Qi Tong , Ynyn Hung , , , Mingkeng Dun , Qinfei Zho

a Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science & Technology, Nanjing, China

b Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai, China

c Dalian Meteorological Bureau of Liaoning Province, Dalian, China

Keywords:Western pacific subtropical high Decadal climate variation Pacific decadal oscillation

ABSTRACT Previous studies have demonstrated that the western Pacific subtropical high (WPSH) has experienced an eastward retreat since the late 1970s. In this study, the authors propose that this eastward retreat of the WPSH can be partly attributed to atmospheric responses to the positive phase of the Pacific decadal oscillation (PDO), based on idealized SST forcing experiments using the Community Atmosphere Model, version 4. Associated with the positive phase of the PDO, convective heating from the Indian Peninsula to the western Pacific and over the eastern tropical Pacific has increased, which has subsequently forced a Gill-type response to modulate the WPSH.The resulting cyclonic gyre over the Asian continent and the western Pacific in the lower troposphere is favorable for the eastward retreat of the WPSH. Additionally, the resulting anticyclonic gyre in the upper troposphere is favorable for the strengthening and southward expansion of the East Asian westerly jet, which can modulate the jet-related secondary meridional-vertical circulation over the western Pacific and promote the eastward retreat of the WPSH.

1. Introduction

The western Pacific subtropical high (WPSH) is a synoptic-scale anticyclonic circulation over the Pacific, which usually reaches its peak intensity in the summer. The WPSH is one of the most important components of the East Asian summer monsoon ( Lau and Li, 1984 ; Tao and Chen, 1987 ). Anomalies of the WPSH are significantly associated with anomalies of summer precipitation and temperature over East Asia( Tao and Xu, 1962 ; Zhou and Wang, 2006 ), and can even affect tropical cyclone occurrence frequency over the western Pacific ( Chia and Ropelewski, 2002 ; Zhou and Cui, 2008 ).

In recent decades, the decadal variation of the WPSH has attracted considerable attention amongst scientists. Based upon measurement of the decadal variation in the location of the 5880/5870 gpm contour line in the geopotential height field at 500 hPa, most previous studies have documented that the WPSH has experienced a westward extension since the late 1970s ( Nitta and Hu, 1996 ; Hu, 1997 ). This westward-extended WPSH has been suggested to contribute to the decadal variations of the temperature and precipitation over East Asia ( Nitta and Hu, 1996 ;Hu, 1997 ; Gong and Ho, 2002 ) and the typhoon tracks over the western Pacific ( Ho et al., 2004 ) since the late 1970s. The Indian Ocean-western Pacific warming has been suggested to have caused this westward extension of the WPSH ( He and Gong, 2002 ; Zhou et al., 2009 ).

Fig. 1. (a) Doubled JJA SST difference between 1979 and 2000 and 1950-1978 in the Hadley SST dataset, which is added to the climatological cycle of SST from May to September and used as the boundary forcing in EXP-PDO. (b, c) Climatological JJA eddy geopotential height at 850 (units: gpm) in NCEP during 1950-2000 and EXP-CTL, respectively. (d, e) Characteristic WPSH isolines of H’850 from two periods in NCEP (0-gpm) and two experiments in CAM (15-gpm), respectively.Solid lines (dashed lines) represent the period 1950-78 (1979-2000) and EXP-CTL (EXP-PDO).

However, it has been found that global warming has certain impacts on the decadal variation of the WPSH as quantified by changes in the geopotential height field ( Yang and Sun, 2003 ; Lu et al., 2008 ).Our studies have documented that the WPSH has retreated eastwards since the late 1970s, with a weakened anticyclone at 850 hPa, decreased sea level pressure, increased relative vorticity in the middle and lower troposphere, and an eastward-shifted maximum meridional wind over the western Pacific ( Huang et al., 2015 ). It has been proposed that the eddy geopotential height should be implemented to investigate or project the decadal WPSH variation mainly because the artificial signals caused by global warming can be avoided ( Huang and Li, 2015 ; Huang et al., 2016 ). The eastward retreat of the WPSH has been found to contribute to the south-flood-north-drought pattern over China ( Huang et al., 2015 ). Nevertheless, the mechanism related to the eastward retreat of the WPSH remains unsolved.

