Abstract

The interdecadal Pacific oscillation (IPO) shifted to a negative phase around the late 1990s. Its impact on the atmospheric quasi-biweekly oscillation (QBWO) intensity over the western North Pacific (WNP) during late summer was investigated. Corresponding to the phase transition of the IPO, La Niña–like SST anomalies and an enhanced Walker circulation appeared in the tropical Pacific, which led to decreased precipitation over the equatorial central and eastern Pacific. The decreased precipitation induced a Gill response with an anomalous anticyclone (cyclone) in the lower (upper) troposphere over the WNP. This resulted in anomalous background westerly vertical shear over the tropical WNP. Furthermore, the anomalous anticyclone induced anomalous horizontal divergence and descent motion in the planetary boundary layer, which led to decreased background surface moisture over the tropical WNP. These changes in background atmospheric conditions suppressed the development of QBWO perturbations over the tropical WNP. Therefore, the QBWO intensity weakened over the WNP after the late 1990s. The composite evolution of QBWO events before and after the late 1990s confirm the interdecadal change of the QBWO intensity. A simple model was employed to understand the relative role of the background moisture and vertical shear changes in modulating the QBWO activity. The result shows that the moisture change plays a more important role than the vertical shear change in weakening the QBWO intensity.

1. Introduction

The quasi-biweekly oscillation (QBWO) over the western North Pacific (WNP) during summer has a pronounced impact on climate over the East Asia–WNP region (Li and Wang 2005; Mao and Chan 2005; Zhou and Chan 2005; Yang et al. 2010; Li and Zhou 2013; Ren et al. 2013; Chen et al. 2015). Apparently, the strength of the climate impact of the QBWO is closely related to its intensity. Therefore, it is meaningful to study the variation of the QBWO intensity during summer over the WNP and its mechanisms.

The westward-propagating QBWO over the WNP during summer can be understood as n = 1 equatorial Rossby waves in the presence of convective coupling and monsoon mean flow (Chatterjee and Goswami 2004; Kikuchi and Wang 2009). They can be traced to the westward-propagating perturbations in the equatorial western Pacific (Mao and Chan 2005; Kikuchi and Wang 2009; Chen and Sui 2010). The n = 1 equatorial Rossby waves related to the QBWO can be driven unstable by convective coupling, in which the planetary boundary layer convergence plays an important role (Wang 1988; Wang and Li 1994; Chatterjee and Goswami 2004). The background surface moisture can change the vertical gradient of background moisture and influence vertical moisture transport by the low-frequency vertical motion of the QBWO, which is important to connective coupling (Chatterjee and Goswami 2004). Therefore, background surface moisture is critical to the development of the QBWO and the variation of the QBWO intensity over the WNP during summer. A significant feature of the Asian summer monsoon is the strong background easterly vertical shear (Wang and Xie 1996). The low-frequency moist n = 1 equatorial Rossby wave instability is remarkably enhanced (weakened) by background easterly (westerly) vertical shear (Wang and Xie 1996; Xie and Wang 1996). Background easterly (westerly) vertical shear confines the wave to the lower (upper) level, which generates stronger (weaker) Ekman pumping–induced heating and more (less) meridional heat flux. Furthermore, background easterly vertical shear over the Asian summer monsoon favors the northward propagation of the QBWO (Jiang et al. 2004; Drbohlav and Wang 2005; Wang and Duan 2015). Therefore, background zonal wind vertical shear is important to the development of the QBWO and the variation of the QBWO intensity over the WNP during summer.

Previous studies have revealed sea surface temperature (SST) anomalies in the tropics can influence the interannual and interdecadal variability of the QBWO intensity during summer over the WNP by changing the background atmospheric fields (Zveryaev 2002; Teng and Wang 2003; Lin and Li 2008; Wu and Cao 2016). The QBWO intensity during summer tends to enhance (weaken) over the tropical WNP during the developing El Niño (La Niña) year because of increased background easterly vertical shear and surface moisture (Teng and Wang 2003; Wu and Cao 2016). The developing El Niño events intensify the background easterly vertical shear over the tropical WNP during July–October, and then favor the development and northwestward emanation of moist Rossby waves away from the equatorial western and central Pacific, leading to enhanced 10–20-day westward-propagating intraseasonal oscillation (ISO) between 100°E and 180° (Teng and Wang 2003). The QBWO intensity of precipitation and mean precipitation in summer are significantly and positively correlated over south China and the middle and lower reach of the Yangtze River during past decades (Li et al. 2015a,b). When these regions are wetter than normal, large-scale background atmospheric conditions over the key QBWO activity region (such as the South China Sea and the Philippine Sea) favor the growth of the QBWO perturbations. Using 850-hPa zonal winds, Zveryaev (2002) revealed that the QBWO intensity was enhanced during summer over the tropical western Pacific around the mid- to late 1970s, which was caused by the sea surface warming in the Indian Ocean and associated enhanced convection. The QBWO intensity during mid-April to mid-May over the South China Sea and the Philippine Sea was enhanced around 1993/94, which is related to the sea surface warming near the Philippine Sea (Kajikawa and Wang 2012; Xiang and Wang 2013). Although previous studies investigated the variation of the QBWO intensity during summer over the WNP on the interannual and interdecadal time scales, whether it experienced an interdecadal change during recent decades is unrevealed.

The interdecadal Pacific oscillation (IPO)–Pacific decadal oscillation (PDO) has a pronounced impact on global climate (Huang et al. 1999; Ding et al. 2009; Li et al. 2010; Zhu et al. 2011; Dai 2013; Dong and Lu 2013; Qian and Zhou 2014; Xu et al. 2015; Yu et al. 2015; Zhu et al. 2015). It shifted to a negative phase with La Niña–like SST anomalies in the tropical Pacific (Figs. 1, 2) around the late 1990s, inducing the change in background atmospheric fields. The influence of the phase transition of the IPO around the late 1990s on the seasonal climate (such as summer precipitation) was investigated by previous studies (e.g., Zhu et al. 2011). However, its impact on the QBWO intensity over the WNP during summer is unknown. Therefore, this work tries to address two questions: Can the phase transition of the IPO around the late 1990s influence the QBWO intensity over the WNP during summer on interdecadal time scales via modulating background atmospheric fields? What are the relative roles of the anomalous background atmospheric fields? As revealed in section 4, only the QBWO intensity over the WNP during late summer weakened significantly in response to the phase transition of the IPO. Therefore, the rest of this work is organized as follows. The data and methods used are described in section 2. Section 3 describes the background atmospheric field anomalies during late summer caused by the phase transition of the IPO around the late 1990s. The basic features of the weakened QBWO intensity over the WNP during late summer and the relative contribution of anomalous background atmospheric fields to it are explored in section 4. Finally, conclusions and a discussion are given in section 5.

Fig. 1.

(a) The EOF2 and (b) the corresponding standardized time coefficients of 3-yr running-averaged SST anomalies from 60°S to 60°N during late summer for 1950–2012. The EOF2 is the negative phase of the IPO and accounts for about 20.0% of the total variance. Its time coefficients are the reversed IPO index.

Fig. 1.

(a) The EOF2 and (b) the corresponding standardized time coefficients of 3-yr running-averaged SST anomalies from 60°S to 60°N during late summer for 1950–2012. The EOF2 is the negative phase of the IPO and accounts for about 20.0% of the total variance. Its time coefficients are the reversed IPO index.

Fig. 2.

The difference in SST (°C) during late summer between 1998–2012 and 1980–97. The 0.05 significance level is indicated by the dots.

Fig. 2.

The difference in SST (°C) during late summer between 1998–2012 and 1980–97. The 0.05 significance level is indicated by the dots.

2. Data and method

The datasets employed in this study include 1) daily outgoing longwave radiation (OLR) with a horizontal resolution of 2.5° × 2.5° (Liebmann and Smith 1996), provided by the National Oceanic and Atmospheric Administration (NOAA); 2) 6-hourly and monthly reanalysis datasets from ERA-Interim with a horizontal resolution of 1.5° × 1.5° (Dee et al. 2011); 3) the Global Precipitation Climatology Project (GPCP) monthly precipitation, with a horizontal resolution of 2.5° × 2.5° (Adler et al. 2003); 4) the monthly Climate Prediction Center (CPC) Merged Analysis of Precipitation, with a horizontal resolution of 2.5° × 2.5° (Xie and Arkin 1997); 5) the NOAA Extended Reconstructed SST, version 3b, with a horizontal resolution of 2.0° × 2.0° (Smith et al. 2008). The data during 1979–2013 are used in this study.

