC and NIP as observations (Fig. 9a). They both drove the LBM in the last experiment to examine their joint effect on the WNPSH. We integrated the LBM for 30 days in each experiment and considered the stationary atmospheric responses as the averaged outputs for the last 15 days.
Stationary atmospheric response to the AHS over MC and NIP in September 2021 in the LBM experiments. (a) Forcing of AHS in the LBM experiments. Blue and red lines are for the 400-hPa anomalous diabatic heating over the MC and ISM (values > 0 are plotted with an interval of 1.0 K day−1). The inset panels indicate the vertical profile of diabatic heating (K day−1) in the observation (black curves) and LBM experiments (red and blue curves). (b)–(d) The responses of the 500-hPa streamfunction (shading; 106 m2 s−1) and winds (vectors; m s−1) in the MC, NIP, and combined forcing experiments, respectively.
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
Despite the lack of moist feedback, the LBM experiments still capture the observed circulation features in September 2021. The enhanced diabatic heating over the MC first induces a pair of anomalous cyclones straddling the equator in the middle troposphere of the tropical Indian Ocean. In the meantime, the anticyclone strengthens over the western North Pacific and extends westward onto the East Asian continent (Fig. 9b). In addition, the midtropospheric anomalous cyclone deepens convection and intensifies diabatic heating over the NIP, strengthening the in situ cyclone and the anticyclone to its northeast (Fig. 9c). Finally, this anomalous anticyclone elongates eastward to South China to produce the deeper inland intrusion of the WNPSH, as shown by the experiment with both diabatic anomalies over the MC and NIP (Fig. 9d).
2) Effects of tropical SSTAs
We designed a series of AGCM experiments to validate the influences of tropical SSTAs. The global climatological SST with an annual cycle drives the AGCM in control (CTL) runs. Considering the SSTA intensity and persistence in 2021, the sensitivity runs were forced by the observed SSTAs in August–September 2021 over the Indo-Pacific warm pool (WP), tropical Atlantic (TA), tropical central and eastern Pacific (LN), and their various combinations (Table 2). We integrated each experiment for 40 years with different initial states. The significant differences of the last 35 members between the sensitivity and control runs denoted the robust atmospheric responses to the tropical SSTAs in September. Although the AGCM overestimates the CNC rainfall and WNPSH because of the uncertainty in conventional cumulus parameterization or the lack of air–sea interaction (Wang et al. 2005; Zhang and Chen 2016), the CTL experiment could reproduce the observed relationship between CNC rainfall and WNPSH.
Design of AGCM experiments.
The circulation and East Asian climate anomalies are distinct among the sensitivity experiments forced by sole SSTAs. In the WP experiment, the anomalous large-scale circulation ascends over the warm SSTAs in the Indo-Pacific warm pool and descends over the African continent (Fig. 10a). The airflow thus diverges over the warm pool to strengthen convection and diabatic heating over the MC and NIP in the mid- to upper troposphere (Fig. 10b). The westerly wind accelerates and maintains the equatorial-symmetric cyclone over the tropical central-eastern Pacific and the Atlantic Ocean, along with enhancing the easterly wind and anticyclone over the tropical western Pacific and the Indian Ocean. The anomaly of tropical easterly wind leads to the westward extension of the WNPSH and above-normal (below-normal) rainfall over the CNC (Southwest China) as the observation (Figs. 1a, 11a, and 11b). Meanwhile, the SAH expands eastward to favor warmer temperature and severe drought over Southwest China in autumn (Fig. 11c). In the TA experiment, anomalous large-scale ascending and tropical convection moves onto the warmer tropical Atlantic. The upper-level convergent and sinking flow prevails over the tropical Indian Ocean to suppress convection over the MC and NIP (Figs. 10c,d). As a Rossby-wave response, a couple of anticyclonic anomalies emerge in the upper troposphere of the eastern Pacific. They emanate toward the extratropics and form two teleconnections in the Northern and Southern Hemispheres. The northern branch propagates eastward along the EAJ to affect East Asia via a midlatitude wave train (Fig. 10d). The large internal variability of atmospheric circulation in the midlatitudes increases the uncertainty of East Asian rainfall and air temperature responses (Figs. 11d,f). Concurrently, the WNPSH shifts southward due to the thicker air column in the tropics, acting as a Kelvin wave response to warmer TA (Figs. 10d and 11e). In the LN experiment, sinking is enhanced to suppress tropical convection over the colder tropical eastern Pacific, along with anomalous ascending and convection centered over the tropical western Indian Ocean and Maritime Continent (Fig. 10e). As a Rossby wave response to the anomalies of tropical convection, the anomalous cyclones straddle the equator over the tropical central Pacific, and the anticyclone gets developed over East Asia (Fig. 10f). But we observe no significant responses of precipitation, air temperature, or primary circulation pattern over East Asia (Figs. 11g–i). The AGCM experiments forced by individual SSTAs indicate the dominant role of the warmer Indo-Pacific warm pool in westward extension of the WNPSH and more abundant CNC rainfall in September 2021.
