Abstract

The findings of the study reported in this paper show that, during ENSO decaying summers, rainfall and circulation anomalies exhibit clear subseasonal variation. Corresponding to a positive (negative) December–February (DJF) Niño-3.4 index, a positive (negative) subtropical rainfall anomaly, with a southwest–northeast tilt, appears in South China and the western North Pacific (WNP) in the subsequent early summer (from June to middle July) but advances northward into the Huai River Basin in China as well as Korea and central Japan in late summer (from late July to August). Concurrently, a lower-tropospheric anticyclonic anomaly over the WNP extends northward from early to late summer. The seasonal change in the basic flows, characterized by the northward shift of the upper-tropospheric westerly jet and the WNP subtropical high, is suggested to be responsible for the differences in the above rainfall and circulation anomalies between early and late summer by inducing distinct extratropical responses even under the almost identical tropical forcing of a precipitation anomaly in the Philippine Sea.

A particular focus of the study is to investigate, using station rainfall data, the subseasonal variations in ENSO-related rainfall anomalies in eastern China since the 1950s, to attempt to examine their role in weakening the relationship between the ENSO and summer mean rainfall in eastern China since the late 1970s. It is found that the ENSO-related rainfall anomalies tend to be similar between early and late summer before the late 1970s, that is, the period characterized by a stronger ENSO–summer mean rainfall relationship. After the late 1970s, however, the anomalous rainfall pattern in eastern China is almost reversed between early and late summer, resulting accordingly in a weakened relationship between the ENSO and total summer rainfall in eastern China.

1. Introduction

Interannual variability of summer climate in the western North Pacific (WNP) and East Asia tends to be related to the phases of the El Niño–Southern Oscillation (ENSO) (e.g., Huang and Wu 1989; Wang et al. 2000; Chou et al. 2003; Xue and Liu 2008; Huang and Huang 2009). Over the subtropical WNP, a lower-tropospheric cyclonic (anticyclonic) anomaly occurs during El Niño (La Niña) developing summers, and, in contrast, a lower-tropospheric anticyclonic (cyclonic) anomaly appears during El Niño (La Niña) decaying summers (Wang et al. 2000; Lim et al. 2002; Chou et al. 2003). This lower-tropospheric cyclonic (anticyclonic) anomaly results from rainfall anomalies over the Philippine Sea (Lu 2001b; Lu and Dong 2001) and is crucial for the well-known seesaw pattern of summer rainfall anomalies over the WNP and East Asia.

In ENSO decaying years in particular, precursory SST anomalies can be used to forecast East Asian rainfall in summer, which is a rainy season in the WNP and East Asia and a crucial season for agriculture in East Asia. Subsequently, there has tended to be an emphasis on studying the ENSO’s impacts on summer climate anomalies in the WNP and East Asia during ENSO decaying years, in comparison with those during ENSO developing years. The effects of ENSO on climate in the WNP and East Asia persist from winter to the following summer, through a positive feedback role of atmosphere–ocean interaction in the WNP (Wang et al. 2000) or through an Indian Ocean capacitor effect (Terao and Kubota 2005; Yang et al. 2007; Li et al. 2008; Xie et al. 2009). Summer rainfall anomalies in eastern China are closely related to the variations in East Asian summer monsoon and thus are also affected by the ENSO. In El Niño decaying summers, positive rainfall anomalies tend to appear over South China, while negative rainfall anomalies locate between the Yangtze River and the Huai River (Huang and Wu 1989).

The relationship between East Asian summer monsoon and the ENSO, however, has been found to be unstable on the interdecadal time scale (Wang 2002; Feng and Hu 2004; Kwon et al. 2005; Yim et al. 2008). One of such changes in the monsoon–ENSO relationship can be seen to have happened around the late 1970s. Wu and Wang (2002) investigated the change in the monsoon–ENSO relationship between the two periods before and after the 1970s. They found that in El Niño decaying summers, in the earlier period (1962–77), rainfall was enhanced in North China and suppressed in central Japan, but it tended to be normal in these two regions during the later period (1978–93). Furthermore, Gao et al. (2006) indicated that the relationship between summer mean rainfall in China and preceding winter SST anomalies in the equatorial central and eastern Pacific has weakened remarkably since the late 1970s. They showed that by using the precursory ENSO signals, summer rainfall anomalies can be predicted reasonably well over 43 out of 160 stations across China for the period 1951–74, but that the number of predicted stations decreases to only 15 for the period 1980–2003. This weakened relationship has increased the difficulty in forecasting summer rainfall anomalies in China.

Interestingly, while the ENSO–eastern China summer mean rainfall relationship is weaker after the late 1970s, the relationship between the circulation in summer over the WNP and ENSO becomes stronger (Tanaka 1997; Wang et al. 2008). This is against expectation since, as mentioned above, the East Asian summer monsoon is related closely to the rainfall and circulations in the WNP. In particular, it is widely believed that the preceding winter’s ENSO events affect East Asian summer monsoon through modulation of the WNP climate. How, then, does one explain these contrasting changes that occurred around the late 1970s?

