A severe drought struck southwest China during autumn 2009, which had a huge impact on productivity and the lives of the affected population. A nonconventional El Niño, the so-called warm pool (WP) El Niño, was supposed to be a principal factor of this strong autumn drought. In sharp contrast to a conventional El Niño, in the 2009 WP El Niño year the maximum sea surface temperature (SST) anomalies are confined to the central equatorial Pacific Ocean. Moreover, this WP El Niño was characterized by the relatively farther westward location and the strongest intensity among the WP El Niño events in the past 60 years. Observations and modeling studies both indicate that the rainfall deficits over southwest China are significantly influenced by the combined effects of the location and intensity of the WP El Niño. That is, the drought over southwest China tends to be more severe when the warming SST anomalies associated with the WP El Niño are located farther westward and are stronger. Therefore, the strong autumn drought over southwest China in 2009 can be largely attributed to the concurrent distinctive WP El Niño, which generates a strongly anomalous cyclone over the west North Pacific and leads to a serious reduction in rainfall over southwest China. The influence of the Indian Ocean warming on autumn rainfall over southwest China was also examined but seems to have little contribution to this drought.
Autumn rainfall over southwest China makes a large contribution (above 20%) to total annual rainfall (Wu and Hu 2003) and is a key influence on the timing of crop maturity and harvest. The destruction caused by an autumn drought can be catastrophic. A strong autumn drought occurred over southwest China in 2009 (Fig. 1b), seriously affecting more than 80% of the vegetation ecosystem in Yunnan, Guangxi, and Guizhou provinces (22°–30°N, 100°–110°E) (Wang et al. 2010; Barriopedro et al. 2012). More than 60 million residents suffered drinking-water shortages, and more than 1 million ha of crops died. A better understanding of the mechanisms that led to this disastrous drought would be useful for future drought management and planning.
During the disastrous autumn of 2009, a strong warming in the sea surface temperature (SST) appeared over the central equatorial Pacific Ocean, and it was identified as an El Niño autumn on the basis of the Niño-3.4 (5°S–5°N, 120°–170°W) SST anomaly (SSTA) by the Climate Prediction Center (CPC). Numerous efforts have been made to explore possible ENSO impacts on the East Asian climate (e.g., Zhang et al. 1996; Chang et al. 2000a,b; B. Wang et al. 2000, 2008; Wu and Hu 2003; Niu and Li 2008; L. Wang et al. 2008; Wu et al. 2009; Li et al. 2010, 2011a,b; Wu et al. 2012). These efforts were often focused on the conventional (or canonical) El Niño events, with the warm SSTA center mainly located over the east equatorial Pacific (Rasmusson and Carpenter 1982; Wallace et al. 1998). However, the warming of the SSTA during autumn 2009, in sharp contrast with conventional El Niño events, shifted westward toward the warm pool rim (Lee and McPhaden 2010). The consensus reached by earlier studies defines this tropical Pacific phenomenon as a new type of El Niño (Larkin and Harrison 2005a; Ashok et al. 2007; Yu and Kao 2007; Kao and Yu 2009; Kug et al. 2009; Yeh et al. 2009).
Larkin and Harrison (2005a) identified a number of these new type of El Niño events using a new El Niño definition from the National Oceanic and Atmospheric Administration (NOAA), and refer to them as a date line El Niño because the maximum SSTA is located near the date line. The new type of El Niño may also be defined according to the second empirical orthogonal function (EOF) mode and called El Niño Modoki (Ashok et al. 2007). Kao and Yu (2009) referred to the new type of El Niño as the central Pacific El Niño. Kug et al. (2009) separated the new type of El Niño events from the conventional El Niño events based on the spatial patterns of the SSTA and named them a warm pool (WP) El Niño. Many studies have shown that the new type of El Niño has a different influence on atmospheric circulation compared to the conventional El Niño (Larkin and Harrison 2005a,b; Wang and Hendon 2007; Weng et al. 2007, 2009; Cai and Cowan 2009; Kim et al. 2009; Taschetto and England 2009; Chen and Tam 2010; Feng et al. 2010; Mo 2010; Feng and Li 2011; Lee et al. 2010; Zhang et al. 2011; Wang and Wang 2013; Xie et al. 2012). Despite the variety of names and definitions used in these studies, an emphasis was placed on the difference between the two types of El Niño. In this paper, the new type of El Niño is referred to as the WP El Niño, following Kug et al. (2009) and Ren and Jin (2011).