Responses of large-scale circulation to the individual SSTA forcing in the critical domain in the AGCM experiment. (left) Responses of 150-hPa velocity potential (contours; 105 m2 s−1) and divergent wind (vectors; m s−1; values exceeding the 95% confidence level are in black) in the (a) WP, (c) TA, and (e) LN experiments. Shadings denote the SSTA forcing in each experiment (°C). (right) As in the left column, but for the responses of the 300-hPa streamfunction (shading; 106 m2 s−1; values exceeding the 95% confidence level are stippled) and 500-hPa diabatic heating (contours; K day−1).
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
Responses of circulation, precipitation, and 2-m air temperature to the individual forcing of SSTA in the (top) warm pool, (middle) tropical Atlantic, and (bottom) eastern Pacific in the AGCM experiment. (left) Anomalous precipitation (shading; mm day−1) and winds at 850 hPa (vectors; m s−1). (center) Anomalous geopotential height (shading; gpm) and winds at 500 hPa (vectors; m s−1). The black and blue lines are for the 5880-gpm contour of geopotential height in sensitivity and CTL runs. (right) Anomalous 2-m air temperature (shading; K) and winds at 150 hPa (vectors; m s−1). The black and blue lines are for the 14 280-gpm contour of geopotential height in sensitivity and CTL runs. Black vectors and stipples indicate values exceeding the 95% confidence level.
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
Furthermore, the combinations of tropical SSTAs in the three ocean basins could result in East Asian circulation and climate anomalies closer to the observations. Comparing with the WP experiment, the combined hotter warm pool and colder tropical eastern Pacific further develop anomalies of large-scale ascending and convection over the Maritime Continent by enlarging the zonal SSTA gradient between the tropical Indian and Pacific Oceans (Figs. 12a,b). The anomalous rainfall is thus significantly intensified in the CNC, along with the stronger WNPSH, enhanced Mongolian cyclone, and more eastward extension of SAH (Figs. 13a–c). Note that the anomalous rainband in North China displaces north of its observed position, implicating the exaggerated tropical influences in this experiment. When we attach warmer TA to the above experiment, the warmer TA tends to intensify the anticyclone around the Maritime Continent and decrease the large-scale meridional thermal contrast by stimulating a warm equatorial Kelvin wave (Fig. 12d). As a result, the convection anomalies remain over the MC and NIP, but its intensity attenuates over the NIP, corresponding to the weaker anomalies of large-scale ascent over the Indo-Pacific warm pool (Fig. 12c). In this experiment, the WNPSH is enhanced and anchored to its observed meridional position, along with a deeper trough over Northeast Asia and eastward extension of the SAH over hotter and drier South China (Figs. 13e,f). As the WNPSH intrudes inland, above-normal rainfall takes place in the lower reaches of the Huang River as observed (Fig. 13d). In the AGCM, the tropical Indo-Pacific–Atlantic SSTAs could reproduce the anomalous circulation and climate over East Asia in September 2021, despite some bias regarding the intensity and location of the Mongolian cyclone in the mid- to high latitudes.