Climate in the WNP and East Asia during summer exhibits a clear seasonal variation, characterized by the onset and retreat of rainy seasons in different regions (e.g., Tao and Chen 1987; Ding 1992; Wu and Wang 2001; Wang and LinHo 2002). In the tropical WNP, monsoon rain penetrates into the northern part of the Philippine Sea around mid-July, following its appearance over the South China Sea in mid-May and over the southwestern Philippine Sea in early to mid-June (Ueda et al. 1995; Wu and Wang 2001). In East Asia and the adjacent ocean, a rainband marches northward from May to August. Specifically, in eastern China, the rainy season starts in May in South China and is followed by the Chinese mei-yu, Korean changma, and Japanese baiu that persist from mid-June to mid-July. Afterward, the rainband shifts farther northward to North China and becomes less obvious. Coinciding with this northward shift of the rainband, large-scale atmospheric circulation experiences a considerable seasonal change over East Asia and the WNP. Specifically, in mid-to-late July, the East Asian upper-tropospheric westerly jet jumps northward (Lin and Lu 2008), and the core of this jet shifts abruptly westward into the Eurasian continent (Zhang et al. 2006). At the same time, in the lower troposphere the WNP subtropical high retreats eastward and advances northward (e.g., Lu 2001b).

Most previous studies on the East Asian rainfall–ENSO relationship, however, have treated rainfall anomalies as seasonal averages and thus ignored the possible role of these seasonal variations of basic flows in modulating ENSO-related rainfall anomalies. Although Lim and Kim (2007) and Wang et al. (2009) examined subseasonal ENSO-related rainfall anomalies since 1979, they focused on the large-scale principal modes in Asia and the WNP and did not attempt to consider the role of seasonal changes in basic flows. On the other hand, Lin and Lu (2009) investigated the rainfall anomalies over eastern China in the early summers of ENSO decaying years, but they did not discuss the change in the ENSO–eastern China rainfall relationship around the late 1970s.

The motivation for launching this study was to examine subseasonal rainfall and circulation anomalies in East Asia and the WNP associated with the preceding winter’s ENSO signals and to investigate the role of the seasonal variation in basic flow on the change in the relationship between ENSO and East Asian summer monsoon around the late 1970s. To achieve this, we began by analyzing the relationship between the ENSO and eastern China rainfall anomalies, before examining the ENSO-related rainfall and circulation anomalies in East Asia and the WNP. During summer, rainfall in eastern China exhibits a clear stepwise northward advancement, in good agreement with the meridional shift of the East Asian rainband. Furthermore, there have been over five decades of station rainfall data recorded in eastern China, and these data are adequate for discussion on the interdecadal change, which happened around the 1970s.

The arrangement of the paper is as follows. Section 2 describes the methodology and datasets used in the study. In section 3, we show subseasonal rainfall anomalies in eastern China, associated with preceding ENSO signals, indicating that the subseasonal variation of ENSO-related rainfall anomalies is responsible for the weakening of the ENSO–summer mean rainfall relationship after the late 1970s. We examine the ENSO-related rainfall and circulation anomalies in East Asia and the WNP and illustrate possible mechanisms for the interdecadal change and the subseasonal variation of the ENSO–summer rainfall relationship in section 4. Finally, a summary is provided in section 5.

2. Data and methods

Station rainfall data used in this study include the following: 1) monthly rainfall at 160 stations in China from 1951 to 2006 and 2) daily rainfall at 730 stations from 1951 to 2005. Both datasets were provided by the Chinese Meteorological Data Center. The monthly data were used directly, without any further treatment. The daily data were averaged into 10-day means to remove synoptic disturbances, and a three-point smoothing was then performed to the 10-day means of rainfall before calculating the correlations between ENSO and rainfall. In this way, subseasonal variability in the ENSO–rainfall relationship could be retained and the impacts of SST anomalies on rainfall highlighted. In this study, eastern China is specified by the region east of 105°E in mainland China between the latitudes 20° and 42°N but excluding northeast China. Thus, it includes North China, the Huai River Basin, and South China. In this region, there are 94 stations for the 160-station dataset during 1951–2006 and 177 stations for the 730-station dataset during 1954–2005. Also utilized were pentad rainfall data from the Climate Prediction Center (CPC) Merged Analysis Product (CMAP) for the period 1979–2006 (Xie and Arkin 1997), to investigate the relationship between the ENSO and rainfall in East Asia and the WNP. It is well known that the rainfall anomaly in the subtropical WNP is closely related to the rainfall anomaly in East Asian countries, both being referred to as East Asian summer monsoon rainfall anomalies.

Monthly and daily data of horizontal winds at 850 and 200 hPa were used, derived from the European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) from 1958 to 2002 (Uppala et al. 2005). Also used were monthly SST data from 1951 to 2006 (Smith and Reynolds 2003). Correlations and regressions were performed according to the Niño-3.4 index of the previous winter [December–February (DJF)], with SST anomalies averaged over the region (5°S–5°N, 120°–170°W).

3. Changes in the ENSO–rainfall relationship around the late 1970s

a. Summer mean results

Figure 1 shows the evolution of correlation between eastern China summer mean rainfall and preceding winter Niño-3.4 SST with a 15-yr sliding window. Around the late 1970s, a noticeable change occurs in the ENSO–summer mean rainfall relationship. Before the late 1970s the correlation is steady and statistically significant, albeit marginal, while after the late 1970s there is almost no significant correlation. Moreover, prior to the late 1970s the distribution of correlation coefficients appears to resemble a tripole pattern in the meridional direction, with positive correlations over 36°–42°N (North China) and 27°–30°N and the negative correlations over 31°–35°N (the Huai River region). However, after the late 1970s this tripole pattern changes into a dipole pattern: the correlation coefficients tend to be reverse south and north of the Huai River, and this dipole-like pattern seems to be out of phase before and after the early 1990s. In addition, around the late 1970s, North China experiences a change from positive to negative correlation, which is consistent with previous results (Wu and Wang 2002). Another considerable change—from negative to positive correlation—can also be seen in the Huai River region. Therefore, the ENSO–rainfall relationship changes evidently around the late 1970s, in both statistical significance and distribution pattern of correlation.