The Lanczos bandpass filter (Duchon 1979) was applied to the anomalies of daily data to extract 10–20-day and 20–80-day signals. The active (suppressed) QBWO events were selected when the standardized 10–20-day filtered OLR is less than −1.0 (greater than 1.0) at the peak phase. In this study, the late summer refers to August–November. The ISO intensity for each season is represented by the standard deviation of the filtered atmospheric variables in that season.

In our study, the two-tailed two-sample t test was employed to test the statistical significance of the difference between the averages of two time series. The t statistics is given under the assumption that the two time series come from populations with equal variance:

 
formula

where and are the average of two time series X = {x1, …, xi, …, } and Y = {y1, …, yi, …, }, respectively. The quantities nX and nY are the sample sizes of the two time series. The quantities and , which are the estimates of population variance, are defined as

 
formula
 
formula

Under the null hypothesis, (1) has a t distribution with degrees of freedom N = nX + nY − 2. If the t statistics exceeds the threshold at a certain (0.10 or 0.05) significance level, then the difference between the averages of the two time series is statistically significant at the significance level. The moving t test (MTT), which is based on the two-tailed two-sample t test, was employed to detect the interdecadal changepoint of the area-averaged QBWO intensity around the late 1990s. The MTT was carried out as follows: first, calculate the t statistics in a year using (1) based on the two time series with an equal simple size before and after the year (e.g., the t statistics in 1998 is calculated by the two epochs: 1987–97 and 1998–2008); second, move forward with an interval of one year and repeat the first step. The sample size for the two selected time series is 11 in our study. A year around the late 1990s in which t statistics reached a maximum or minimum exceeding the threshold at the 0.05 significance level is defined as an interdecadal changepoint. If the interdecadal changepoint detected by the MTT is M, then this implies that an interdecadal change occurred around (M − 1)/M. For example, if M is equal to 1998, then an interdecadal change occurred around 1997/98. The Lepage test (Lepage 1971; Liu et al. 2011) was also employed to detect the interdecadal changepoint of area-averaged QBWO intensity around the late 1990s. The results are basically similar to those detected by the MTT (figure omitted).

The intermediate 2.5-layer model with a two-layer free atmosphere and a well-mixed planetary boundary layer (Wang and Xie 1997) was employed to identify the role of background surface moisture and zonal wind vertical shear anomalies in the interdecadal change of the QBWO intensity. Under the equatorial β-plane approximation, the linearized primitive equations in pressure coordinates are employed in the model. Therefore, the background fields in the model are fixed, making it suitable to study the impact of background surface moisture and flow on the intensity and evolution of the ISO. Furthermore, prescribed by the realistic summer background fields, the model captures the essential structure and evolution characteristics of the boreal summer ISO (BSISO), such as the emanation of moist Rossby waves in the western Pacific and the westward and northward propagation of the BSISO in the Asian monsoon region (Wang and Xie 1997). Because of the two advantages described above, the model was used to explore the impact of background fields on the intensity and evolution of the BSISO by previous studies (Wang and Xie 1997; Tao et al. 2015; Deng and Li 2016; Deng et al. 2016).

A detail description of the model and the realistic 3D background fields prescribed in our study are shown in the  appendix. Several experiments are designed to study the changes of the ISO intensity over the WNP under different background atmospheric fields. See more details about experiment designs in Table 1. The EXP-ctrl was integrated for 30 days. The sensitivity experiments are identical to EXP-ctrl except that the background surface moisture or/and zonal flow were changed. Note that the argument based on the experiments changes slightly if the strength of the initial perturbation increases or decreases 50% or if the center of the initial perturbation shifts to 2°N, 175°E (figure omitted).

Table 1.

Experiment designs.

Experiment designs.
Experiment designs.

3. The background atmospheric field anomalies caused by the phase transition of the IPO around the late 1990s

As revealed by previous studies (e.g., Zhu et al. 2011; Wang et al. 2012), the IPO shifted to a negative phase around the late 1990s. In this section, the background atmospheric field anomalies during late summer caused by the phase transition of the IPO are explored. Similar to Wang et al. (2012), the empirical orthogonal function (EOF) analysis was applied to the 3-yr running averaged SST anomalies from 60°S to 60°N during late summer for 1950–2012. The second EOF (EOF2) mode depicts the IPO (Fig. 1). For convenience, the negative phase of the IPO and the reversed IPO index are shown in Fig. 1. The MTT of the reversed IPO index shows the IPO experienced an interdecadal change around 1997/98 (figure omitted). Therefore, the research period (1980–2012) is divided into two epochs: 1980–97 (epoch 1) and 1998–2012 (epoch 2). The difference in variables between epoch 2 and epoch 1 was calculated to explore the anomalous background fields caused by the phase transition of the IPO around the late 1990s.

Corresponding to the negative phase of the IPO, there were La Niña–like SST anomalies, with cold (warm) SST anomalies in the tropical central and eastern (western) Pacific (Figs. 1a, 2), and an enhanced Walker circulation (Dong and Lu 2013). Precipitation decreased over the equatorial central and eastern Pacific as a result of the enhanced Walker circulation (Fig. 3). The decreased precipitation induced a Gill pattern to the west of it (Gill 1980), with a pair of anticyclones (cyclones) in the lower (upper) troposphere (Fig. 4). Anomalous easterlies (westerlies) dominated in the lower (upper) troposphere over the tropical WNP (Fig. 4). This led to anomalous westerly vertical shear over the tropical WNP (Fig. 5d). The anomalous anticyclone in the lower troposphere (Fig. 4a) over the WNP triggered anomalous horizontal divergence and descent motion in the planetary boundary layer through Ekman pumping (Figs. 5a,b) that favored decreased surface moisture (Fig. 5c). The maximum center of the climatological surface moisture during late summer is located east of the Philippines (EOP; figure omitted). Therefore, the anomalous lower-level easterlies led to decreased surface moisture over the east of the EOP by advecting the climatological surface moisture. This may partly explain why the lower-level anomalous divergence is somewhat local over the east of the EOP (Fig. 5b), but the decreased surface moisture is large scale (Fig. 5c). Furthermore, the large-scale climatological convergence (figure omitted), which is related to the monsoon trough, enhanced the decreased surface moisture over the WNP during late summer. The uncertainty of the ERA-Interim data may also have partly led to the inconsistency between the spatial pattern of lower-level anomalous divergence and the decreased surface moisture over the east of the EOP. The decreased surface moisture dominated in ICOADS data over the east of the EOP (figure omitted), but there are some small regions with increased surface moisture. This indicates that there is some uncertainty in the ERA-Interim data. The large-scale spatial pattern of the decreased surface moisture based on ERA-Interim data is basically similar to that based on ICOADS data. Therefore, the decreased surface moisture revealed by ERA-Interim data is basically reliable.

Fig. 3.

The difference in precipitation (mm day−1) during late summer between 1998–2012 and 1980–1997: (a) GPCP and (b) CPC. The 0.05 significance level is indicated by the dots.

Fig. 3.

The difference in precipitation (mm day−1) during late summer between 1998–2012 and 1980–1997: (a) GPCP and (b) CPC. The 0.05 significance level is indicated by the dots.

Fig. 4.

The difference in wind (m s−1) at the (a) 850- and (b) 200-hPa levels during late summer between 1998–2012 and 1980–97. The 0.10 (light shading) and 0.05 (dark shading) significance levels are indicated. The red letters A and C denote anticyclones and cyclones, respectively.

Fig. 4.

The difference in wind (m s−1) at the (a) 850- and (b) 200-hPa levels during late summer between 1998–2012 and 1980–97. The 0.10 (light shading) and 0.05 (dark shading) significance levels are indicated. The red letters A and C denote anticyclones and cyclones, respectively.

Fig. 5.