The Pacific decadal oscillation (PDO) is the dominant mode of decadal sea surface temperature (SST) variability over the North Pacific ( Mantua et al., 1997 ). It is not only connected to other components of the climate system, but also affects natural resources and both marine and terrestrial ecosystems ( Mantua et al., 1997 ; Newman et al.,2016 ). An abrupt change in the PDO occurred in the late 1970s, and many other decadal climate changes, such as the ENSO variability( An and Wang, 2000 ; Vimont et al., 2003 ), ENSO-monsoon relationship ( Chan and Zhou, 2005 ; Wang et al., 2008 ; Xie et al., 2010 ; Fan and Fan, 2017 ), East Asian summer monsoon ( Zhu and Yang, 2003 ), rainfall in Australia ( Arblaste et al., 2002 ), and the climate of North America( McCabe et al., 2004 , 2012 ), have been attributed to this phase change in the PDO. Considering the significant influence of the PDO on the decadal variation of regional climate, this study investigates the possible contributions of the PDO to the decadal eastward retreat of the WPSH.

Fig. 2. Differences in (a) velocity potential (units: 10 6 m 2 s ? 1 ) and the divergent component of wind vectors (units: m s ? 1 ) and (b) JJA precipitation (units: mm d ? 1 )between EXP-PDO and EXP-CTL. The shaded regions in (a) and dotted regions in (b) are for values significant at or above the 0.01 level.

2. Data and methods

This study focuses on the summer season (June-July-August, JJA).Two datasets are used: geopotential height data from the NCEP-NCAR(hereafter simply “NCEP ”) reanalysis dataset ( Kalnay et al., 1996 ); and SST data from the Met-Office Hadley centre ( Rayner et al., 2003 ).

Compared to the SST during 1950-1978, positive SST anomalies(SSTAs) occurred over the eastern North Pacific and eastern tropical Pacific, while negative SSTAs occurred over the central and western North Pacific during 1979-2000. Such an SSTA pattern indicates that the PDO changed from a negative to positive phase after the late 1970s( Fig. 1 (a)). In this study, we investigate the impacts of the positive PDO phase on summer climate. Two experiments were conducted using the Community Atmosphere Model, version 4 (CAM4). In the control experiment (EXP-CTL), the climatological annual cycle of SST released by Hurrell et al. (2008) was used as the boundary forcing. In the sensitivity experiment (EXP-PDO), the doubled JJA SST difference between 1979 and 2000 and 1950-1978 over the Pacific (60°S-60°N, 120°E-60°W)( Fig. 1 (a)) was added to the climatological cycle of SST from May to September and used as the boundary forcing. Both experiments were integrated for 60 years. Here, we analyze the JJA outputs of the last 50 years, with the first 10 years taken as the spin-up time for the model to reach a relative equilibrium state. The difference between EXP-PDO and EXP-CTL is used to investigate regional climate responses to the positive PDO. Statistical significance was tested using the Student’s

t

-test.The resolution of the CAM4 is 2.5°(longitude) ×1.9°(latitude), and there are 26 hybrid levels in the vertical direction.Following Huang et al. (2015) , the WPSH is measured by the eddy geopotential height at 850 hPa (

H

’850), which is calculated by subtracting the zonal-mean 850-hPa geopotential height between 0° and 40°N from the original 850-hPa geopotential height. The WPSH reflected by positive

H

’850 over the Pacific in the reanalysis dataset ( Fig. 1 (b)) can be successfully reproduced by CAM4 in EXP-CTL, but is located more westward and northward ( Fig. 1 (c)). To address this issue, we use the 0-gpm isoline in the reanalysis and 15-gpm isoline in the CAM4 experiments as the characteristic WPSH isoline.