Based on the results shown in Fig. 1, the period of analysis was divided into the following two episodes: 1951–77, characterized by a stronger ENSO–rainfall relationship, and 1978–2006, characterized by a weaker relationship. The lengths of the two periods are almost the same (the former period being 27 yr and the latter 29 yr). Since the intensity of the ENSO–rainfall relationship is the emphasis in this study, the latter 29 yr are considered as the period of a weakened relationship, despite the evidence of changes in correlation around the early 1990s. Correlation coefficients are basically not significant during the whole later period. Moreover, an analysis over a longer period of data can lead to more statistically reliable results. The point of the division (between 1977 and 1978) is identical to that used in previous studies (Chang et al. 2000; Wu and Wang 2002; Chen et al. 2005). In addition, Gao et al. (2006) compared the results between the periods 1951–74 and 1980–2003, again similar to those being covered in the present study.

Figure 2 presents the correlation between summer mean rainfall at each station and the previous winter Niño-3.4 SST for the two periods representing a stronger and weaker ENSO–rainfall relationship. In the stronger period, corresponding to above-normal Niño-3.4 SST, there is significantly heavier rainfall in North and South China and lighter rainfall between the Yellow and Yangtze Rivers (Fig. 2a). The distribution of the correlation displays a tripole pattern, which is in agreement with the results shown in Fig. 1. In the later period, however, the correlation distribution is not well organized (Fig. 2b). In addition, the correlation coefficients decline greatly and there are almost no areas of statistical significance.

The number of stations at which the correlation coefficients are statistically significant at the 90% confidence level decreases remarkably in the second period, relative to the first. In the first period, rainfall at 16 stations is well correlated with the ENSO (Fig. 2a), but there are only six stations in the second period (Fig. 2b). These numbers confirm that the relationship between eastern China summer mean rainfall and the preceding winter ENSO weakens noticeably around the late 1970s, consistent with the findings of Gao et al. (2006). Furthermore, a Monte Carlo test (Livezey and Chen 1983) confirms that the relationship between ENSO and eastern China summer mean rainfall is significant at the 90% level in the first period but not in the second period. We have also analyzed the ENSO–rainfall relationship by using a 730-station dataset and obtained results similar to those shown in Figs. 1 and 2 (not shown). In particular, when the 730-station dataset is used, 25 stations at which summer mean rainfall is significantly correlated to the preceding winter Niño-3.4 SST are found for the first period while there are only 14 for the second period.

b. Subseasonal results

Figure 3 displays monthly rainfall regressed onto the previous winter Niño-3.4 SST for the two periods. In June, more rainfall occurs south to the Yangtze River and less rainfall locates between the Yangtze and Yellow rivers in the first period (Fig. 3a). This pattern of rainfall anomalies is roughly similar to that for the later period (Fig. 3b). In July, positive rainfall anomalies occupy a large portion of eastern China in the first period (Fig. 3c), while they shrink into much smaller areas in the second period (Fig. 3d).

In August, before the late 1970s, positive rainfall anomalies appear in North China and negative anomalies appear between the Yellow and Yangtze rivers (Fig. 3e). After the late 1970s, however, ENSO-related rainfall anomalies are positive between the Yellow and Yangtze Rivers, opposite to those in the first period. Moreover, positive rainfall anomalies are not evident in North China, while significant negative rainfall anomalies appear in South China (Fig. 3f). Thus, the relationship between the ENSO and late summer rainfall displays great differences between the two periods.

In addition, after the late 1970s, rainfall anomalies in August show a totally different representation in comparison with those in June, that is, more (less) rainfall in South China and less (more) rainfall in between the Yellow and Yangtze Rivers in June (August) (Figs. 3b and 3f). This out-of-phase distribution between June and August suggests a clear subseasonal evolution of the ENSO–rainfall relationship in the second period. In contrast, before the late 1970s, rainfall anomalies tend to be consistent between June and August, with decreased rainfall between the Yellow and Yangtze Rivers in both months (Figs. 3a and 3e). Thus, it can be implied that the opposite rainfall anomalies in June and August result in a weakened ENSO–summer mean rainfall relationship after the late 1970s.

It can also be seen from Fig. 3 that the association between monthly rainfall and the ENSO appears not to be weak, although the seasonal-average results indicate a significantly weakened relationship in the second period. In each month of the summer season, the areas of significant correlation do not shrink in the later period relative to the earlier period. Moreover, the numbers of significantly correlated stations for each month can also illustrate that the ENSO–rainfall relationship has not weakened significantly. Before the late 1970s, there are 15, 14, and 12 significantly correlated stations in June, July, and August, respectively. These numbers change to 15, 7, and 14, respectively, after the late 1970s. Compared with the sharp decrease in significantly correlated stations from 16 to 6 shown in the seasonal-average results (Fig. 2), the decrease in the “well correlated” station numbers is much weaker for each month.

Since rainfall is noticeably dependent on local characteristics such as topography, it is necessary to reexamine the ENSO–rainfall relationship by using another dataset of rainfall. Thus, we used the 730-station data to repeat the analysis and found that the results (not shown) resemble well those obtained by using the 160-station dataset (Fig. 3). The results confirm the conclusion that the monthly rainfall–ENSO relationship has not weakened since the late 1970s, but the distribution of monthly rainfall anomalies are different between the two periods, especially in August. Among the 177 stations in eastern China, the numbers of significantly correlated stations at the 90% confidence level are 17, 22, and 22 for the first period during June, July, and August, respectively, and 17, 15, and 38 for the second period.