The WP El Niño is very different from the conventional El Niño and contributed to an autumn drought over south China (Zhang et al. 2011). During the WP El Niño, an anomalous cyclone occurs over the west North Pacific (WNP) and a northeasterly wind anomaly forms over East Asia, leading to autumn rainfall deficits over south China. However, there remains the question of why the autumn drought over southwest China was so strong during autumn 2009 compared to other WP El Niño events. Whether this WP El Niño involved some unusual characteristics, or whether other possible abnormalities played an additional role, is unclear. In this study, the linkage between the WP El Niño and the autumn drought over southwest China is examined through intercomparison with other WP El Niño events and modeling experiments. It is our hypothesis that the 2009 WP El Niño, with relatively farther westward location and the strongest intensity among the WP El Niño events that occurred in the past 60 years, was largely responsible for this strong autumn drought.
In the remainder of the paper, section 2 describes data, the definition of WP El Niño events, and experimental designs used in this study. Section 3 illustrates the strong autumn drought of 2009 in terms of the atmospheric and oceanic anomalies. In section 4, the distinctive nature of the 2009 El Niño is discussed, with a focus on the intensity and location of the warm SSTA center, and modeling experiments are performed to investigate the effects related to the intensity and location of the warm SSTA center. The major findings are summarized in section 5, and this section also discusses the possible contribution of the Indian Ocean SSTA in this drought event.
2. Data and methods
The monthly station rainfall dataset (1951–2009) from the China Meteorological Administration was used to analyze such strong drought in autumn [September–November (SON)] 2009. The SST over the tropical Pacific and Indian Oceans was investigated based on the monthly global SST data from the Hadley Centre Sea Ice and Sea Surface Temperature dataset (HadISST), which was provided by the Met Office Hadley Centre (Rayner et al. 2003). Anomalous circulations were identified using the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis data (Kalnay et al. 1996). Anomalies for all variables were conducted as the deviation from the 30-yr climatological mean (1961–90).
Adopting the approach of our previous work (Zhang et al. 2011), the autumns of 1969, 1977, 1991, 1994, 2002, 2003, 2004, 2006, and 2009 were identified as WP El Niño autumns from all 18 El Niño autumns issued by the CPC, which were defined on the basis of the location of the maximum SSTA center (i.e., whether the event has larger SSTA in the central Pacific to the west of 150°W). Discrepancies exist in the definition of the WP El Niño cases between our work and earlier studies (e.g., Kim et al. 2009; Kug et al. 2009; Yeh et al. 2009), and two reasons are considered here. First, most of the previous studies defined the WP El Niño events based on the winter SSTA. However, the autumn season is the focus here. Second, five events (1969, 1991, 1994, 2002, and 2004) were identified by Kim et al. (2009) based on the autumn SSTA. The differences between our definition and their cases are the autumns of 1977, 2003, 2006, and 2009. As shown in Fig. 2, warm SSTA centers of up to 0.5°C appear over the central or west equatorial Pacific during these four autumns (Figs. 2a–d). Their associated convective anomaly centers indicated by the convergence of water vapor (integrated from the surface to 300 hPa), mainly emerge over the west equatorial Pacific to the west of the positive SSTA center (Figs. 2e–h) where there is a high basic SST (Wallace et al. 1998; Kug et al. 2009). It is similar to that of the WP El Niño, but significantly different from that of the conventional El Niño (Ashok et al. 2007; Weng et al. 2007; Kug et al. 2009). Although SSTA also shows another warm anomaly greater than 1.0°C over the far eastern equatorial Pacific during autumn 2006 (Fig. 2c), the equatorial convection anomaly mainly emerges to the west of the date line (Fig. 2g). This is because the central Pacific warming is much more effective at forming deep convection than the eastern Pacific warming because of a higher background SST. Because the warm SSTA forces the atmosphere through the convection, the location of convection can be used as another effective indicator when attempting to define WP El Niño events.