The difference in (a) 850-hPa omega (10−2 Pa s−1), (b) 1000-hPa horizontal divergence (10−6 s−1), (c) 1000-hPa specific humidity (g kg−1), and (d) zonal flow vertical shear (200-hPa zonal wind minus 850-hPa zonal wind, m s−1) during late summer between 1998–2012 and 1980–97. The 0.05 significance level in (a)–(c) and 0.10 significance level in (d) are indicated by dots. The small rectangle indicates the EOP (10°–20°N, 132.5°–150°E), which is the core region with the weakened QBWO intensity. The large rectangle (4°–28°N, 125°–170°E) indicates the region where the decreased background specific humidity at the 1000-hPa level and anomalous background westerly vertical shear are put into the model.

Fig. 5.

The difference in (a) 850-hPa omega (10−2 Pa s−1), (b) 1000-hPa horizontal divergence (10−6 s−1), (c) 1000-hPa specific humidity (g kg−1), and (d) zonal flow vertical shear (200-hPa zonal wind minus 850-hPa zonal wind, m s−1) during late summer between 1998–2012 and 1980–97. The 0.05 significance level in (a)–(c) and 0.10 significance level in (d) are indicated by dots. The small rectangle indicates the EOP (10°–20°N, 132.5°–150°E), which is the core region with the weakened QBWO intensity. The large rectangle (4°–28°N, 125°–170°E) indicates the region where the decreased background specific humidity at the 1000-hPa level and anomalous background westerly vertical shear are put into the model.

Overall, the phase transition of the IPO around the late 1990s led to decreased surface moisture and anomalous westerly vertical shear over the tropical WNP via air–sea interaction in the tropical Pacific. These background atmospheric field anomalies modulated the interdecadal variation of the QBWO intensity over the WNP, as revealed in the following section.

4. The weakened intensity of atmospheric QBWO over the WNP during late summer around the late 1990s

a. Basic features

As revealed in section 3, the decreased background surface moisture and the anomalous background westerly vertical shear over the WNP during late summer are remarkable in epoch 2. As a result, the QBWO intensity decreased significantly there, which is reflected by various atmospheric variables (Fig. 6). The core region with the weakened QBWO intensity is located in the EOP. Because of the Rossby wave response to the anomalous convection over the EOP, the QBWO intensity of zonal wind in the lower troposphere decreased southwest and northwest of it (Fig. 6g). The significantly weakened QBWO intensity of specific humidity, geopotential height, and meridional wind in the lower troposphere also appeared north of the EOP (Figs. 6c,d,h). We checked the difference in the QWBO intensity (represented by standard deviations of 10–20-day filtered OLR) by sliding bimonthly from May to December over the WNP between 1998–2012 and 1980–1997 (figure omitted) and found that the QBWO intensity decreased from August to November. Because of the long memory of the ocean, the climatological background surface moisture and the zonal wind vertical shear in November over the tropical WNP are somewhat similar to those in August–October (figure omitted). Therefore, the late summer is defined as August–November in our study. Besides, over the far northwest of the EOP, six of eight variables show enhanced QBWO intensity (Fig. 6), which is closely related to the local increased surface moisture (Fig. 5c) and enhanced background convection caused by the local background sea surface warming (Figs. 2, 3). Based on the 20–80-day filtered OLR, we found that there was weakened intensity of 20–80-day ISO along 10°N over the WNP, but the statistical significance is weak (figure omitted).

Fig. 6.

Difference in the QBWO intensity during late summer between 1998–2012 and 1980–97. The QBWO intensity is represented by the standard deviations of the following variables: (a) OLR (W m−2), (b) 600–400-hPa omega (10−2 Pa s−1), (c) 1000–700-hPa specific humidity (g kg−1), (d) 1000–700-hPa geopotential height (gpm), (e) 1000–700-hPa vertical relative vorticity (10−5 s−1), (f) 1000–700-hPa horizontal divergence (10−6 s−1), (g) 1000–700-hPa zonal wind (m s−1), and (h) 1000–700-hPa meridional wind (m s−1). The standard deviations are calculated for each pressure level and then the average of the multi-pressure levels is taken. The 0.10 (dashed contour) and 0.05 (solid contour) significance levels are indicated. The red rectangle indicates the EOP (10°–20°N, 132.5°–150°E) where the area-averaged QBWO intensity of variables except the zonal wind is calculated. The purple rectangle in (g) indicates the region (southwest of the EOP; 7.5°–16.5°N, 120°–139.5°E) where the area-averaged QBWO intensity of zonal wind is calculated.

Fig. 6.

Difference in the QBWO intensity during late summer between 1998–2012 and 1980–97. The QBWO intensity is represented by the standard deviations of the following variables: (a) OLR (W m−2), (b) 600–400-hPa omega (10−2 Pa s−1), (c) 1000–700-hPa specific humidity (g kg−1), (d) 1000–700-hPa geopotential height (gpm), (e) 1000–700-hPa vertical relative vorticity (10−5 s−1), (f) 1000–700-hPa horizontal divergence (10−6 s−1), (g) 1000–700-hPa zonal wind (m s−1), and (h) 1000–700-hPa meridional wind (m s−1). The standard deviations are calculated for each pressure level and then the average of the multi-pressure levels is taken. The 0.10 (dashed contour) and 0.05 (solid contour) significance levels are indicated. The red rectangle indicates the EOP (10°–20°N, 132.5°–150°E) where the area-averaged QBWO intensity of variables except the zonal wind is calculated. The purple rectangle in (g) indicates the region (southwest of the EOP; 7.5°–16.5°N, 120°–139.5°E) where the area-averaged QBWO intensity of zonal wind is calculated.

Figure 7 shows the vertical profile of the difference in the QBWO intensity over the EOP between the two epochs. The weakened QBWO intensity mainly occurred in the middle and lower troposphere. In the upper troposphere, the QBWO intensity of the vertical relative vorticity, the horizontal divergence, the zonal wind, and the meridional wind was enhanced in spite of weak statistical significance (Figs. 7d–g). This may be partly caused by the anomalous background westerly vertical shear over the WNP, which tends to trap the Rossby waves in the upper troposphere (Wang and Xie 1996). Detected by MTT, the area-averaged QBWO intensity over the EOP in the middle and lower troposphere experienced an interdecadal decrease around 1997/98 (Figs. 8a,c–f,h), corresponding to the phase transition of the IPO (Fig. 1b). To the southwest of the EOP, the QBWO intensity of the zonal wind in the lower troposphere also underwent an interdecadal decrease around 1997/98 (Fig. 8g). The QBWO intensity decreased about 10.0%–20.0% for the atmospheric variables (Fig. 8).

Fig. 7.

Vertical profile of the difference in the QBWO intensity during late summer between 1998–2012 and 1980–97. The QBWO intensity is represented by the standard deviations of the following variables: (a) omega (10−2 Pa s−1), (b) specific humidity (g kg−1), (c) geopotential height (gpm), (d) vertical relative vorticity (10−5 s−1), (e) horizontal divergence (10−6 s−1), (f) zonal wind (m s−1), and (g) meridional wind (m s−1). The area-averaged QBWO intensity is calculated over the southwest of the EOP for zonal wind and over the EOP for other variables. The 0.10 (blue circles) and 0.05 (red circles) significance levels are indicated.

Fig. 7.

Vertical profile of the difference in the QBWO intensity during late summer between 1998–2012 and 1980–97. The QBWO intensity is represented by the standard deviations of the following variables: (a) omega (10−2 Pa s−1), (b) specific humidity (g kg−1), (c) geopotential height (gpm), (d) vertical relative vorticity (10−5 s−1), (e) horizontal divergence (10−6 s−1), (f) zonal wind (m s−1), and (g) meridional wind (m s−1). The area-averaged QBWO intensity is calculated over the southwest of the EOP for zonal wind and over the EOP for other variables. The 0.10 (blue circles) and 0.05 (red circles) significance levels are indicated.

Fig. 8.