The subseasonal evolution of the ENSO–rainfall relationship can be more clearly illustrated by zonal mean rainfall regressed onto the previous winter Niño-3.4 SST (Fig. 4). During the first period, positive anomalies lie over South China (22°–30°N) in roughly all months of the summer and negative anomalies appear between the Yellow and Yangtze Rivers (30°–37°N) except for a small patch of weak positive rainfall anomalies in July (Fig. 4a). Meanwhile, North China is occupied by significantly heavier rainfall in July and August. During the second period, however, in the lower latitudes more rainfall is discernable in June while lighter rainfall prevails in July and August (Fig. 4b). In addition, negative rainfall anomalies in early summer are followed by positive ones in late summer between the Yellow and Yangtze Rivers. Furthermore, there are almost no significant anomalies over North China.

Figure 5 displays the subseasonal evolution of zonal mean rainfall regressed onto the preceding winter Niño-3.4 SST, using the 730-station rainfall data. Similar to the results shown in Fig. 4, ENSO-related rainfall anomalies tend to be somewhat consistent during summer before the late 1970s (Fig. 5a) but exhibit a sharp difference between early and late summer after the late 1970s (Fig. 5b). During the second period, in South China there is a positive rainfall anomaly in early summer but a negative anomaly in late summer. Moreover, between the Yellow and Yangtze Rivers, a negative anomaly in early summer changes to a positive anomaly in late summer. These consistent or reversed patterns of rainfall anomalies between early and late summer confirm the result obtained using the 160-station dataset (Fig. 4).

Therefore, we can conclude that a consistent or inconsistent ENSO–subseasonal rainfall relationship during summer is responsible for a strong or weak ENSO–summer mean rainfall relationship. Before the late 1970s, the subseasonal rainfall anomalies related to ENSO are roughly consistent during the whole summer, and thus the ENSO–summer mean rainfall relationship is stronger. In contrast, after the late 1970s, opposite rainfall anomalies between early and late summer result in a weakened ENSO–summer mean rainfall relationship.

4. Mechanisms for the changes in the ENSO–rainfall relationship between early and late summer

The results shown in the preceding section indicate that the inconsistent ENSO–rainfall relationship between early and late summer is responsible for the weakening of the relationship between the ENSO and summer mean rainfall in eastern China after the late 1970s. However, the mechanisms responsible for the change in the ENSO–rainfall relationship between early and late summer remain unknown. In this section, the circulation and SST anomalies associated with the previous winter Niño-3.4 SST are examined, to explore possible mechanisms for the subseasonal change in the ENSO–rainfall relationship.

It has been found from Fig. 5, as well as from Figs. 3 and 4, that the ENSO–rainfall relationship experiences a transition in July. On the other hand, previous studies have indicated that in middle and late July the East Asian upper-tropospheric westerly jet jumps northward (Lin and Lu 2008) and the WNP subtropical high retreats eastward and advances northward (e.g., Lu 2001b). Alongside these circulation changes, the East Asian summer monsoon withdraws and the rainband shifts northward into North China and becomes less obvious. These considerable changes in mean state may have an effect on the EN SO–subseasonal rainfall relationship. Therefore, following these examples from previous work, we divide the summer season into two parts: early summer (1 June–20 July) and late summer (21 July–31 August), thus enabling an investigation of the ENSO–rainfall relationship in early and late summer, respectively.

Figure 6 shows early and late summer rainfall anomalies regressed onto the preceding winter Niño-3.4 SST, for the first and second periods. In the first period, the spatial pattern of rainfall anomalies is roughly similar between early and late summer, that is, positive anomalies in North and South China and negative anomalies between the Yellow and Yangtze Rivers (Figs. 6a and 6c). In the second period, the spatial pattern of early summer rainfall anomalies (Fig. 6b) resembles that of the first period (Fig. 6a); however, late summer rainfall anomalies (Fig. 6d) are completely different with both early summer anomalies in the second period (Fig. 6b) and late summer anomalies in the first period (Fig. 6c). The late summer rainfall anomalies in the second period are characterized by lighter rainfall in South China and heavier rainfall between the Yellow and Yangtze Rivers (Fig. 6d). In addition, the ENSO–subseasonal rainfall relationship does not weaken in the second period, confirming the results presented in the previous section and suggesting that the weakened ENSO–summer mean rainfall relationship in this period is due to the contrast of rainfall anomalies between early and late summer. In the first period, there are 10 and 18 significantly correlated stations in early and late summer, respectively (Figs. 6a and 6c), while in the second period there are 7 and 31 stations in early and late summer, respectively (Figs. 6b and 6d). The number of significantly correlated stations in late summer is much larger in the second period than the first.

Figure 7 shows early and late summer 850-hPa winds regressed onto previous winter Niño-3.4 SST in the first and second periods. There is an anomalous anticyclone over the WNP in early and late summer in both the periods. However, this anticyclonic anomaly, for both early and late summer, is much weaker in the first period and becomes stronger and more significant in the second period. Furthermore, in the second period, the anticyclonic anomaly is stronger and stretched more northward in late summer than in early summer (Figs. 7b and 7d).

During the second period, the lower-tropospheric circulation anomalies are dynamically consistent with the rainfall anomalies shown in Fig. 6. The anticyclonic anomaly suggests a westward extension of the WNP subtropical high (Lu 2001b) and more moisture flux along the northwest flank of the high. Therefore, more rainfall tends to occur in South China in early summer (Fig. 6b). For late summer, however, because of the extraordinary northward and westward extension of the WNP subtropical high, more rainfall occurs between the Yellow and Yangtze Rivers and less rainfall is found in South China (Fig. 6d), where it is under the control of the anticyclone. Such a dynamical consistency between rainfall and circulation anomalies does not exist in the first period, likely because of the weaker influence of ENSO on these anomalies over the WNP and East Asia in this period (e.g., Wang et al. 2008; Xie et al. 2010).