Composite and regression analyses were used to investigate climatic impacts associated with the WP El Niño. In this study, our interest is why the autumn drought was so strong in autumn 2009 compared to other WP El Niño cases, so the 2009 WP El Niño event was excluded from the composite analyses. All statistical significance tests were performed using the two-tailed Student's t test.
The atmospheric general model used here is the NCAR Community Atmosphere Model, version 3 (CAM3) (Collins et al. 2004). The horizontal resolution is T42 (approximately 2.8° longitude × 2.8° latitude), with 26 hybrid vertical levels. First, climatological (seasonal varying) SST was used to force the model and integrate 5 years. Then, 10 independent atmospheric fields were selected as the initial conditions from the simulated results in July of the fifth year, and 10 ensemble members were calculated for each experiment. Four simulations were performed to verify the respective contribution from the SSTA location and intensity. In the first simulation [control (CTRL)], the SST was prescribed as the climatological SST. In the second simulation [eastern warming modeling (EWM)], an anomalous SST warming center was added at 165°W, with the maximum SSTA reaching 1.5°C. In the third simulation [western warming modeling (WWM)], the anomalous SST warming was shifted westward by 10°. The warming center was enhanced to 2.0°C in the fourth simulation [enhanced western warming modeling (EWWM)] relative to the third experiment. Therefore, a comparison of the EWM and WWM runs may identify the contribution of the SSTA location, while a comparison of the WWM and EWWM may indicate the effect of the SSTA intensity. Each run was initialized on 1 July and integrated for 5 months. The ensemble mean for the autumn (SON) was calculated for analysis.
3. Ocean and atmospheric features associated with the autumn drought of 2009
Figure 1c displays the standard time series of the autumn rainfall anomaly over southwest China (22°–30°N, 100°–110°E) between 1951 and 2009. It is clear that the most severe autumn drought for 60 years was recorded in 2009, and the associated rainfall anomaly exceeds two standard deviations. Although a slight decreasing trend can be detected in the time series, it can be discounted here because the autumn rainfall deficit in 2009 remains the largest even after detrending (figure omitted). The rainfall deficit exceeded 50% in most regions of southwest China (Fig. 1b), and in some regions, such as the east part of Yunnan province, the rainfall deficits reached 80%; deficits of 90% were previously reported in some regions of Yunnan province (Huang et al. 2011). In addition, an abnormally high temperature appears over south China, which increases evaporation and exacerbates the local drought condition (Lu et al. 2011).
In the climatological autumn (Fig. 3a), an anticyclone covers south China. The easterly wind to its south, and the southwesterly wind to its west, converge over south China and carry large amounts of water vapor from the tropical oceans. These two moisture transport branches supply the autumn rainfall over south China. One is from the Bay of Bengal and the other is from the South China Sea, but they originate from the Indian Ocean and the west tropical Pacific, respectively. Sufficient water vapor leads to a relatively wet season over south China, with accumulated rainfall of up to 200 mm in most regions and over 300 mm in some areas (Fig. 1a).
However, in autumn 2009, a strong cyclonic anomaly occurred over the WNP with anomalous low pressure exceeding −1.5 hPa (Fig. 3b). In the northwest section of the anomalous WNP cyclone, an anomalously southwestward moisture transport prevailed over south China (Fig. 3b), which inhibited the northeastward transport of water vapor from the south. To the northeast and southwest of the WNP cyclone anomaly, there was a strong anticyclone anomaly to the east of Japan and a weak anticyclone anomaly over the Indian Ocean.