As in Fig. 10, but indicating the responses to the combined SSTA forcing. (a),(b) Combined SSTA forcing in tropical Indo-Pacific warm pool and eastern Pacific. (c),(d) Combined SSTA forcing in tropical Pacific and Atlantic in September 2021.
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
As in Fig. 11, but indicating the responses to the combined SSTA forcing. (a),(b) SSTA forcing in tropical Indo-Pacific warm pool and eastern Pacific. (c),(d) SSTA forcing in tropical Pacific and Atlantic in September 2021.
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
In addition, we exhibit the responses of WNPSH and CNC rainfall to the tropical SSTAs in a probabilistic manner. Comparing with the CTL experiment, the ensemble response of the 500-hPa geopotential height area-averaged over South China and western Pacific is largest in the ALL experiment, followed by the TA, WP + LN, and WP experiments, in contrast to little increase of geopotential height over the region in the LN experiment (Fig. 14a). The probability distribution functions (PDFs) of the responses demonstrate that the combined Indo-Pacific–Atlantic SSTAs can most evidently improve the probability of geopotential height greater than 5880-gpm at 500 hPa over the region, while the individual influence of colder tropical eastern Pacific on the westward extension of the WNPSH is limited (Fig. 14b). Consistent with the WNPSH response, the ensemble response of the CNC rainfall amount is largest in the ALL but smallest in the LN and CTL experiment (Fig. 14c). In particular, the probability of extreme rainfall event improves greatly in the WP, WP + LN, and ALL experiment comparing with the CTL result (Fig. 14d). Although the warmer tropical Atlantic could induce evident westward extension of the WNPSH, the probability of above-normal CNC rainfall increases moderately, implicating larger noise due to the atmospheric internal variability in the TA experiment.
Probability of (a),(b) 500-hPa geopotential height (gpm) over 20°–30°N, 90°–120°E and (c),(d) rainfall (mm day−1) over the CNC (31°–41°N, 104°–120°E) responses to the tropical SST anomalies in different experiments. (a),(c) Box plot of the responses to different SSTA forcing (colored bars). (b),(d) Probability distribution functions (PDFs) of responses to different SSTA forcing (colorful curves). Black bar and curves indicate the results in the control run. Each experiment includes 35 members with a bootstrap method (1500 resamplings).
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
The SSTA pattern in September resembles that in August because of large thermal capacity in seawater. However, the responses of CNC rainfall present larger spreading among the members and less difference of ensemble mean between sensitivity experiments and control run in August (figure not shown). Since the atmospheric circulation associated with the CNC rainfall in August is significant in both tropics and midlatitude, the influences of tropical SSTAs on the CNC rainfall may be partly distorted by the stronger atmospheric internal variability in August.
d. Physically based statistical prediction model
The influences of SSTAs on the WNPSH allow us to build a physically based statistical model to predict the WEI in advance of one month. In this model, the training period is from 1982 to 2011, and the independent forecast period is from 2012 to 2021. We considered the predictors as PC1 and PC2 of the tropical SSTAs in August to forecast the WEI in September. The predictor of PC1 could well fit the WEI in the training period and show skillful prediction of WEI in the independent forecast period (Fig. 15a). When the predictors include both PC1 and PC2, the correlation coefficient between reconstructed and observed WEI increases in the training period. Meanwhile, the reforecast WEI shows a more significant correlation with observations in the independent forecast period (Fig. 15b). Therefore, the interannual variability of the combined effect of tropical SSTAs is dominant in the westward extension of WNPSH in early autumn, while the global warming trend could further amplify the anomalies of WNPSH by enlarging the warming in the Indo-Pacific warm pool and tropical Atlantic.