Figure 8 shows the regression of monthly mean SSTs during summer onto the previous winter Niño-3.4 index for the two periods. Here, monthly mean results are shown, rather than early and late summer results, owing to the shortage of SST data with a higher temporal resolution in both periods. However, as SSTs change slowly, results on SST anomalies for early and late summer can be well inferred by monthly mean results. In the first period, positive SST anomalies appear in the Indian Ocean and South China Sea in June and the positive anomalies shrink to the Southeast Indian Ocean and South China Sea in July and August (Figs. 8a, 8c, and 8e). In addition, the equatorial eastern Pacific is dominated by negative SST anomalies from June to August.

On the contrary, during the second period, positive SST anomalies in the Indian Ocean are stronger relatively and persist throughout the whole summer, although they become weaker in late summer (Figs. 8b, 8d, and 8f). This difference in intensity of the Indian Ocean SST anomalies between the two periods is in agreement with Xie et al. (2010), who explained this difference by the strengthening and slow decaying of the ENSO in the second period. In the WNP, negative SST anomalies appear in early summer but become less obvious in late summer. In addition, there are warm SSTs in the equatorial eastern Pacific.

These ENSO-related SST anomalies can explain the difference in the anticyclonic anomaly over the WNP between the two periods, that is, a weaker anomaly in the first period and stronger anomaly in the second period (Fig. 7). On the one hand, many studies have indicated that positive SST anomalies in the Indian Ocean, following wintertime El Niño events in the equatorial eastern Pacific, induce an anticyclonic anomaly over the WNP in the El Niño decaying summer through a Kelvin wave (Yang et al. 2007; Li et al. 2008; Ding et al. 2009; Xie et al. 2009; Wu et al. 2010). On the other hand, negative SST anomalies in the WNP also favor the formation and maintenance of the WNP anticyclonic anomaly through a Rossby wave, particularly in early summer (Wu et al. 2010). Therefore, based on these previous results, it can be concluded that, in the second period, a combination of the stronger positive SST anomalies in the Indian Ocean and negative SST anomalies in the WNP induce a stronger anticyclonic anomaly over the WNP.

The ENSO-related SST anomalies cannot explain the difference in circulation and rainfall between early and late summer, although they can explain the difference between the two periods. Both the positive SST anomalies in the Indian Ocean and negative anomalies in the WNP are stronger in early summer and dissipate significantly in late summer for both the periods. This weakening of the ENSO-related SST anomalies from early to late summer is apparently contradictory to the simultaneous intensification of the WNP anticyclonic anomaly in the second period. Therefore, in the following analysis, we focus on the difference between early and late summer in the second period, in an attempt to explain the physical reason for this apparent contradiction.

Figure 9 shows the ENSO-related rainfall anomalies in East Asia and the western Pacific in the second period, for early and late summer. In both early and late summer, the rainfall anomalies display a well-known seesaw pattern in the meridional direction in the WNP and East Asia, that is, a negative anomaly in the tropical WNP (10°–20°N, 110°–160°E), and positive anomaly in the subtropical region of East Asia and the WNP. The negative anomaly in the tropical WNP is of roughly the same location (the Philippine Sea) and amplitude in early and late summer. The subtropical positive anomaly, however, exhibits a remarkable change in location from early to late summer. This subtropical anomaly is located in South China and the WNP in early summer (Fig. 9a) but advances northward into the Huai River Basin in China, Korea, and central Japan in late summer (Fig. 9b). This subtropical anomaly with a southwest–northeast tilt, for both early and late summer, is located just south of the axis of the upper-tropospheric westerly jet, which experiences an abrupt northward jump in mid summer (Lin and Lu 2008). This association between the rainfall anomalies and the climatological westerly jet implies a possible role of the seasonal change in basic flows in affecting the features of subseasonal anomalies, which will be investigated further in the following analysis. It should be noted that the seasonal march of the jet for the first period (not shown) is almost identical to that for the second period.

The eastern China rainfall anomalies in Fig. 9 are consistent with those obtained using station rainfall data (Figs. 6b and 6d). In early summer, a weak positive anomaly appears in South China (Figs. 9a and 6b), and in late summer a negative (positive) anomaly locates in South China (between the Yellow River Yangtze rivers) (Figs. 9b and 6d). The results from both the station data and CMAP data indicate that, in the second period, late summer rainfall anomalies are more closely related to the previous winter Niño-3.4 SST than early summer rainfall anomalies in eastern China.

To illustrate the possible role of seasonal variation in basic flows in affecting ENSO-related rainfall anomalies, 850-hPa winds regressed onto the rainfall anomalies averaged over the Philippine Sea are presented in Fig. 10. There is a negative rainfall anomaly for both early and late summer in the Philippine Sea, corresponding to the previous winter Niño-3.4 SST (Fig. 9). Since the location of this negative rainfall anomaly is roughly the same in early and late summer, we define a Philippine Sea rainfall anomaly index (PSRI) by simply averaging rainfall anomalies over the region (10°–20°N, 110°–160°E) and then multiplying the average by minus one. This region is determined by the results shown in Fig. 9 and is identical to that used to represent the tropical WNP in previous studies (Lu 2001a; Wu and Wang 2001; Kobayashi et al. 2005). The purpose of the sign change in the index is to facilitate comparison with the Niño-3.4 SST-related anomalies. A positive PSRI indicates a negative precipitation anomaly in the tropical WNP and thus corresponds to a positive DJF Niño-3.4 index.