3. Results

As discussed in Huang et al. (2015) and Huang and Li (2015) , the WPSH has retreated eastwards since the late 1970s, with the mean position of the western edge located at 137.5°E (140°E) during 1950-78(1979-2000) ( Fig. 1 (d)). The mean position of the western edge of the WPSH is located at 137.5°E in EXP-CTL, but shifts eastwards to 145°E in EXP-PDO ( Fig. 1 (e)). These results indicate that the positive PDO can lead to an eastward retreat of the WPSH. The possible mechanism is discussed in the following sections.

Fig. 2 shows the anomalies of the velocity potential and divergent wind component at 200 hPa and the response of summer rainfall to the positive PDO. Generally, the low-level convective activity over the eastern tropical Pacific warmer water (central Pacific cooler water) has increased (reduced), as seen by the local anomalous upper-tropospheric divergence (convergence) ( Fig. 2 (a)). Correspondingly, the rainfall has decreased over the central Pacific, but increased over the eastern tropical Pacific ( Fig. 2 (b)). Zhou et al. (2009) demonstrated that negative rainfall anomalies over the central and eastern tropical Pacific are crucial for the westward extension of the WPSH according to the Gill-type response ( Gill, 1980 ). Here, increased precipitation over the eastern tropical Pacific may influence the WPSH in the opposite way. It is worth noting that the regions of anomalous upper-tropospheric divergence are not only located over the western Pacific warmer water, but also extend westwards to the Indian peninsula, indicating significant above-normal convective activity over that region ( Fig. 2 (a)). Therefore, the largest positive precipitation anomalies are located over the western Pacific-Indian peninsula ( Fig. 2 (b)). The streamfunction anomalies show a vigorous cyclonic couplet straddling this large positive rainfall anomaly region ( Fig. 3 (a)). The positive precipitation anomalies correspond to positive heat sources due to greater latent heat release. The streamfunction responses are consistent with the theory of Gill (1980) . Associated with the cyclonic streamfunction anomalies, there is an anomalous low occupying the Asian continent and western Pacific ( Fig. 3 (b)). Since negative rainfall anomalies cannot prevail over the Pacific, a slightly anomalous high only appears over the central Pacific ( Fig. 3 (b)). Generally, the Gill-type response to positive rainfall anomalies over the Indian peninsula-western Pacific and the eastern tropical Pacific are favorable for the eastward retreat of the WPSH.

Fig. 3. (a) Differences in horizontal streamfunction (contours; units: 10 6 m 2 s ? 1 ) and the non-divergent wind component (vectors; units: m s ? 1 ) at 850 hPa between EXP-PDO and EXP-CTL. The shaded regions represent values significant at or above the 0.01 level. (b) As in (a) but for sea level pressure (contours; units: Pa) and horizontal wind at 850 hPa (vectors; units: m s ? 1 ).

In the upper levels, the streamfunction anomalies show two strong anticyclonic couplets straddling the southern Indian Ocean-Maritime Continent and the eastern tropical Pacific, respectively ( Fig. 4 (a)). The streamfunction responses in the upper troposphere are generally opposite to those in the lower troposphere, which is consistent with the vertical structure of Gill’s model ( Gill, 1980 ; Sutton and Hodson, 2007 ).Therefore, these two anticyclonic couplets primarily derive from the positive heating sources forced by increased convective activities in the Indian peninsula-western Pacific and the eastern tropical Pacific,respectively. Associated with the strong anticyclonic streamfunction anomalies over the Asian continent and the western Pacific, zonal winds in the midlatitudes (30°-40°N) of the Asian continent and the western Pacific have increased. Correspondingly, the East Asian westerly jet (EAWJ) axis has moved southwards in EXP-PDO ( Fig. 4 (b)). The response of the EAWJ to the positive PDO is consistent with previous studies ( Yu and Zhou, 2007 ; Zhang and Huang, 2011 ; Lu et al., 2013 ). The changes in the EAWJ can influence summer climate over the surrounding areas by modulating the jet-related secondary meridional-vertical circulation ( Ding et al., 2008 ; Zhu et al., 2015 ). Associated with the strengthened and southward-shifted EAWJ, anomalous ascending motion develops to its south and descending motion develops to its north over the western Pacific ( Fig. 4 (c)). Although the descent anomalies are not remarkable, the significant ascent anomalies occupy south of 35°N over the western Pacific, which corresponds to the anomalous cyclone at 850 hPa over the western Pacific shown in Fig. 3 (b). Generally,the response of the EAWJ to the positive PDO modulates the secondary meridional-vertical circulation and further promotes the eastward retreat of the WPSH.