Statistical reforecast of the WEI based on the PC1 and PC2 of tropical SSTAs in August. Bars: observed WEI; solid red curve: reconstructed WEI in the training period of 1982–2011; blue dashed curve: predicted WEI in the independent forecast period of 2012–21.
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
6. Conclusions and discussion
In September 2021, the CNC rainfall broke its record since 1961 and extremely prolonged the rainy season in North China, giving rise to continuous floods in the lower reaches of the Yellow River. The present study ascribes the above-normal CNC rainfall to the westward extension of the WNPSH in the middle troposphere and suggests the crucial role of tropical Indo-Pacific–Atlantic SSTAs in this relationship. The conclusions are summarized as follows (Fig. 16).
Schematic diagram of combined effects of tropical Indo-Pacific–Atlantic SSTAs on extreme floods over the CNC in September 2021. (bottom) Horizontal distribution of the SSTAs. (middle) Atmospheric responses in the middle troposphere to the SSTAs, including a couple of cycle and anticyclone anomalies due to the tropical Indo-Pacific SSTAs and a Kelvin wave response to the tropical Atlantic SSTA. (top) Anomalous moisture transport and rainfall in China, presenting above-normal precipitation in the CNC and below-normal rainfall in South China. See the details in the text.
Citation: Journal of Climate 35, 18; 10.1175/JCLI-D-21-0988.1
a. Conclusions
The case study, statistical analyses, and numerical experiments demonstrate that the westward extension of the WNPSH in the middle troposphere determines the climate anomalies over East China in September. As the WNPSH intrudes into the East Asian continent, more abundant water vapor is transported to the CNC by the enhanced southwesterly wind west of the anomalous WNPSH. However, with less moisture convergence, South China experiences drier and hotter weather beneath the WNPSH.
The anomalous convection over the MC and NIP acts as the atmospheric forcing that extends the WNPSH westward in September. When convection intensifies over the MC, the midtropospheric cyclone strengthens over the northern Indian Ocean and NIP as a Rossby wave response to anomalous diabatic heating. Then, as a Kelvin wave response, the easterly wind accelerates over the tropical western Pacific to intensify the WNPSH and expand its western boundary to the East Asian continent. In the meantime, the anomalous cyclone over the NIP further deepens the local convection. It then releases more diabatic heating and induces additional anticyclonic anomalies over East Asia and the western Pacific, which leads the WNPSH to intrude inland over South China.
The linkage among anomalous tropical convection, the westward extension of the WNPSH, and climate anomalies over East Asia relies on the combined effects of tropical Indo-Pacific–Atlantic SSTAs as external forcings. The warm SSTAs in the Indo-Pacific warm pool are critical for the deeper convection over the MC and NIP and the westward extension of the WNPSH, which results in floods throughout the CNC and drought in South China. In contrast, the cold SSTAs in the tropical central-eastern Pacific could amplify the influences of warm pool SSTAs by increasing the zonal SSTA gradient around the MC. The warmer tropical Atlantic changes the tropical geopotential height over the Indo-Pacific warm pool via atmospheric Kelvin wave responses and finally modulates the intensity and meridional position of the WNPSH and CNC rainfall responses to the tropical Indian and Pacific SSTAs in September.