Standardized QBWO intensity during late summer from 1980 to 2012. The QBWO intensity is represented by the standard deviations of the following variables: (a) OLR, (b) 600–400-hPa omega, (c) 1000–700-hPa specific humidity, (d) 1000–700-hPa geopotential height, (e) 1000–700-hPa vertical relative vorticity, (f) 1000–700-hPa horizontal divergence, (g) 1000–700-hPa zonal wind, and (h) 1000–700-hPa meridional wind. The standard deviations are calculated for each pressure level and then the average of the multi-pressure levels is taken. The area-averaged QBWO intensity is calculated over the southwest of the EOP for zonal wind and over the EOP for other variables. The decreased percent of the QBWO intensity for each variable in 1998–2012 relative to that in 1980–97 is given in the top-right corner of each plot. The interdecadal changepoint detected by the MTT is also given. The asterisk (*) with the MTT in (f) indicates that the interdecadal changepoint is significant only at 0.10 significance level.

Fig. 8.

Standardized QBWO intensity during late summer from 1980 to 2012. The QBWO intensity is represented by the standard deviations of the following variables: (a) OLR, (b) 600–400-hPa omega, (c) 1000–700-hPa specific humidity, (d) 1000–700-hPa geopotential height, (e) 1000–700-hPa vertical relative vorticity, (f) 1000–700-hPa horizontal divergence, (g) 1000–700-hPa zonal wind, and (h) 1000–700-hPa meridional wind. The standard deviations are calculated for each pressure level and then the average of the multi-pressure levels is taken. The area-averaged QBWO intensity is calculated over the southwest of the EOP for zonal wind and over the EOP for other variables. The decreased percent of the QBWO intensity for each variable in 1998–2012 relative to that in 1980–97 is given in the top-right corner of each plot. The interdecadal changepoint detected by the MTT is also given. The asterisk (*) with the MTT in (f) indicates that the interdecadal changepoint is significant only at 0.10 significance level.

To further demonstrate the weakened QBWO intensity over the WNP, the evolution of the QBWO events in the two epochs are shown in Figs. 9 and 10. The QBWO active (suppressed) events were selected when the standardized 10–20-day filtered OLR over the EOP exceeds minus (plus) one standard deviation at the peak phase. There are 106 (69) active and 110 (59) suppressed events in epoch 1 (epoch 2). Fewer QBWO events in epoch 2 also indicate that the QBWO intensity weakened in epoch 2. The composite evolution of the QBWO events are the difference between the QBWO active and suppressed events. The results are similar when the evolution of the QBWO active or suppressed events are analyzed solely. Day 0 indicates the reference time when OLR anomalies over the EOP are at the peak phase. From day −6 to day +3, the phase evolution of the QBWO events are similar in the two epochs (Fig. 9). At day −6, there are positive vertical relative vorticity perturbations and an anomalous cyclone at the 850-hPa level, corresponding to the enhanced convection over the southeast region of the EOP. The perturbations become enhanced while moving northwestward from day −6 to day 0 and reach the peak phase over the EOP at day 0. From day 0 to day +3, the perturbations weaken while propagating northwestward. However, because of the decreased background surface moisture and the anomalous background westerly vertical shear not favoring the development of perturbations, the perturbations are much weaker in epoch 2 from day −3 to day +3 relative to those in epoch 1 (Fig. 9). At day−7 and day−5 (figure omitted), the perturbations are basically the same in the two epochs over the equatorial western Pacific, from which the QBWO events originated (Mao and Chan 2005; Chen and Sui 2010). Therefore, the anomalous background atmospheric fields mainly influenced the growth of the QBWO perturbations while they propagate northwestward over the WNP instead of their initiation. The local evolution of the perturbations over the EOP during QBWO events are also similar in the two epochs (Fig. 10), but the amplitude of the perturbations is weaker in epoch 2. For the OLR, the 850-hPa specific humidity, the 850-hPa vertical relative vorticity, and the 850-hPa kinetic energy, the amplitude decreased about 13.2%, 27.6%, 19.6%, and 23.9%, respectively (Fig. 10). The difference in perturbations between the two epochs is at its maximum when the perturbations reach the peak phase (Fig. 10).

Fig. 9.

Composite evolution of the 10–20-day filtered OLR (shaded, W m−2), 850-hPa vertical relative vorticity (contours, 10−5 s−1), and 850-hPa wind (vectors, m s−1) for the QBWO events during late summer over the WNP at day −6, −3, 0, and 3 for (left) 1980–97 and (right) 1998–2012. Contours change from −1.2 to 1.2 at intervals of 0.6, and the 0 contour is omitted. The anomalies of the QBWO events are the difference between the QBWO active and suppressed events. The QBWO active (suppressed) events were selected when the standardized 10–20-day filtered OLR over the EOP exceeds minus (plus) one standard deviation at the peak phase. Day 0 indicates the reference time when the 10–20-day filtered OLR over the EOP is at the peak phase. The plotted fields are significant at the 0.05 significance level.

Fig. 9.

Composite evolution of the 10–20-day filtered OLR (shaded, W m−2), 850-hPa vertical relative vorticity (contours, 10−5 s−1), and 850-hPa wind (vectors, m s−1) for the QBWO events during late summer over the WNP at day −6, −3, 0, and 3 for (left) 1980–97 and (right) 1998–2012. Contours change from −1.2 to 1.2 at intervals of 0.6, and the 0 contour is omitted. The anomalies of the QBWO events are the difference between the QBWO active and suppressed events. The QBWO active (suppressed) events were selected when the standardized 10–20-day filtered OLR over the EOP exceeds minus (plus) one standard deviation at the peak phase. Day 0 indicates the reference time when the 10–20-day filtered OLR over the EOP is at the peak phase. The plotted fields are significant at the 0.05 significance level.

Fig. 10.

Composite local evolution of 10–20-day filtered (a) OLR (W m−2), (b) 850-hPa specific humidity (g kg−1), (c) 850-hPa vertical relative vorticity (10−5 s−1), and (d) 850-hPa kinetic energy (m2 s−2) over the EOP for the QBWO events during late summer for 1980–97 (red curves) and 1998–2012 (blue curves). Day 0 indicates the reference time when 10–20-day filtered OLR over the EOP is at the peak phase. Different evolution periods were selected based on the time when the anomalies of variables are at the peak phase. For OLR and 850-hPa specific humidity, the period from day −3 to day 3 is given. For 850-hPa vertical relative vorticity and kinetic energy, the period from day −4 to day 2 and day −1 to day 5 are given, respectively. The kinetic energy was calculated based on 10–20-day filtered zonal and meridional winds. The decreased percent of the accumulated anomaly during the selected period for the QBWO events, and the QBWO active and suppressed events in 1998–2012 relative to those in 1980–97 are given in each plot.

Fig. 10.

Composite local evolution of 10–20-day filtered (a) OLR (W m−2), (b) 850-hPa specific humidity (g kg−1), (c) 850-hPa vertical relative vorticity (10−5 s−1), and (d) 850-hPa kinetic energy (m2 s−2) over the EOP for the QBWO events during late summer for 1980–97 (red curves) and 1998–2012 (blue curves). Day 0 indicates the reference time when 10–20-day filtered OLR over the EOP is at the peak phase. Different evolution periods were selected based on the time when the anomalies of variables are at the peak phase. For OLR and 850-hPa specific humidity, the period from day −3 to day 3 is given. For 850-hPa vertical relative vorticity and kinetic energy, the period from day −4 to day 2 and day −1 to day 5 are given, respectively. The kinetic energy was calculated based on 10–20-day filtered zonal and meridional winds. The decreased percent of the accumulated anomaly during the selected period for the QBWO events, and the QBWO active and suppressed events in 1998–2012 relative to those in 1980–97 are given in each plot.

b. Relative role of decreased background surface moisture and anomalous background westerly vertical shear based on a modeling study