To quantitatively explore the moisture budget, the moisture flux at four boundaries of southwest China is calculated as shown in Fig. 4. During autumn 2009, the vertically integrated moisture divergence reaches as high as 1.32 mm day−1, which is almost the same as the value of rainfall deficits (Fig. 4). It indicates that the autumn rainfall deficits are mainly caused by the moisture transport anomalies. From Fig. 4, the moisture fluxes from the south (QVS) and east (QUE) are strongly decreased and increased, respectively. The moisture flux anomalies at the west and north boundaries are much weaker than those at the south and east boundaries. The westward transport anomalies from the east (QUE) are in favor of moisture increasing over southwest China. So, the sharp reduction of moisture flux at the south boundary (QVS) is primarily responsible for the rainfall deficit because the northward transport of moisture from the south is inhibited by the anomalous northerly.
In addition, the net radiative flux into the atmospheric column is also calculated here because it is pointed out that the radiative cooling may also play some role on the regional drought (Sooraj et al. 2013). The net radiative flux (FNET) is the difference between the net flux at the top of the atmosphere and that at the surface following Sooraj et al. (2013). During autumn 2009, anomalous net flux shows a radiative cooling with a value of −83 W m−2. Sooraj et al. (2013) indicated that the atmospheric radiative cooling can enhance the local rainfall deficits through strengthening the anomalous divergent circulation. Therefore, the radiative cooling also makes a contribution to maintaining the autumn drought of 2009 to some degree.
Simultaneously, a remarkable warming of SST appeared in the tropical Pacific, with the maximum SST anomalies located in the central equatorial Pacific rather than the eastern equatorial Pacific (Fig. 2d). This warming phenomenon has been identified as a WP El Niño event. In our previous work, we demonstrated that the WP El Niño can produce a cyclonic anomaly over the WNP with its center located near 20°N, 130°E, and this causes an autumn rainfall deficit over south China (Zhang et al. 2011). As shown in Fig. 3b, a similar teleconnection pattern developed during autumn 2009, indicating that the strong autumn drought of 2009 may be attributed to the WP El Niño event.
Figure 5 shows the anomalous rainfall, wind at 925 hPa, and sea level pressure (SLP) in autumn 2009 and the composite of eight previous WP El Niño events, as well as the respective differences. Rainfall over southwest China is less than normal during a WP El Niño autumn (Fig. 5b). It is noticeable that the autumn rainfall deficits in 2009 are much greater than those of the previous eight WP El Niño autumns. The autumn rainfall deficits in 2009 exceed 1.5 mm day−1 over most regions of southwest China (Fig. 5a). The largest difference between 2009 and the other events occurs over southwest China (Fig. 5c).
The SLP anomaly over the WNP was about −1.5 hPa during autumn 2009 (Fig. 5d), whereas the composite SLP anomaly is only one-third of that in 2009 (Fig. 5e). The large SLP difference occurs mainly over the South China Sea (Fig. 5f). An anomalous cyclone appears over the South China Sea and south China in the wind difference. The northerly anomaly during autumn 2009 is 0.19 m s−1 stronger than the composition over southwest China, which may cause more severe rainfall deficits (Fig. 5f).
These observations suggest that the severe drought of autumn 2009 was associated with a strong atmospheric anomaly, which was probably caused by the concurrent WP El Niño. A question remains: How could the 2009 WP El Niño event result in such strong atmospheric anomalies? In the next section, the features of the WP El Niño in 2009 are considered in terms of its intensity and location through a comparison with other WP El Niño events.