b. Discussion
The present study used the AGCM to identify the relative and combined effects of the SSTAs in the three basins on the WNPSH and CNC rainfall in September but did not consider the interaction among the three basins in the tropics (Cai et al. 2019; Wang 2019). It is still confusing why the weak La Niña accompanies the strong warming in the tropical Indo-Pacific warm pool and Atlantic in early autumn 2021, which cannot be explained by the remote SST responses to ENSO events (Alexander et al. 2002; Klein et al. 1999). More numerical experiments based on the coupled ocean–atmosphere model are essential to address this issue. Except for the anomalous WNPSH, although the midlatitude wave train is not well organized over East Asia in September, the CNC rainfall significantly correlates with the anomalous East Asian trough (EAT) in the upper troposphere, which has been simulated by the AGCM with the SSTAs in the three ocean basins (Figs. 3b and 12d). The TCC between the CNC rainfall regressed against both WEI and EAT index (defined as the 150-hPa streamfunction area-averaged over 30°–50°N, 140°–170°E) and the observation is +0.54 during 1981–2021. It exceeds the 99% confidence level and is more significant than the correlation between WEI-regressed CNC rainfall and observation (TCC = +0.42). The collaboration of EAT and WNPSH on the CNC rainfall in early autumn requires further study in the future. Furthermore, the extreme CNC rainfall in September 2021 also reflects the enhancement of the East Asian summer monsoon, which may be associated with the reduction in aerosols during the COVID-19 pandemic (He et al. 2022; Kripalani et al. 2022; Z. Wang et al. 2020). At the interdecadal time scale, the TCC between the WEI and CNC rainfall is +0.62 and exceeds the 90% confidence level. The CNC rainfall in September tends to increase after the late 1990s, consistent with the higher frequency of inland intrusion of the WNPSH defined by either WEI or other WNPSH indices. This is different from the eastward retreat of the WNPSH in boreal summer after the late 1970s (Huang et al. 2015), which implies the seasonal dependence of the interdecadal variability of the WNPSH. Consequently, the WNPSH behavior may link the long-term changes in the global climate pattern (e.g., Atlantic multidecadal oscillation, Pacific decadal oscillation, or interdecadal Pacific oscillation) with the length of the EASM rainy season in a warmer world. This specific physical process also deserves further investigation.
Acknowledgments.
This work was jointly funded by the National Natural Science Foundation of China (41830969), the Second Tibetan Plateau Scientific Expedition and Research (STEP) program (2019QZKK0105), the National Natural Science Foundation of China (42005131), the Basic Scientific Research and Operation Foundation of the Chinese Academy of Meteorological Sciences (CAMS) (2021Z004), and the S&T Development Foundation of CAMS (2020KJ009, 2020KJ012).
Data availability statement.
The in-situ rainfall records were downloaded from http://data.cma.cn/en/?r=site/index. The NOAA High-Resolution SST data were provided by the NOAA/OAR/ESRL PSL, Boulder, Colorado, USA, from https://psl.noaa.gov/data/gridded/data.noaa.oisst.v2.highres.html. The JRA-55 dataset was collected and provided under the Data Integration and Analysis System (DIAS) and developed and operated by a project supported by the Ministry of Education, Culture, Sports, Science and Technology.
REFERENCES
Alexander, M. A., I. Bladé, M. Newman, J. R. Lanzante, and N.-C. Lau, 2002: The atmospheric bridge: The influence of ENSO teleconnections on air–sea interaction over the global oceans. J. Climate, 15, 2205–2231, https://doi.org/10.1175/1520-0442(2002)015<2205:TABTIO>2.0.CO;2.
Bai, H., and W. Dong, 2004: Climate features and formation causes of autumn rain over southwest China (in Chinese). Plateau Meteor., 23, 884–889.
Bretherton, C. S., M. Widmann, V. P. Dymnikov, J. M. Wallace, and I. Bladé, 1999: The effective number of spatial degrees of freedom of a time-varying field. J. Climate, 12, 1990–2009, https://doi.org/10.1175/1520-0442(1999)012<1990:TENOSD>2.0.CO;2.
Cai, W., and Coauthors, 2019: Pantropical climate interactions. Science, 363, eaav4236, https://doi.org/10.1126/science.aav4236.
Chen, L., B. Zhang, and Y. Zhang, 2006: Progress in research on the East Asian monsoon (in Chinese). J. Appl. Meteor. Sci., 17, 711–724.
China Meteorological Administration, 2021: 2021 China Climate Bulletin, 62 pp.
Ding, Y., 2004: Seasonal march of the East-Asian summer monsoon. East Asian Monsoon, World Scientific, 3–53.