Both the anomalous decreased background surface moisture and the background westerly vertical shear contributed to the weakened QBWO intensity. What about their relative contribution? The question is addressed by a modeling study in this subsection. We mainly focus on the core region (that is the EOP) with the weakened QBWO intensity. The precipitation and lower-level kinetic energy over this region are used to measure the ISO intensity in the model. In the control experiment, prescribed by climatological background fields and initiated by a perturbation centered at 4°N, 170°E, the 2.5-layer model can basically capture the propagation and development of the initial perturbation over the equatorial western Pacific related to the observed northwestward-propagating QBWO over the WNP during late summer for 1980–97 (cf. Figs. 9, 11). The wavelength, northward, and westward-propagating speeds of the low-frequency perturbation in the control experiment (EXP-ctrl) are similar to those of the observed QBWO. The wavelength of the low-frequency perturbation in the EXP-ctrl is about 6000 km that is consistent with that of the observed QBWO (cf. Figs. 9, 11). The northward- and westward-propagating speeds of the low-frequency perturbation in the EXP-ctrl are about 1.2 and 6.4 m s−1 respectively, which are similar to those of the observed QBWO (about 1.6 and 4.1 m s−1, respectively). The low-frequency perturbation in the EXP-ctrl also reaches its peak phase around the EOP (Fig. 11). The amplitude of low-frequency zonal wind in the EXP-ctrl is comparable to that related to the observed QBWO (cf. Figs. 9, 11). Note that Fig. 9 shows the difference between the QBWO active and suppressed events. Therefore, the amplitude of wind for the QBWO active or suppressed events during 1980–1997 is about half of that shown in the left panel of Fig. 9. The amplitude of the low-frequency meridional wind in the EXP-ctrl is much weaker than that related to the observed QBWO (cf. Figs. 9, 11). Based on the discussion above, the 2.5-layer model can be used to carry out experiments to identify the influence of the background atmospheric fields on the development of the initial perturbation over the equatorial western Pacific related to the observed QBWO, which is the issue with which our study is concerned. Because the 2.5-layer model is a simplified model, it cannot reproduce the oscillation period of the QBWO. The mechanisms for the periodic oscillation of the QBWO are beyond the scope of our study.

Fig. 11.

Evolution of the precipitation (contours, mm day−1) and lower-level wind (vectors, m s−1) at days 2, 3, 5, and 7 in EXP-ctrl.

Fig. 11.

Evolution of the precipitation (contours, mm day−1) and lower-level wind (vectors, m s−1) at days 2, 3, 5, and 7 in EXP-ctrl.

For the sensitivity experiments, the decreased background surface moisture and/or anomalous background westerly vertical shear over the WNP were added to the climatology (Table 1). In the sensitivity experiments, the evolution of the low-frequency perturbation is similar to that in EXP-ctrl (Fig. 11). However, the amplitude of the low-frequency perturbation over the EOP weakens remarkably (Fig. 12). The amplitude is weakest in EXP-qs&u_shear. The amplitude is weaker in EXP-qs than that in EXP-u_shear (Fig. 12). This indicates that the decreased background surface moisture plays a more important role than the anomalous background westerly vertical shear in weakening the QBWO intensity. Compared to EXP-ctrl, the accumulated precipitation (lower level of kinetic energy) over the EOP in the EXP-qs&u_shear, EXP-qs, and EXP-u_shear decreases 24.1% (33.3%), 17.2% (22.2%), and 9.5% (16.3%), respectively (Fig. 13). EXP-qs (EXP-u_shear) accounts for about 70% (45%) of the decreased accumulated precipitation and the lower level of kinetic energy in the EXP-qs&u_shear. These results further support the finding that the decreased background surface moisture is more important to the weakened QBWO intensity over the EOP.

Fig. 12.

Local evolution of the precipitation (mm day−1) and lower-level kinetic energy (m2 s−2) over the EOP from day 3 to day 10 in the EXP-ctrl (red), EXP-qs&u_shear (blue), EXP-qs (green), and EXP-u_shear (purple). For the modeling results, the EOP refers to 10°–20°N, 130°–150°E, which is slightly different from the observation (10°–20°N, 132.5°–150°E).

Fig. 12.

Local evolution of the precipitation (mm day−1) and lower-level kinetic energy (m2 s−2) over the EOP from day 3 to day 10 in the EXP-ctrl (red), EXP-qs&u_shear (blue), EXP-qs (green), and EXP-u_shear (purple). For the modeling results, the EOP refers to 10°–20°N, 130°–150°E, which is slightly different from the observation (10°–20°N, 132.5°–150°E).

Fig. 13.

Accumulated precipitation (mm) at left and lower-level kinetic energy (m2 s−2) at right over the EOP from day 3 to day 10 in the EXP-ctrl (red), EXP-qs&u_shear (blue), EXP-qs (green), and EXP-u_shear (purple). The decreased percent of accumulated precipitation and lower level kinetic energy in the EXP-qs&u_shear, EXP-qs, and EXP-u_shear relative to those in the EXP-ctrl are given on the top of the corresponding bars.

Fig. 13.

Accumulated precipitation (mm) at left and lower-level kinetic energy (m2 s−2) at right over the EOP from day 3 to day 10 in the EXP-ctrl (red), EXP-qs&u_shear (blue), EXP-qs (green), and EXP-u_shear (purple). The decreased percent of accumulated precipitation and lower level kinetic energy in the EXP-qs&u_shear, EXP-qs, and EXP-u_shear relative to those in the EXP-ctrl are given on the top of the corresponding bars.

5. Conclusions and discussion

The IPO shifted to a negative phase around the late 1990s. In the present study, how this interdecadal change influenced the background atmospheric fields and hence modulated the QBWO intensity over the WNP during late summer is demonstrated. There were La Niña–like SST anomalies and an enhanced Walker circulation in response to the phase transition of the IPO. This led to decreased precipitation in the equatorial central and eastern Pacific. The decreased precipitation induced a Gill pattern to the west of it, with an anomalous anticyclone (cyclone) in the lower (upper) troposphere over the WNP. Therefore, anomalous background westerly vertical shear dominated over the tropical WNP. Furthermore, the anomalous anticyclone triggered anomalous horizontal divergence and descent motion in the planetary boundary layer, which led to decreased surface moisture over the tropical WNP. The climatological convergence and the zonal advection of climatological surface moisture by the anomalous lower-level easterlies also contributed to the decreased surface moisture over the tropical WNP.

The decreased background surface moisture and anomalous background westerly vertical shear suppressed the development of QBWO perturbations and weakened the QBWO intensity over the WNP during late summer. As reflected by various atmospheric variables, the core region with the weakened QBWO intensity is located in the lower and middle troposphere over the EOP. Based on the standardized 10–20-day filtered OLR over the EOP, QBWO active and suppressed events in the two epochs were selected. Fewer QBWO events occurred after the late 1990s. The composite evolution of the QBWO events in the two epochs further demonstrates that the QBWO intensity weakened after the late 1990s.

Prescribed by the decreased background surface moisture or/and anomalous background westerly vertical shear over the tropical WNP, a modeling study was carried out to explore the relative contribution of the two factors to the weakened QBWO intensity over the EOP. Based on the evolution of area-averaged precipitation and lower kinetic energy over the EOP and their accumulation, the decreased background surface moisture plays a more important role than the anomalous background westerly vertical shear. Note that the convective parameterization in the model is based on the vertically integrated moisture convergence. A further study with different convective parameterizations is needed to identify the relative role of the decreased background surface moisture and the anomalous background westerly vertical shear.

The QBWO events over the WNP during late summer originate from the equatorial western Pacific (Mao and Chan 2005; Chen and Sui 2010). The QBWO intensity over the WNP mainly is modulated by the local background atmospheric fields. Therefore, the QBWO intensity weakened remarkably in response to the phase transition of the IPO around the late 1990s. Whereas the 20–80-day ISO events over the WNP originate from the equatorial Indian Ocean (Annamalai and Slingo 2001; Kemball-Cook and Wang 2001), the intensity of the 20–80-day ISO over the WNP is not only influenced by the local background atmospheric fields, but also modulated by other factors. This may explain why the response of the 20–80-day ISO intensity over the WNP to the phase transition of the IPO is weak. Besides, the anomalies of the background atmospheric fields induced by the phase transition of the IPO (Fig. 3) and the QBWO activities over the tropical Pacific during summer are region dependent (Kikuchi and Wang 2009). Therefore, influence of the phase transition of the IPO on the QBWO intensity over other regions in the tropical Pacific during late summer needs further study.