4. Summary and discussion

Fig. 4. Difference in (a) horizontal streamfunction (contours; units: 10 6 m 2 s ? 1 ) and the non-divergent wind component (vectors; units: m s ? 1 ) and (b) zonal wind at 200 hPa between EXP-PDO and EXP-CTL. Shaded regions indicate statistical significance at or above the 0.01 level. The red (blue) line in (b) represents the westerly jet axis in EXP-CTL (EXP-PDO), which is defined as the location of mean maximum zonal wind speed at each longitude. (c) Differences in meridional-vertical circulation between 120°? 150°E (omega: 10 ? 2 Pa s ? 1 ; meridional wind: m s ? 1 ). Shaded regions are values significant at or above the 0.01 level.

Previous studies have demonstrated that the WPSH has retreated eastwards since the late 1970s, which has subsequently caused decadal variations of summer climate in Asia. In this study, we investigate the contribution of the PDO to the eastward retreat of the WPSH by using an idealized SST-forcing CAM4 approach. The results of numerical experiments demonstrate that the positive PDO mode can lead to the eastward retreat of the WPSH. Particularly, associated with the positive PDO mode, the positive heating in the Indian peninsula-western Pacific region forces a Gill-type response that modulates the WPSH. The resulting cyclonic gyre over the Asian continent and the western Pacific in the lower troposphere is favorable for the eastward retreat of the WPSH.On the other hand, the resulting anticyclonic gyre in the upper troposphere over the Asian continent and the western Pacific is favorable for a strengthening and southward shift of the EAWJ, which can lead to anomalous ascending motion to its south and anomalous descending motion to its north, and further promotes the eastward retreat of the WPSH. The positive heating anomalies in the eastern tropical Pacific also play an important role for the eastward retreat of the WPSH. They weaken the atmospheric response to negative heating over the central Pacific and meanwhile force a Gill-like response that further weakens the subtropical high over the Pacific.

Zhou et al. (2009) demonstrated that Indian Ocean warming can lead to a westward extension of the WPSH, based on numerical studies of five AGCMs. The eastward retreat of the WPSH seems to contradict the atmospheric responses for the Indian Ocean warming. Actually, the Indian Ocean may stay in a recharging phase on the decadal scale ( Li et al.,2008 ). The PDO may be the more visible contributor to the eastward retreat of the WPSH. Actually, decadal climate change is contributed by multiple factors, including global warming and decadal variability,such as the PDO. Based on previous studies about the response of the subtropical high to global warming, the center of the North Pacific subtropical high (NPSH) may be weakened, and the western portion of the NPSH (i.e., the WPSH) will also be weakened in the mid troposphere( He et al., 2015 , 2017 ) but changes ambiguously in the lower troposphere ( He and Zhou, 2015 ; Huang et al., 2016 ). But how is the decadal change in the WPSH presented in our study related to ongoing global warming? Is the positive PDO-related change in the WPSH similar to the global warming-related change in the WPSH? Meanwhile, there is an interaction between the atmosphere and the Indian Ocean and western Pacific. This study only addresses how the atmosphere responds to SSTAs. More objective experiments (such as using a coupled climate model) should be carried out in the future.

Funding

This study was funded by the Major Program of the National Natural Science Foundation of China [grant number 41991283 ], the National Key Research and Development Program of China [grant number 2016YFA0600703 ], and the Funding of the Jiangsu Innovation &Entrepreneurship Team, the Priority Academic Program Development(PAPD) of Jiangsu Higher Education Institutions .