Corresponding to a negative rainfall anomaly in the tropical WNP (or a positive PSRI), there is an anticyclonic anomaly northwest to the region of the rainfall anomaly in both early and late summer (Figs. 10a and 10b). This is consistent with the findings of Lu (2001b), who suggested that an anomalous anticyclone can be induced over the WNP by a negative precipitation anomaly in the Philippine Sea. However, there is a distinct difference in the scope of the anticyclonic anomaly between early and late summer; this anticyclonic anomaly extends significantly farther northward in late summer than in early summer. In addition, this change between early and late summer is in agreement with the change in the ENSO-related WNP anticyclonic anomaly (Figs. 7b and 7d). Furthermore, the PSRI-related rainfall anomalies in the Philippine Sea are of roughly the same scope and amplitude between early and late summer (not shown), consistent with the ENSO-related rainfall anomalies shown in Fig. 9.

The distinct difference in the WNP anticyclonic anomaly between early and late summer under similar tropical heat forcing suggests that the basic flow, which exhibits a clear change from early to late summer, may play a role. Recently, using a linear quasigeostrophic two-layer model, Kosaka and Nakamura (2010) performed two numerical experiments: tropical forcing remaining identical but the basic-state subtropical jet axis being located at 35°N in one experiment and at 50°N in the other. They found that the circulation responses to the tropical forcing are sensitive to the upper-tropospheric westerly jet position and the meridional scale of the WNP subtropical high at the lower troposphere. Despite the unchanged tropical forcing between the two experiments, the lower-tropospheric anticyclonic response extends poleward significantly and is stronger when the jet axis is located poleward (at 50°N). This experimental result from Kosaka and Nakamura (2010) confirms that the seasonal change in the basic flow plays a crucial role in inducing the difference of the ENSO-related WNP anticyclonic anomaly between early and late summer.

5. Summary

The relationship between summer mean rainfall in eastern China and the preceding winter’s ENSO, which can be used for seasonal forecasting of potential floods and droughts, has been significantly weakened since the late 1970s, as indicated by previous studies. In the present study, the reason for this weakening has been explored, uncovering the finding that subseasonal variation in ENSO-related rainfall anomalies plays a crucial role. Thus, we investigated specifically the subseasonal variation of ENSO-related rainfall and circulation anomalies during summer in East Asia and the WNP.

Results show that the subseasonal variation in ENSO-related rainfall causes the weakening of the ENSO–eastern China summer mean rainfall relationship after the late 1970s. The ENSO-related rainfall anomalies tend to be similar between early and late summer before the late 1970s, that is, the period characterized by a stronger ENSO–summer mean rainfall relationship. After the late 1970s, however, corresponding to a positive DJF Niño-3.4 index, rainfall is enhanced in South China and suppressed between the Yellow River and Yangtze Rivers in early summer, but the pattern of anomalous rainfall is almost reversed in late summer. Therefore, this subseasonal variation in ENSO-related rainfall anomalies results in a weakened relationship between the ENSO and total summer rainfall in eastern China since the late 1970s, although the relationships between the ENSO and both early and late summer rainfall anomalies do not weaken.

Furthermore, subseasonal variations in ENSO-related rainfall anomalies in East Asia in the second period (i.e., after the late 1970s) were investigated, a period for which satellite-observed rainfall data over the WNP are available. Corresponding to a positive DJF Niño-3.4 index, a positive subtropical rainfall anomaly, with a southwest–northeast tilt, was found to locate just south of the axis of the upper-tropospheric westerly jet for both early and late summer. This subtropical anomaly appears in South China and the WNP in early summer but advances northward into the Huai River Basin in China, Korea, and central Japan in late summer. This northward shift of the positive subtropical rainfall anomaly from early to late summer can be explained by the simultaneous northward extension of the ENSO-related WNP anticyclonic anomaly. It is suggested that seasonal variation in the basic flows plays a crucial role in affecting the features of ENSO-related rainfall and circulation anomalies in East Asia and the WNP. The northward shift of the upper-tropospheric westerly jet and concurrent change in the WNP subtropical high is responsible for both the northward shift of the ENSO-related positive subtropical rainfall anomaly and the northward extension of the ENSO-related WNP anticyclonic anomaly from early to late summer, even under an almost identical tropical forcing (a negative precipitation anomaly in the Philippine Sea). This observational result is supported by results from a simple model (Kosaka and Nakamura 2010).

In the first period (i.e., before the late 1970s), in contrast, the ENSO-related WNP anticyclonic anomaly was found to be much weaker than in the second period. This is consistent with previous studies (Tanaka 1997; Wang et al. 2008), which show that there is not a clear relationship between the preceding winter ENSO and summer circulation anomaly over the WNP. On the other hand, the seasonal march of the basic flow is almost identical between the first and second periods. Therefore, the change in intensity of the ENSO-related WNP anticyclonic anomaly is the essential factor to produce the different patterns of rainfall anomalies between the two periods. Furthermore, in the first period, because of the weakness of tropical forcing over the WNP, this tropical forcing may not induce a clear subseasonal difference in rainfall anomalies over eastern China, even under the circumstance of the seasonal change in basic flow.

Based on the present results, as well as those from previous results (Yang et al. 2007; Li et al. 2008; Ding et al. 2009; Xie et al. 2009; Wu et al. 2010), it is suggested that in the second period, in summer following a positive DJF Niño-3.4 index, a combination of the stronger positive SST anomaly in the Indian Ocean and negative SST anomaly in the WNP induces a stronger anticyclonic anomaly over the WNP, through a Kelvin wave and Rossby wave, respectively. These SST anomalies are much weaker or invisible in the first period. Therefore, this result on the SST anomalies is consistent with the differences in the ENSO-related WNP lower-tropospheric circulation and SST anomalies between the first and second periods.