The La Niña–like SST anomalies in the tropical Pacific (Fig. 2) are the combined contribution of the IPO, global warming, and the Atlantic multidecadal oscillation (AMO) (Dong and Zhou 2014). The cooling over the tropical central and eastern Pacific is primarily dominated by the phase transition of the IPO, which overwhelms the global warming signal and warming caused by the AMO (Dong and Zhou 2014). In the tropical western Pacific, the warming caused by the phase transition of the IPO is enhanced by the global warming signal. There is significant warming in the tropical Indian Ocean (TIO) and the tropical Atlantic (TA; Fig. 2). Recent studies show that the sea surface warming in the two regions also contributed to the La Niña–like SST anomalies and enhanced the Walker circulation in the tropical Pacific during the past three decades (Luo et al. 2012; Han et al. 2014; McGregor et al. 2014; Li et al. 2015c; Kucharski et al. 2016). The correlations of the QBWO intensity over the EOP and the background atmospheric fields over the WNP during late summer with low-frequency (greater than 3 years) SST modes indicates that the IPO may be more crucial to the weakened QBWO intensity (Table 2). When the reversed IPO signal is removed, the correlations of the QBWO intensity and the background atmospheric fields with low-frequency SST modes in the tropical Indian and Atlantic Oceans change remarkably (columns 4–6 of Table 2), whereas the changes for the correlations of the QBWO intensity and background atmospheric fields with the reversed IPO are relatively small (column 3 of Table 2) when removing all of the low-frequency SST modes in the tropical Indian and Atlantic Ocean. Furthermore, the interdecadal changepoint of the weakened QBWO intensity around the late 1990s is consistent with that of the IPO. Therefore, the phase transition of the IPO probably is more important to the weakened QBWO intensity. The mechanisms for the phase transition of the IPO around the late 1990s and its connection with the sea surface warming in the tropical Indian Ocean and Atlantic Ocean, which need to be studied further, are beyond the scope of this study.

Table 2.

Correlations of the QBWO intensity over the EOP (see Fig. 8) and the background atmospheric fields over the WNP during late summer with low-frequency (greater than 3 yr) SST modes. The low-frequency SST modes include the reversed IPO, the EOF1 of 3-yr running averaged SST anomalies in the tropical Indian Ocean (25°S–25°N, 40°–120°E), and the EOF1 and EOF2 of 3-yr running averaged SST anomalies in the tropical Atlantic Ocean (25°S–25°N, 290°–358°E). Values in the parentheses of column 3 are correlations of the QBWO intensity and background atmospheric fields with the reversed IPO when all of the low-frequency SST modes in the tropical Indian and Atlantic Ocean are removed. Values in the parentheses of columns 4–6 are the correlations of the QBWO intensity and background atmospheric fields, respectively, with the low-frequency SST modes in the TIO and TA when the reversed IPO signal is removed. A 3-yr running average is applied to the time series of the QBWO intensity and background atmospheric fields before calculating the correlations and partial correlations. The WNP refers to the region marked by the large rectangle in Fig. 5.

Correlations of the QBWO intensity over the EOP (see Fig. 8) and the background atmospheric fields over the WNP during late summer with low-frequency (greater than 3 yr) SST modes. The low-frequency SST modes include the reversed IPO, the EOF1 of 3-yr running averaged SST anomalies in the tropical Indian Ocean (25°S–25°N, 40°–120°E), and the EOF1 and EOF2 of 3-yr running averaged SST anomalies in the tropical Atlantic Ocean (25°S–25°N, 290°–358°E). Values in the parentheses of column 3 are correlations of the QBWO intensity and background atmospheric fields with the reversed IPO when all of the low-frequency SST modes in the tropical Indian and Atlantic Ocean are removed. Values in the parentheses of columns 4–6 are the correlations of the QBWO intensity and background atmospheric fields, respectively, with the low-frequency SST modes in the TIO and TA when the reversed IPO signal is removed. A 3-yr running average is applied to the time series of the QBWO intensity and background atmospheric fields before calculating the correlations and partial correlations. The WNP refers to the region marked by the large rectangle in Fig. 5.
Correlations of the QBWO intensity over the EOP (see Fig. 8) and the background atmospheric fields over the WNP during late summer with low-frequency (greater than 3 yr) SST modes. The low-frequency SST modes include the reversed IPO, the EOF1 of 3-yr running averaged SST anomalies in the tropical Indian Ocean (25°S–25°N, 40°–120°E), and the EOF1 and EOF2 of 3-yr running averaged SST anomalies in the tropical Atlantic Ocean (25°S–25°N, 290°–358°E). Values in the parentheses of column 3 are correlations of the QBWO intensity and background atmospheric fields with the reversed IPO when all of the low-frequency SST modes in the tropical Indian and Atlantic Ocean are removed. Values in the parentheses of columns 4–6 are the correlations of the QBWO intensity and background atmospheric fields, respectively, with the low-frequency SST modes in the TIO and TA when the reversed IPO signal is removed. A 3-yr running average is applied to the time series of the QBWO intensity and background atmospheric fields before calculating the correlations and partial correlations. The WNP refers to the region marked by the large rectangle in Fig. 5.

Acknowledgments

The authors thank Dr. Lu Wang for sharing the model code. The authors are grateful to the three anonymous reviewers for their insightful comments, which led to a significant improvement of the manuscript. This research was jointly supported by the National Natural Science Foundation of China (Grant 41421004), the China National 973 project 2015CB453200, the National Natural Science Foundation of China (Grants 41325018, 41730964, and 41575079), and the CAS/SAFEA International Partnership Program for Creative Research Team “Regional environmental high resolution numerical simulation.”

APPENDIX

A 2.5-Layer Atmospheric Model

The model includes a two-level free atmosphere and a well-mixed planetary boundary layer (Wang and Xie 1997). For the free atmosphere, the low-frequency motion is controlled by the following linearized primitive equations in a pressure coordinate and equatorial β plane:

 
formula
 
formula
 
formula
 
formula
 
formula

where an overbar and a prime denote the summer background basic-state and low-frequency perturbation quantities, respectively; u, υ, ω, ϕ, and T denote zonal and meridional winds, vertical pressure velocity, geopotential, and temperature, respectively; ϵ and μ denote the Rayleigh friction and Newtonian cooling coefficients respectively; K is the horizontal momentum or thermal diffusion coefficient; and S is the dry static stability parameter. To establish a standard two-level free atmosphere model, the momentum and continuity equations are written at the upper- and lower-tropospheric levels of p1 and p2, and the thermodynamic and hydrostatic equations are written at the middle level, pm = (1/2)(p1 + p2).

Upon neglecting the local surface evaporation and the local change of moisture content, the precipitation rate P′ is estimated by the vertically integrated moisture convergence of unit cross section. Further neglecting the horizontal gradient of moisture content, the precipitation rate can be written as

 
formula

where b is a coefficient measuring the fraction of the moisture convergence converting to precipitation; v is three-dimensional wind; ps and pt are pressure at the surface and top of atmosphere, respectively; represents background specific humidity; g represents gravity acceleration; Δp is the half-depth of the free atmosphere; is vertical pressure velocity at the top of the planetary boundary layer; and , , and represent the vertically averaged background specific humidity in the boundary layer, and the lower and upper troposphere, respectively. Based on the moisture and heat budget following Kuo (1974), the condensational heating rate Q′ as a result of deep convection can be written as

 
formula

where δ is the heating switch-on function with a value of 1 (0) for column moisture convergence (divergence). Therefore, condensation heating rate at the middle level pm is parameterized as follows

 
formula

A well-mixed planetary boundary layer is used; for large-scale low-frequency tropical motion, the momentum balance approximately holds among the pressure gradient force, the Coriolis force, and a Rayleigh friction that represents the friction and the high-frequency transient eddy effects. The governing equations are

 
formula
 
βyuBϕeyEyυB=0,and
formula
(A5b)
 
formula

where and are the vertically averaged perturbation zonal and meridional winds, respectively, in the planetary boundary layer; is the pressure at the top of the planetary boundary; represents the perturbation geopotential at the top of the planetary boundary layer; and Ex and Ey are Rayleigh friction coefficients in the zonal and meridional directions, respectively.