Ding, Y., and Z. Wang, 2008: A study of rainy seasons in China. Meteor. Atmos. Phys., 100, 121–138, https://doi.org/10.1007/s00703-008-0299-2.
Gao, Y., and Q. Guo, 1958: On the autumn raining area in China. Acta Meteor. Sin., 29, 264–273.
Gill, A. E., 1980: Some simple solutions for heat-induced tropical circulation. Quart. J. Roy. Meteor. Soc., 106, 447–462, https://doi.org/10.1002/qj.49710644905.
Gong, D., and S. Wang, 1999: Impacts of ENSO on rainfall of global land and China. Chin. Sci. Bull., 44, 852–857, https://doi.org/10.1007/BF02885036.
Harada, Y., and Coauthors, 2016: The JRA-55 reanalysis: Representation of atmospheric circulation and climate variability. J. Meteor. Soc. Japan, 94, 269–302, https://doi.org/10.2151/jmsj.2016-015.
He, C., T. Zhou, A. Lin, B. Wu, D. Gu, C. Li, and B. Zheng, 2015: Enhanced or weakened Western North Pacific subtropical high under global warming? Sci. Rep., 5, 16771, https://doi.org/10.1038/srep16771.
He, C., A. Lin, D. Gu, C. Li, B. Zheng, B. Wu, and T. Zhou, 2018: Using eddy geopotential height to measure the western North Pacific subtropical high in a warming climate. Theor. Appl. Climatol., 131, 681–691, https://doi.org/10.1007/s00704-016-2001-9.
He, C., W. Zhou, T. Li, T. Zhou, and Y. Wang, 2022: East Asian summer monsoon enhanced by COVID-19. Climate Dyn., https://doi.org/10.1007/s00382-022-06247-8.
He, M., 1984: Distribution and long-term forecast of Chinese autumn rainfalls. Meteor. Mon., 10, 10–13.
Hu, J., W. Wang, and Y. Tan, 2021: Contrasting impacts of three types of ENSO event on boreal autumn rainfall over Southwest China. J. Geosci. Environ. Prot., 9, 14–27, https://doi.org/10.4236/gep.2021.910002.
Huang, B., and Coauthors, 2021: Improvements of the daily optimum interpolation sea surface temperature (DOISST) version 2.1. J. Climate, 34, 2923–2939, https://doi.org/10.1175/JCLI-D-20-0166.1.
Huang, Y., H. Wang, K. Fan, and Y. Gao, 2015: The western Pacific subtropical high after the 1970s: Westward or eastward shift? Climate Dyn., 44, 2035–2047, https://doi.org/10.1007/s00382-014-2194-5.
Klein, S. A., B. J. Soden, and N.-C. Lau, 1999: Remote sea surface temperature variations during ENSO: Evidence for a tropical atmospheric bridge. J. Climate, 12, 917–932, https://doi.org/10.1175/1520-0442(1999)012<0917:RSSTVD>2.0.CO;2.
Kobayashi, S., Y. Ota, and Y. Harada, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 5–48, https://doi.org/10.2151/jmsj.2015-001.
Kripalani, R., K. J. Ha, C. H. Ho, J. H. Oh, B. Preethi, M. Mujumdar, and A. Prabhu, 2022: Erratic Asian summer monsoon 2020: COVID-19 lockdown initiatives possible cause for these episodes? Climate Dyn., https://doi.org/10.1007/s00382-021-06042-x.
Kuang, X., and Y. Zhang, 2005: Seasonal variation of the East Asian subtropical westerly jet and its association with the heating field over East Asia. Adv. Atmos. Sci., 22, 831–840, https://doi.org/10.1007/BF02918683.
Li, C., and T. Zhao, 2019: Seasonal responses of precipitation in China to El Niño and positive Indian Ocean dipole modes. Atmosphere, 10, 372, https://doi.org/10.3390/atmos10070372.