The present results suggest that ENSO events can serve as a potential predictor for subseasonal rainfall anomalies in East Asia, as well as seasonal mean anomalies. This argument is in agreement with Kim et al. (2008), who argued that subseasonal variability, relative to seasonal mean variability, has comparable or even more SST-forced variance.

Acknowledgments

We thank two anonymous reviewers for their valuable comments and suggestions, which greatly help us improve the presentation of this paper. This study was supported by the Ministry of Finance of China (Grant GYHY200906017) and by National Natural Science Foundation of China (Grants 40810059005 and 40725016).

REFERENCES

REFERENCES
Chang
,
C.-P.
,
Y.
Zhang
, and
T.
Li
,
2000
:
Interannual and interdecadal variations of the East Asian summer monsoon and tropical Pacific SSTs. Part I: Roles of the subtropical ridge
.
J. Climate
,
13
,
4310
4325
.
Chen
,
J.-M.
,
F.-C.
Lu
,
S.-L.
Kuo
, and
C.-F.
Shih
,
2005
:
Summer climate variability in Taiwan and associated large-scale processes
.
J. Meteor. Soc. Japan
,
83
,
499
515
.
Chou
,
C.
,
J.-Y.
Tu
, and
J.-Y.
Yu
,
2003
:
Interannual variability of the western North Pacific summer monsoon: Differences between ENSO and non-ENSO years
.
J. Climate
,
16
,
2275
2287
.
Ding
,
R.
,
K.-J.
Ha
, and
J.
Li
,
2009
:
Interdecadal shift in the relationship between the East Asian summer monsoon and the tropical Indian Ocean
.
Climate Dyn.
,
34
,
1059
1071
.
Ding
,
Y.
,
1992
:
Summer monsoon rainfalls in China
.
J. Meteor. Soc. Japan
,
70
,
373
396
.
Feng
,
S.
, and
Q.
Hu
,
2004
:
Variations in the teleconnection of ENSO and summer rainfall in northern China: A role of the Indian summer monsoon
.
J. Climate
,
17
,
4871
4881
.
Gao
,
H.
,
Y.
Wang
, and
J.
He
,
2006
:
Weakening significance of ENSO as a predictor of summer precipitation in China
.
Geophys. Res. Lett.
,
33
,
L09807
,
doi:10.1029/2005GL025511
.
Huang
,
P.
, and
R.-H.
Huang
,
2009
:
Relationship between the modes of winter tropical Pacific SST anomalies and the intraseasonal variations of the following summer rainfall anomalies in China
.
Atmos. Ocean. Sci. Lett.
,
2
,
295
300
.
Huang
,
R. H.
, and
Y. F.
Wu
,
1989
:
The influence of ENSO on the summer climate change in China and its mechanism
.
Adv. Atmos. Sci.
,
6
,
21
32
.
Kim
,
K.-Y.
,
A.
Kitoh
, and
K.-J.
Ha
,
2008
:
The SST-forced predictability of the sub-seasonal mode over East Asia with an atmospheric general circulation model
.
Int. J. Climatol.
,
28
,
1599
1606
.
Kobayashi
,
C.
,
S.
Maeda
,
A.
Ito
,
Y.
Matsushita
, and
K.
Takano
,
2005
:
Relation between SSTs and predictability of seasonal mean precipitation over the western tropical Pacific
.
J. Meteor. Soc. Japan
,
83
,
919
929
.
Kosaka
,
Y.
, and
H.
Nakamura
,
2010
:
Mechanisms of meridional teleconnection observed between a summer monsoon system and a subtropical anticyclone. Part I: The Pacific–Japan pattern
.
J. Climate
,
23
,
5085
5108
.
Kwon
,
M. H.
,
J.-G.
Jhun
,
B.
Wang
,
S.-I.
An
, and
J.-S.
Kug
,
2005
:
Decadal change in relationship between East Asian and WNP summer monsoons
.
Geophys. Res. Lett.
,
32
,
L16709
,
doi:10.1029/2005GL023026
.
Li
,
S.
,
J.
Lu
,
G.
Huang
, and
K.
Hu
,
2008
:
Tropical Indian Ocean Basin warming and East Asian summer monsoon: A multiple AGCM study
.
J. Climate
,
21
,
6080
6088
.
Lim
,
Y. K.
, and
K. Y.
Kim
,
2007
:
ENSO impact on the space–time evolution of the regional Asian summer monsoons
.
J. Climate
,
20
,
2397
2415
.
Lim
,
Y. K.
,
K. Y.
Kim
, and
H. S.
Lee
,
2002
:
Temporal and spatial evolution of the Asian summer monsoon in the seasonal cycle of synoptic fields
.
J. Climate
,
15
,
3630
3644
.
Lin
,
Z.
, and
R.
Lu
,
2008
:
Abrupt northward jump of the East Asian upper-tropospheric jet stream in mid-summer
.
J. Meteor. Soc. Japan
,
86
,
857
866
.
Lin
,
Z.
, and
R.
Lu
,
2009
:
The ENSO’s effect on the eastern China rainfall in the following early summer
.
Adv. Atmos. Sci.
,
26
,
333
342
.
Livezey
,
R.
, and
W.
Chen
,
1983
:
Statistical field significance and its determination by Monte Carlo techniques
.
Mon. Wea. Rev.
,
111
,
46
59
.
Lu
,
R.
,
2001a
:
Atmospheric circulations and sea surface temperatures related to the convection over the western Pacific warm pool on the interannual scale
.
Adv. Atmos. Sci.
,
18
,
270
282
.
Lu
,
R.
,
2001b
:
Interannual variability of the summertime North Pacific subtropical high and its relation to atmospheric convection over the warm pool
.
J. Meteor. Soc. Japan
,
79
,
771
783
.
Lu
,
R.
, and
B.
Dong
,
2001
:
Westward extension of North Pacific subtropical high in summer
.
J. Meteor. Soc. Japan
,
79
,
1229
1241
.
Smith
,
T. M.
, and
R. W.
Reynolds
,
2003
:
Extended reconstruction of global sea surface temperatures based on COADS data (1854–1997)
.
J. Climate
,
16
,
1495
1510
.
Tanaka
,
M.
,
1997
:
Interannual and interdecadal variations of the western North Pacific monsoon and baiu rainfall and their relationship to the ENSO cycles
.
J. Meteor. Soc. Japan
,
75
,
1109
1123
.
Tao
,
S.-Y.
, and
L.-X.
Chen
,
1987
:
A review of recent research on the East Asian summer monsoon in China
.
Monsoon Meteorology, C.-P. Chang and T. N. Krishnamurti, Eds., Oxford University Press, 60–92
.
Terao
,
T.
, and
T.
Kubota
,
2005
:
East-west SST contrast over the tropical oceans and post El Niño western North Pacific summer monsoon
.
Geophys. Res. Lett.
,
32
,
L15706
,
doi:10.1029/2005GL023010
.
Ueda
,
H.
,
T.
Yasunari
, and
R.
Kawamura
,
1995
:
Abrupt seasonal change of large-scale convective activity over the western Pacific in the northern summer
.
J. Meteor. Soc. Japan
,
73
,
795
809
.
Uppala
,
S. M.
, and
Coauthors
,
2005
:
The ERA-40 Re-Analysis
.
Quart. J. Roy. Meteor. Soc.
,
131
,
2961
3012
.
Wang
,
B.
, and
LinHo
,
2002
:
Rainy season of the Asian–Pacific summer monsoon
.
J. Climate
,
15
,
386
398
.
Wang
,
B.
,
R.
Wu
, and
X.
Fu
,
2000
:
Pacific–East Asian teleconnection: How does ENSO affect East Asian climate?
.
J. Climate
,
13
,
1517
1536
.
Wang
,
B.
,
J.
Yang
,
T.
Zhou
, and
B.
Wang
,
2008
:
Interdecadal changes in the major modes of Asian–Australian monsoon variability: Strengthening relationship with ENSO since the late 1970s
.
J. Climate
,
21
,
1771
1789
.
Wang
,
B.
,
J.
Liu
,
J.
Yang
,
T.
Zhou
, and
Z.
Wu
,
2009
:
Distinct principal modes of early and late summer rainfall anomalies in East Asia
.
J. Climate
,
22
,
3864
3875
.
Wang
,
H. J.
,
2002
:
The instability of the East Asian summer monsoon-ENSO relations
.
Adv. Atmos. Sci.
,
19
,
1
11
.
Wu
,
B.
,
T.
Li
, and
T.
Zhou
,
2010
:
Relative contributions of the Indian Ocean and local SST anomalies to the maintenance of the western North Pacific anomalous anticyclone during El Niño decaying summer
.
J. Climate
,
23
,
2974
2986
.
Wu
,
R.
, and
B.
Wang
,
2001
:
Multi-stage onset of the summer monsoon over the western North Pacific
.
Climate Dyn.
,
17
,
277
289
.
Wu
,
R.
, and
B.
Wang
,
2002
:
A contrast of the East Asian summer monsoon–ENSO relationship between 1962–77 and 1978–93
.
J. Climate
,
15
,
3266
3279
.
Xie
,
P.
, and
P.
Arkin
,
1997
:
Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs
.
Bull. Amer. Meteor. Soc.
,
78
,
2539
2558
.
Xie
,
S.-P.
,
K.
Hu
,
J.
Hafner
,
H.
Tokinaga
,
Y.
Du
,
G.
Huang
, and
T.
Sampe
,
2009
:
Indian Ocean capacitor effect on Indo–western Pacific climate during the summer following El Niño
.
J. Climate
,
22
,
730
747
.
Xie
,
S.-P.
,
Y.
Du
,
G.
Huang
,
X.-T.
Zheng
,
H.
Tokinaga
,
K.
Hu
, and
Q.
Liu
,
2010
:
Decadal shift in El Niño influences on Indo-western Pacific and East Asian climate in the 1970s
.
J. Climate
,
23
,
3352
3368
.
Xue
,
F.
, and
C.
Liu
,
2008
:
The influence of moderate ENSO on summer rainfall in eastern China and its comparison with strong ENSO
.
Chin. Sci. Bull.
,
53
,
791
800
.
Yang
,
J.
,
Q.
Liu
,
S.-P.
Xie
,
Z.
Liu
, and
L.
Wu
,
2007
:
Impact of the Indian Ocean SST basin mode on the Asian summer monsoon
.
Geophys. Res. Lett.
,
34
,
L02708
,
doi:10.1029/2006GL028571
.
Yim
,
S.-Y.
,
S.-W.
Yeh
,
R.
Wu
, and
J.-G.
Jhun
,
2008
:
The influence of ENSO on decadal variations in the relationship between the East Asian and western North Pacific summer monsoons
.
J. Climate
,
21
,
3165
3179
.
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
,
doi:10.1029/2006GL026377
.