Based on (A5), the vertical pressure velocity at the top of the planetary boundary layer is expressed as

 
formula

where the coefficients are

 
formula
 
formula
 
formula
 
formula

Meanwhile, is related to free atmosphere convergence by mass conservation in a vertical column:

 
formula

where and are the zonal and meridional winds, respectively, in the free atmosphere. An elliptical equation can be derived for by (A6) and (A7):

 
formula

With the approximation of , (A1) and (A8) form a closed set of equations.

The model covers from 40°S to 40°N with a horizontal resolution of 5° longitude × 2° latitude. The fluxes of mass, momentum, and heat normal to the meridional boundaries vanish. The zonal boundary condition is periodic around the globe. Finite differences are used in both space and time. The time step is 10 min. A Fourier transform is carried out in the zonal direction with a truncation of one-fourth of the zonal grid points to solve (A8). For each wavenumber, an equation with y derivatives can be obtained and solved by matrix inversion. The model parameters used in the numerical calculation are listed in Table A1.

Table A1.

Model parameters.

Model parameters.
Model parameters.

The background flow used in the model is the climatological circulation during late summer of 1980–97 based on ERA-Interim data. The climatological wind at the 850- and 200-hPa levels, which exhibits strong easterly vertical shear over the Asian summer monsoon region (Fig. A1), is used as the background flow at the lower and upper levels in the model, respectively. The background vertical pressure velocity at the middle level of the model was calculated by the background flow according to the continuity equation. The climatological specific humidity at the 1000-hPa level is defined as the background surface specific humidity of the model (Fig. A2). Following Wang (1988), the vertically averaged background specific humidity at a layer between pi and pj (i < j) is determined by the background surface specific humidity and can be calculated as follows:

 
formula

where m = H/H1 (see Table A1); and H and H1 are the air density and water density scale heights, respectively. The background temperature in the middle level was derived from the geopotential thickness between 200 and 850 hPa based on a hydrostatic balance.

Fig. A1.

The climatological wind (m s−1) at the (a) 850- and (a) 200-hPa levels during late summer for 1980–97.

Fig. A1.

The climatological wind (m s−1) at the (a) 850- and (a) 200-hPa levels during late summer for 1980–97.

Fig. A2.

The climatological specific humidity (g kg−1) at the 1000-hPa level during late summer for 1980–97. A nine-point smoothing is carried out.

Fig. A2.

The climatological specific humidity (g kg−1) at the 1000-hPa level during late summer for 1980–97. A nine-point smoothing is carried out.