Li, H.-X., H.-P. Chen, and H.-J. Wang, 2017: Influence of North Pacific SST on heavy precipitation events in autumn over North China. Atmos. Oceanic Sci. Lett., 10, 21–28, https://doi.org/10.1080/16742834.2017.1237256.
Li, T., B. Wang, B. Wu, T. Zhou, C.-P. Chang, and R. Zhang, 2017: Theories on formation of an anomalous anticyclone in western North Pacific during El Niño: A review. J. Meteor. Res., 31, 987–1006, https://doi.org/10.1007/s13351-017-7147-6.
Liu, Y., P. Liang, and Y. Sun, 2019: Characteristics of the western Pacific subtropical high and summer rainfall anomalies. The Asian Summer Monsoon, Y. Liu, P. Liang, and Y. Sun, Eds., Elsevier, 85–95.
Luo, X., D. Li, and H. Wang, 2013: New evolution features of autumn rainfall in west China and its responses to atmospheric circulation (in Chinese). Plateau Meteor., 32, 1019–1031.
Qian, W., Y. Zhu, A. Xie, and Q. Ye, 1998: Seasonal and interannual variations of upper tropospheric water vapor band brightness temperature over the global monsoon regions. Adv. Atmos. Sci., 15, 337–345, https://doi.org/10.1007/s00376-998-0005-9.
Qian, W., H. S. Kang, and D. K. Lee, 2002: Distribution of seasonal rainfall in the East Asian monsoon region. Theor. Appl. Climatol., 73, 151–168, https://doi.org/10.1007/s00704-002-0679-3.
Qin, M., D. Li, A. Dai, W. Hua, and H. Ma, 2018: The influence of the Pacific decadal oscillation on North Central China precipitation during boreal autumn. Int. J. Climatol., 38, e821–e831, https://doi.org/10.1002/joc.5410.
Qiu, Y., W. Cai, X. Guo, and B. Ng, 2014: The asymmetric influence of the positive and negative IOD events on China’s rainfall. Sci. Rep., 4, 4943, https://doi.org/10.1038/srep04943.
Reynolds, R. W., T. M. Smith, C. Y. Liu, D. B. Chelton, K. S. Casey, and M. G. Schlax, 2007: Daily high-resolution-blended analyses for sea surface temperature. J. Climate, 20, 5473–5496, https://doi.org/10.1175/2007JCLI1824.1.
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, https://doi.org/10.1175/1520-0469(2001)058<0608:AFOAPI>2.0.CO;2.
Wang, B., and LinHo, 2002: Rainy season of the Asian–Pacific summer monsoon. J. Climate, 15, 386–398, https://doi.org/10.1175/1520-0442(2002)015<0386:RSOTAP>2.0.CO;2.
Wang, B., and Q. Ding, 2008: Global monsoon: Dominant mode of annual variation in the tropics. Dyn. Atmos. Oceans, 44, 165–183, https://doi.org/10.1016/j.dynatmoce.2007.05.002.
Wang, B., Q. Ding, X. Fu, I.-S. Kang, K. Jin, J. Shukla, and F. Doblas-Reyes, 2005: Fundamental challenge in simulation and prediction of summer monsoon rainfall. Geophys. Res. Lett., 32, L15711, https://doi.org/10.1029/2005GL022734.
Wang, B., B. Xiang, and J.-Y. Lee, 2013: Subtropical high predictability establishes a promising way for monsoon and tropical storm predictions. Proc. Natl. Acad. Sci. USA, 110, 2718–2722, https://doi.org/10.1073/pnas.1214626110.
Wang, B., X. Luo, and J. Liu, 2020: How robust is the Asian precipitation–ENSO relationship during the industrial warming period (1901–2017)? J. Climate, 33, 2779–2792, https://doi.org/10.1175/JCLI-D-19-0630.1.
Wang, C., 2019: Three-ocean interactions and climate variability: A review and perspective. Climate Dyn., 53, 5119–5136, https://doi.org/10.1007/s00382-019-04930-x.