4. Distinctive features of the 2009 WP El Niño
Comparing Fig. 2d with Fig. 6a, the SSTA pattern in autumn 2009 is very similar to the composite of the previous eight WP El Niño autumns, but the warming center is much stronger and farther to the west. Simultaneously, a strong moisture convergence anomaly associated with the warm SSTA appeared over the western equatorial Pacific, to the west of the date line, during autumn 2009, whose amplitude and zonal scale were much greater than those of the composite of the earlier WP El Niño events (Figs. 2h, 6b). Figures 6c and 6d display the SSTA patterns along the equatorial Pacific and Niño-4 indices [SSTA in the region (5°S–5°N, 160°E–150°W)] for the nine WP El Niño episodes. As shown in Fig. 6c, the maximum SSTA is confined to the central equatorial Pacific, and the values decrease to the east and west. The rate of decrease in the size of the SSTA along the equator is much slower to the east than to the west. Among all of the WP El Niño events, the SSTA intensity of the 2009 event is the highest based on the Niño-4 index (Fig. 6d). A higher-intensity WP El Niño tends to result in a more severe drought over southwest China because it can produce stronger convection over the western equatorial Pacific and, thus, a stronger anomalous WNP cyclone by means of a Rossby wave response (Gill 1980). To investigate the relationship between the intensity of the WP El Niño and the autumn rainfall anomalies over southwest China, we roughly divided the nine WP El Niño events into two groups according to their intensity as indicated by the Niño-4 index (Fig. 7a). The stronger events were in 1994, 2002, 2004, 2006, and 2009, while the weaker events comprise 1969, 1977, 1991, and 2003. The composite rainfall deficit tends to be more severe for the stronger WP El Niño events than that for the weaker events. It seems that the Niño-4 index could be considered as one factor that affects the autumn rainfall response to the WP El Niño over southwest China. However, although the Niño-4 indices in 1969 and 2003 are the two lowest among these WP El Niño events, the autumn rainfall anomalies during these two autumns are similar to those in the relatively high Niño-4 index years, such as 2002 and 2004 (Fig. 7a). It can be seen that the intensity alone does not adequately represent the complicated relationship between the WP El Niño and autumn rainfall anomalies over southwest China. Other factors need to be taken into consideration.
Based on the SSTA patterns along the equator for the nine WP El Niño events (Fig. 6c), these events naturally fall into two separate groups according to their position of the west boundaries of the maximum SSTA near 165°E. The group with the west edge of the SSTA located farther westward includes the years 1969, 2003, 2004, and 2009. In the other group (1977, 1991, 1994, 2002, and 2006), the west edge of the SSTA appears to shift relatively eastward. To better describe the two groups, the longitude of the maximum zonal gradient of SSTA (; °E) is defined as a measure of the position of the warming SST anomaly. The locations defined here are consistent with those by the zonal location of the maximum SSTA (not shown). Based on this definition, the WP El Niño events can also be classified into two groups, as shown in Fig. 7b. For the group to the west of 163°E, their composite rainfall deficit over southwest China is much larger than that for the group to the east of 163°E. Therefore, the response of the autumn rainfall over southwest China to the WP El Niño over southwest China also depends on the location of warming SSTA, as well as its intensity.
Consequently, to better demonstrate the relationship between the WP El Niño and the autumn rainfall anomaly over southwest China, both the location and intensity of the WP El Niño should be considered. Combining the two normalized factors (i.e., the location and intensity of the SSTA), an empirical formula for y (the normalized autumn rainfall anomaly over southwest China) was constructed based on a linear regression analysis:
where °E, °C, denotes the Niño-4 index, and and indicate the standard deviations of (°E) and (°C), respectively. The correlation between the regressed and observed rainfall anomalies reaches as high as 0.75 (Fig. 8). This indicates that the location and intensity of the SSTA during the WP El Niño autumn collectively contribute to the autumn drought over southwest China. The autumn drought tends to be more severe when the warming SSTA of the WP El Niño is located farther westward and is stronger. During the autumn of 2009, the SST anomalies were located farther westward, and the intensity was the strongest of the nine WP El Niño events. Therefore, the 2009 WP El Niño, with its distinct characteristics in terms of location and intensity, was possibly responsible for this strong drought over southwest China.
b. Model simulations
According to the empirical formula above, the serious drought of 2009 seems to be attributed to the collective effect of the SSTA location and intensity. To verify the respective contribution of these two factors, three experiments were designed and performed in which positive SST anomalies in the central equatorial Pacific, with different locations and intensities, were added to the SST climatology (Fig. 9). The experimental design has been described in section 2.