REFERENCES

REFERENCES
Adler
,
R. F.
, and Coauthors
,
2003
:
The Version-2 Global Precipitation Climatology Project (GPCP) monthly precipitation analysis (1979–present)
.
J. Hydrometeor.
,
4
,
1147
1167
, doi:.
Annamalai
,
H.
, and
J.
Slingo
,
2001
:
Active/break cycles: Diagnosis of the intraseasonal variability of the Asian summer monsoon
.
Climate Dyn.
,
18
,
85
102
, doi:.
Chatterjee
,
P.
, and
N.
Goswami
,
2004
:
Structure, genesis and scale selection of the tropical quasi-biweekly mode
.
Quart. J. Roy. Meteor. Soc.
,
130
,
1171
1194
, doi:.
Chen
,
G.
, and
C.-H.
Sui
,
2010
:
Characteristics and origin of quasi-biweekly oscillation over the western North Pacific during boreal summer
.
J. Geophys. Res.
,
115
,
D14113
, doi:.
Chen
,
J.
,
Z.
Wen
,
R.
Wu
,
Z.
Chen
, and
P.
Zhao
,
2015
:
Influences of northward propagating 25–90-day and quasi-biweekly oscillations on eastern China summer rainfall
.
Climate Dyn.
,
45
,
105
124
, doi:.
Dai
,
A.
,
2013
:
The influence of the inter-decadal Pacific oscillation on US precipitation during 1923–2010
.
Climate Dyn.
,
41
,
633
646
, doi:.
Dee
,
D.
, and Coauthors
,
2011
:
The ERA‐Interim reanalysis: Configuration and performance of the data assimilation system
.
Quart. J. Roy. Meteor. Soc.
,
137
,
553
597
, doi:.
Deng
,
L.
, and
T.
Li
,
2016
:
Relative roles of background moisture and vertical shear in regulating interannual variability of boreal summer intraseasonal oscillations
.
J. Climate
,
29
,
7009
7025
, doi:.
Deng
,
L.
,
T.
Li
,
J.
Liu
, and
M.
Peng
,
2016
:
Factors controlling the interannual variations of MJO intensity
.
J. Meteor. Res.
,
30
,
328
340
, doi:.
Ding
,
Y.
,
Y.
Sun
,
Z.
Wang
,
Y.
Zhu
, and
Y.
Song
,
2009
:
Inter-decadal variation of the summer precipitation in China and its association with decreasing Asian summer monsoon Part II: Possible causes
.
Int. J. Climatol.
,
29
,
1926
1944
, doi:.
Dong
,
B.
, and
R.
Lu
,
2013
:
Interdecadal enhancement of the Walker circulation over the Tropical Pacific in the late 1990s
.
Adv. Atmos. Sci.
,
30
,
247
262
, doi:.
Dong
,
L.
, and
T.
Zhou
,
2014
:
The formation of the recent cooling in the eastern tropical Pacific Ocean and the associated climate impacts: A competition of global warming, IPO, and AMO
.
J. Geophys. Res. Atmos.
,
119
,
11 272
11 287
, doi:.
Drbohlav
,
H.
, and
B.
Wang
,
2005
:
Mechanism of the northward-propagating intraseasonal oscillation: Insights from a zonally symmetric model
.
J. Climate
,
18
,
952
972
, doi:.
Duchon
,
C. E.
,
1979
:
Lanczos filtering in one and two dimensions
.
J. Appl. Meteor.
,
18
,
1016
1022
, doi:.
Gill
,
A. E.
,
1980
:
Some simple solutions for heat‐induced tropical circulation
.
Quart. J. Roy. Meteor. Soc.
,
106
,
447
462
, doi:.
Han
,
W.
, and Coauthors
,
2014
:
Intensification of decadal and multi-decadal sea level variability in the western tropical Pacific during recent decades
.
Climate Dyn.
,
43
,
1357
1379
, doi:.
Huang
,
R. H.
,
Y. H.
Xu
, and
L. T.
Zhou
,
1999
:
The interdecadal variation of summer rainfall in China and the drought trend in North China (in Chinese)
.
Plateau Meteor.
,
18
,
465
475
.
Jiang
,
X.
,
T.
Li
, and
B.
Wang
,
2004
:
Structures and mechanisms of the northward propagating boreal summer intraseasonal oscillation
.
J. Climate
,
17
,
1022
1039
, doi:.
Kajikawa
,
Y.
, and
B.
Wang
,
2012
:
Interdecadal change of the South China Sea summer monsoon onset
.
J. Climate
,
25
,
3207
3218
, doi:.
Kemball-Cook
,
S.
, and
B.
Wang
,
2001
:
Equatorial waves and air–sea interaction in the boreal summer intraseasonal oscillation
.
J. Climate
,
14
,
2923
2942
, doi:.
Kikuchi
,
K.
, and
B.
Wang
,
2009
:
Global perspective of the quasi-biweekly oscillation
.
J. Climate
,
22
,
1340
1359
, doi:.
Kucharski
,
F.
, and Coauthors
,
2016
:
Atlantic forcing of Pacific decadal variability
.
Climate Dyn.
,
46
,
2337
2351
, doi:.
Kuo
,
H. L.
,
1974
:
Further studies of the parameterization of the influence of cumulus convection on large-scale flow
.
J. Atmos. Sci.
,
31
,
1232
1240
, doi:.
Lepage
,
Y.
,
1971
:
A combination of Wilcoxon’s and Ansari-Bradley’s statistics
.
Biometrika
,
58
,
213
217
, doi:.
Li
,
C.
,
T.
Li
,
D.
Gu
,
A.
Lin
, and
B.
Zheng
,
2015a
:
Relationship between summer rainfall anomalies and sub-seasonal oscillation intensity in the ChangJiang Valley in China
.
Dyn. Atmos. Oceans
,
70
,
12
29
, doi:.
Li
,
C.
,
T.
Li
,
A.
Lin
,
D.
Gu
, and
B.
Zheng
,
2015b
:
Relationship between summer rainfall anomalies and sub-seasonal oscillations in South China
.
Climate Dyn.
,
44
,
423
439
, doi:.
Li
,
H.
,
A.
Dai
,
T.
Zhou
, and
J.
Lu
,
2010
:
Responses of East Asian summer monsoon to historical SST and atmospheric forcing during 1950–2000
.
Climate Dyn.
,
34
,
501
514
, doi:.
Li
,
R. C. Y.
, and
W.
Zhou
,
2013
:
Modulation of western North Pacific tropical cyclone activity by the ISO. Part I: Genesis and intensity
.
J. Climate
,
26
,
2904
2918
, doi:.
Li
,
T.
, and
B.
Wang
,
2005
:
A review on the western North Pacific monsoon: Synoptic-to-interannual variabilities
.
Terr. Atmos. Oceanic Sci.
,
16
,
285
314
, doi:.
Li
,
X.
,
S.-P.
Xie
,
S. T.
Gille
, and
C.
Yoo
,
2015c
:
Atlantic-induced pan-tropical climate change over the past three decades
.
Nat. Climate Change
,
6
,
275
279
, doi:.
Liebmann
,
B.
, and
C. A.
Smith
,
1996
:
Description of a complete (interpolated) outgoing longwave radiation dataset
.
Bull. Amer. Meteor. Soc.
,
77
,
1275
1277
.
Lin
,
A.
, and
T.
Li
,
2008
:
Energy spectrum characteristics of boreal summer intraseasonal oscillations: Climatology and variations during the ENSO developing and decaying phases
.
J. Climate
,
21
,
6304
6320
, doi:.
Liu
,
Y.
,
G.
Huang
, and
R.
Huang
,
2011
:
Inter-decadal variability of summer rainfall in Eastern China detected by the Lepage test
.
Theor. Appl. Climatol.
,
106
,
481
488
, doi:.
Luo
,
J.-J.
,
W.
Sasaki
, and
Y.
Masumoto
,
2012
:
Indian Ocean warming modulates Pacific climate change
.
Proc. Natl. Acad. Sci. USA
,
109
,
18 701
18 706
, doi:.
Mao
,
J.
, and
J. C. L.
Chan
,
2005
:
Intraseasonal variability of the South China Sea summer monsoon
.
J. Climate
,
18
,
2388
2402
, doi:.
McGregor
,
S.
,
A.
Timmermann
,
M. F.
Stuecker
,
M. H.
England
,
M.
Merrifield
,
F.-F.
Jin
, and
Y.
Chikamoto
,
2014
:
Recent Walker circulation strengthening and Pacific cooling amplified by Atlantic warming
.
Nat. Climate Change
,
4
,
888
892
, doi:.
Qian
,
C.
, and
T.
Zhou
,
2014
:
Multidecadal variability of North China aridity and its relationship to PDO during 1900–2010
.
J. Climate
,
27
,
1210
1222
, doi:.
Ren
,
X.
,
X.
Yang
, and
X.
Sun
,
2013
:
Zonal oscillation of western Pacific subtropical high and subseasonal SST variations during Yangtze persistent heavy rainfall events
.
J. Climate
,
26
,
8929
8946
, doi:.
Smith
,
T. M.
,
R. W.
Reynolds
,
T. C.
Peterson
, and
J.
Lawrimore
,
2008
:
Improvements to NOAA’s historical merged land–ocean surface temperature analysis (1880–2006)
.
J. Climate
,
21
,
2283
2296
, doi:.
Tao
,
L.
,
J.
Zhao
, and
T.
Li
,
2015
:
Trend analysis of tropical intraseasonal oscillations in the summer and winter during 1982–2009
.
Int. J. Climatol.
,
35
,
3969
3978
, doi:.
Teng
,
H.
, and
B.
Wang
,
2003
:
Interannual variations of the boreal summer intraseasonal oscillation in the Asian–Pacific region
.
J. Climate
,
16
,
3572
3584
, doi:.
Wang
,
B.
,
1988
:
Dynamics of tropical low-frequency waves: An analysis of the moist Kelvin wave
.
J. Atmos. Sci.
,
45
,
2051
2065
, doi:.
Wang
,
B.
, and
T.
Li
,
1994
:
Convective interaction with boundary-layer dynamics in the development of a tropical intraseasonal system
.
J. Atmos. Sci.
,
51
,
1386
1400
, doi:.
Wang
,
B.
, and
X.
Xie
,
1996
:
Low-frequency equatorial waves in vertically sheared zonal flow. Part I: Stable waves
.
J. Atmos. Sci.
,
53
,
449
467
, doi:.
Wang
,
B.
, and
X.
Xie
,
1997
:
A model for the boreal summer intraseasonal oscillation
.
J. Atmos. Sci.
,
54
,
72
86
, doi:.
Wang
,
B.
,
J.
Liu
,
H.-J.
Kim
,
P. J.
Webster
, and
S.-Y.
Yim
,
2012
:
Recent change of the global monsoon precipitation (1979–2008)
.
Climate Dyn.
,
39
,
1123
1135
, doi:.
Wang
,
M.
, and
A.
Duan
,
2015
:
Quasi-biweekly oscillation over the Tibetan Plateau and its link with the Asian summer monsoon
.
J. Climate
,
28
,
4921
4940
, doi:.
Wu
,
R.
, and
X.
Cao
,
2016
:
Relationship of boreal summer 10–20-day and 30–60-day intraseasonal oscillation intensity over the tropical western North Pacific to tropical Indo-Pacific SST
.
Climate Dyn.
,
48
,
3529
3546
, doi:.
Xiang
,
B.
, and
B.
Wang
,
2013
:
Mechanisms for the advanced Asian summer monsoon onset since the mid-to-late 1990s
.
J. Climate
,
26
,
1993
2009
, doi:.
Xie
,
P.
, and
P. A.
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
, doi:.
Xie
,
X.
, and
B.
Wang
,
1996
:
Low-frequency equatorial waves in vertically sheared zonal flow. Part II: Unstable waves
.
J. Atmos. Sci.
,
53
,
3589
3605
, doi:.
Xu
,
Z.
,
K.
Fan
, and
H.
Wang
,
2015
:
Decadal variation of summer precipitation over China and associated atmospheric circulation after the late 1990s
.
J. Climate
,
28
,
4086
4106
, doi:.
Yang
,
J.
,
B.
Wang
,
B.
Wang
, and
Q.
Bao
,
2010
:
Biweekly and 21–30-day variations of the subtropical summer monsoon rainfall over the lower reach of the Yangtze River basin
.
J. Climate
,
23
,
1146
1159
, doi:.
Yu
,
L.
,
T.
Furevik
,
O. H.
Otterå
, and
Y.
Gao
,
2015
:
Modulation of the Pacific Decadal Oscillation on the summer precipitation over East China: A comparison of observations to 600-years control run of Bergen Climate Model
.
Climate Dyn.
,
44
,
475
494
, doi:.
Zhou
,
W.
, and
J. C.
Chan
,
2005
:
Intraseasonal oscillations and the South China Sea summer monsoon onset
.
Int. J. Climatol.
,
25
,
1585
1609
, doi:.
Zhu
,
Y.
,
H.
Wang
,
W.
Zhou
, and
J.
Ma
,
2011
:
Recent changes in the summer precipitation pattern in East China and the background circulation
.
Climate Dyn.
,
36
,
1463
1473
, doi:.
Zhu
,
Y.
,
H.
Wang
,
J.
Ma
,
T.
Wang
, and
J.
Sun
,
2015
:
Contribution of the phase transition of Pacific Decadal Oscillation to the late 1990s’ shift in East China summer rainfall
.
J. Geophys. Res. Atmos.
,
120
,
8817
8827
, doi:.
Zveryaev
,
I.
,
2002
:
Interdecadal changes in the zonal wind and the intensity of intraseasonal oscillations during boreal summer Asian monsoon
.
Tellus
,
54A
,
288
298
, doi:.

Footnotes

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).