Wang, Z., and B. Zhou, 2019: Observed decadal transition in trend of autumn rainfall over central China in the late 1990s. J. Climate, 32, 1395–1409, https://doi.org/10.1175/JCLI-D-18-0112.1.
Wang, Z., J. Mu, M. Yang, and X. Yu, 2020: Reexamining the mechanisms of East Asian summer monsoon changes in response to non–East Asian anthropogenic aerosol forcing. J. Climate, 33, 2929–2944, https://doi.org/10.1175/JCLI-D-19-0550.1.
Watanabe, M., and M. Kimoto, 2000: Atmosphere–ocean thermal coupling in the North Atlantic: A positive feedback. Quart. J. Roy. Meteor. Soc., 126, 3343–3369, https://doi.org/10.1002/qj.49712657017.
Worley, P. H., and J. B. Drake, 2005: Performance portability in the physical parameterizations of the Community Atmospheric Model. Int. J. High Perform. Comput. Appl., 19, 187–201, https://doi.org/10.1177/1094342005056095.
Xie, S.-P., Y. Kosaka, Y. Du, K. Hu, J. S. Chowdary, and G. Huang, 2016: Indo-Western Pacific Ocean capacitor and coherent climate anomalies in post-ENSO summer: A review. Adv. Atmos. Sci., 33, 411–432, https://doi.org/10.1007/s00376-015-5192-6.
Xu, H.-L., J. Feng, and C. Sun, 2013: Impact of preceding summer North Atlantic Oscillation on early autumn precipitation over central China. Atmos. Oceanic Sci. Lett., 6, 417–422, https://doi.org/10.1080/16742834.2013.11447118.
Yang, Q., Z. Ma, and B. Xu, 2016: Modulation of monthly precipitation patterns over East China by the Pacific decadal oscillation. Climatic Change, 144, 405–417, https://doi.org/10.1007/s10584-016-1662-9.
Yeh, T.-C., S.-Y. Tao, and M.-T. Li, 1958: The abrupt change of circulation over the Northern Hemisphere during June and October (in Chinese). Acta Meteor. Sin., 29, 249–263.
Zeng, X., and E. Lu, 2004: Globally unified monsoon onset and retreat indexes. J. Climate, 17, 2241–2248, https://doi.org/10.1175/1520-0442(2004)017<2241:GUMOAR>2.0.CO;2.
Zhang, S., and B. Wang, 2008: Global summer monsoon rainy seasons. Int. J. Climatol., 28, 1563–1578, https://doi.org/10.1002/joc.1659.
Zhang, Y., and H. Chen, 2016: Comparing CAM5 and superparameterized CAM5 simulations of summer precipitation characteristics over continental East Asia: Mean state, frequency–intensity relationship, diurnal cycle, and influencing factors. J. Climate, 29, 1067–1089, https://doi.org/10.1175/JCLI-D-15-0342.1.
Zhang, Y., X. Kuang, W. Guo, and T. Zhou, 2006: Seasonal evolution of the upper-tropospheric westerly jet core over East Asia. Geophys. Res. Lett., 33, L11708, https://doi.org/10.1029/2006GL026377.
Zhou, B., and Z. Wang, 2019: Enlightenment from heavy autumn rain of West China in 2017: Synergic role of atmospheric circulation at mid-high latitudes and oceanic background. Theor. Appl. Climatol., 138, 263–274, https://doi.org/10.1007/s00704-019-02809-9.
Zhu, C., W.-S. Lee, H. Kang, and C.-K. Park, 2005: A proper monsoon index for seasonal and interannual variations of the East Asian monsoon. Geophys. Res. Lett., 32, L02811, https://doi.org/10.1029/2004GL021295.
Zhu, Z., R. Lu, H. Yan, W. Li, T. Li, and J. He, 2020: Dynamic origin of the interannual variability of West China autumn rainfall. J. Climate, 33, 9643–9652, https://doi.org/10.1175/JCLI-D-20-0097.1.