Figure 10 shows differences in low-level wind, SLP, and rainfall in the EWM, WWM, and EWWM runs relative to the CTRL run. All three simulations clearly capture the general features of the observed anomalies associated with the central Pacific warming during autumns (Figs. 1–3), including rainfall increasing over the western equatorial Pacific and decreasing over south China, the anomalous cyclone over the WNP, and the anomalous anticyclone to the east of Japan. This indicates that the warming SSTA in the central Pacific can produce an autumn drought over southwest China. Under the forcing of the central Pacific warming (Fig. 10), rainfall increases over the western equatorial Pacific, which can induce an anomalous cyclone over the WNP through a Rossby wave response (Gill 1980). The northeasterly wind anomalies (in the northwest section of the anomalous cyclone) that prevail over south China inhibit the transport of moist air and cause the rainfall deficits there. It is noteworthy that the response in Figs. 10b,c is much stronger than that in Fig. 3b. It is possibly because the SSTA specified in the simulations is stronger than that in the observation. Over the western and central tropical Pacific, a small change in SST could cause a large response in atmosphere owing to a high background SST.
Over the Indian Ocean, the simulated wind anomalies in the three experiments differ from the observations. This is because no SST forcing was added in the Indian Ocean in the simulations, whereas warming was seen in the observational data. The possible effect of the SSTA in the Indian Ocean will be discussed in the next section.
Comparing the EWM and WWM runs, it can be seen that the atmospheric response over the WNP to the central Pacific warming is very sensitive to the location of the SSTA. The more westward SSTA forcing produces a much larger atmospheric response over the WNP (Figs. 10a,b), which is possibly associated with the climatological basic states of ocean and atmosphere (Hirota and Takahashi 2012). The anomalous SLP over the WNP in the WWM run is almost 3 times as strong as that in the EWM run. Correspondingly, stronger northeasterly wind anomalies over east China in the WWM run lead to larger rainfall deficits over southwest China. The rainfall deficit over southwest China in the WWM run reaches −0.79 mm day−1, nearly 4 times as much as that in the EWM run (Table 1). Relative to the EWM run, the rainfall in the WWM run decreases by 28%. Therefore, a more westward location of the central Pacific warming can cause much greater rainfall deficits over southwest China.
When the SSTA warming is enhanced in the EWWM run, the atmospheric responses over the WNP become stronger relative to the WWM run (Figs. 10b,c). In the EWWM run, the anomalous SLP over the WNP is as high as −7 hPa, and the intense cyclone anomalies lead to enhanced northeasterly anomalies over east China. The sharp response to the warm SSTA over the central Pacific gives rise to a rainfall deficit up to 54% over south China relative to the CTRL run (Table 1).
Comparisons among the three experiments indicate that the location and intensity of the warming SSTA over the central Pacific are both of importance for autumn rainfall over southwest China. Therefore, the greater westward displacement of the WP El Niño tends to produce more severe autumn rainfall deficits over southwest China as a result of the stronger atmospheric anomalies, while at the same time, the stronger warming SSTA of the WP El Niño enhances the anomalous cyclone over the WNP, and this also favors more severe autumn drought over southwest China.
Furthermore, a series of experiments are conducted under the forcings of the observed SST anomalies over the central Pacific (20°S–20°N, 140°E–120°W) during the nine WP El Niño autumns. The experimental designs are similar to the idealized simulations. The modeling results (not shown) show that all of the central Pacific warming can produce autumn rainfall deficits over southwest China through a similar atmospheric response, as shown in Fig. 10. Again, this shows that the SST warming in the central Pacific can cause an anomalous cyclone over the WNP and, thus, an autumn drought over southwest China. The composite rainfall anomalies over southwest China are −0.35 mm day−1 under the forcings of the realistic central Pacific SST warming during the eight previous WP El Niño autumns, which is comparable to the observation (−0.4 mm day−1). The simulated rainfall deficits are weaker than the observation during the 2009 WP El Niño autumn. This suggests that other factors may also have some effect on the serious autumn drought of 2009; for example, the higher-than-normal local temperature is not considered in the simulations. Among the nine events, the rainfall deficits are the strongest under the forcing of the central Pacific warming associated with the 2009 event (Table 2), which is the same as the observation. The simulations reproduce well the extremity of the rainfall deficits during autumn 2009 compared to those during the other eight previous WP El Niño events. Therefore, in 2009, the combined effects of the highest-intensity WP El Niño and its farther westward location among the nine WP El Niño events recorded very likely led to the severe autumn drought over southwest China.
5. Summary and discussion
During autumn 2009, a severe drought struck southwest China, and rainfall deficits were as high as 80%, causing major economic losses and drinking water shortages. A possible mechanism for this strong drought has been demonstrated in this study.
The WP El Niño, an SST warming event in which the maximum SSTA is confined to the central equatorial Pacific, had an important contribution to this severe drought over southwest China. The warming SSTA can produce a strong cyclonic anomaly over the WNP. The anomalous northeasterlies related to the anomalous cyclone prevailed over southwest China and brought in the anomalously dry air that caused the large autumn rainfall deficits. Both the observational data and model simulations show that the scale of the rainfall deficits over southwest China could be attributed mainly to the combined effects of the location and strength of SSTA over the central equatorial Pacific during the WP El Niño autumn. When the location of the warming SSTA is located farther westward, and its intensity increases, a stronger cyclone anomaly develops over the WNP. This stronger cyclone anomaly causes a more severe autumn drought over southwest China. Among the nine WP El Niño autumns, the 2009 event exhibited distinctive characteristics with regards to the location and intensity of the warming SST anomaly: the location was farther to the west, and the intensity was the highest of the nine events. Therefore, this WP El Niño is possibly a key factor for the severe autumn drought over southwest China. Additionally, the observed and simulated analyses exhibit a high sensitivity of responses of the WNP atmosphere and autumn rainfall over southwest China to the central Pacific SST warming. It increases the difficulty in seasonal prediction because it is a big challenge for current models to capture both position and intensity of maximum SST anomaly. A great effort needs be made to improve the model performance in accurately simulating regional SST anomalies and the associated atmospheric responses in the future.
Accompanying the El Niño events, a weak warming is always observed in the Indian Ocean, possibly associated with the ENSO-induced heat flux anomaly (Klein et al. 1999). A similar warming is also found during the WP El Niño, as mentioned previously (e.g., Ashok et al. 2007; Kug et al. 2009). There remains one question about whether the Indian Ocean warming had an effect on the autumn drought over southwest China in 2009. As a significant phenomenon in the Indian Ocean, the Indian Ocean dipole (IOD) may influence the extratropical climate through a Rossby wave response (Cai et al. 2011). As shown in Fig. 11a, positive IOD event would lead to more autumn rainfall over southwest China.
However, in our case during autumn 2009, a basinwide warming, rather than an IOD-like pattern, happened in the Indian Ocean (Fig. 11b). Many studies have turned their attention to the impact of the basinwide Indian Ocean warming on the Asian climate (e.g., Watanabe and Jin 2002; Li et al. 2005; Yang et al. 2007; Xie et al. 2009; Wu et al. 2010). They demonstrated that such basinwide warming can excite an anomalous anticyclone over the WNP as a Kelvin wave response, which tends to bring more moist air to East Asia. It is also supported by modeling experiments by adding only the observed SST anomaly in the Indian Ocean (40°–100°E, 20°S–20°N) during autumn 2009 to the climatological SST. The results show that an anomalous anticyclone covers the WNP (not shown), similar to the previous studies (e.g., Watanabe and Jin 2002; Yang et al. 2007). During autumn 2009, there is still a weak anticyclone anomaly occurring over the Indian Ocean (Fig. 3b), which is possibly associated with the Indian Ocean warming. It seems that the anticyclone anomaly is displaced westward because of the inhibitory effect of the strong WNP cyclone anomaly. Therefore, the Indian Ocean warming is not the main cause for the strong drought.
In this study, we have mainly discussed the large-scale atmospheric circulation anomalies associated with the WP El Niño and its impacts on the autumn drought over southwest China. Other processes, such as local feedback, may also have played a role in the drought.
We thank Dr. Yuqing Wang for useful discussion during the course of this work. This work is supported by the National Nature Science Foundation of China (41005049), the National Basic Research Program “973” (2012CB417403), National Science Foundation Grant ATM 1034798, NOAA Grant NA10OAR4310200, DOE Grant DESC0005110, and the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).