Vertical structure of variabilities in the tropical easterly jet and associated factors

Ye Yo ,Yunyun Guo,b,＊ ,Zhiping Wen,c,d ,Sihu Hung

a Department of Atmospheric and Oceanic Sciences &Institute of Atmospheric Sciences, Fudan University, Shanghai, China

b CMA-FDU Joint Laboratory of Marine Meteorology, Shanghai, China

c Shanghai Key Laboratory of Ocean-land-atmosphere Boundary Dynamics and Climate Change, Fudan University, Shanghai, China

d Jiangsu Collaborative Innovation Center for Climate Change, Nanjing, China

Keywords: Tropical easterly jet Vertical structure Interannual variability Interdecadal variability ENSO Quasi-biennial oscillation

ABSTRACT The tropical easterly jet (TEJ) is an easterly jet stream that occurs from the upper troposphere to lower stratosphere in boreal summer.Owing to its wide vertical extension from 300 to 70 hPa,the TEJ may exhibit distinct characteristics at different levels,the details of which remain thus far unclear.In this study,two empirical orthogonal function (EOF) modes of the year-to-year variability in the vertical structure of the TEJ were investigated.The leading EOF mode represents a consistent strengthening or weakening of the TEJ’s main body in the vertical direction and varies on both interannual and interdecadal time scales.It has been suggested that ENSO can modulate this vertically consistent mode interannually,whereas the Atlantic Multidecadal Oscillation and Pacific Decadal Oscillation can influence its interdecadal variability.In contrast,the second EOF mode exhibits an out-of-phase relationship between the zonal wind anomalies of the upper troposphere and lower stratosphere,linked with the changes in the TEJ’s vertical movement and dominating on the quasi-biennial time scale.The Quasi-Biennial Oscillation could contribute to variations in the TEJ’s vertical movement by changing the tropopause winds as a direct pathway and inducing anomalous convection over the tropical Indian Ocean and Maritime Continent as an indirect pathway.

1.Introduction

The tropical easterly jet (TEJ),a strong easterly belt in boreal summer,is characterized by a wide horizontal extent from the tropical western Pacific to Africa(Fig.1(a))and a broad vertical spread from 300 to 70 hPa (Fig.1(b)),i.e.,covering the upper troposphere and lower stratosphere(UTLS).Many studies have used the zonal wind anomaly at a single level (e.g.,200,150,or 100 hPa) as a proxy of the TEJ’s variability(Rao et al.,2004;Sathiyamoorthy et al.,2007;Naidu et al.,2011;Abish et al.,2013;Huang et al.,2021).However,the characteristics of the TEJ shown in these previous studies may largely depend on the specific level considered (Chen et al.,2007;Sreekala et al.,2014).For instance,the weakening trend of the TEJ core is more significant at 100 hPa than other levels (Sathiyamoorthy,2005).Chen et al.(2007)proposed that the TEJ between 100 and 150 hPa experienced a significant interdecadal shift around the late 1970s,whereas its counterpart at 200 hPa experienced only a faint interdecadal variation.These results suggest that the variabilities of the TEJ may differ vertically,but there is thus far little understanding in this respect.Therefore,studying the TEJ’s vertical profile in summer will help build a more comprehensive understanding of the TEJ’s variation and monsoon dynamics.

Many studies have found that the TEJ varies greatly both on interannual and interdecadal time scales (Abish et al.,2013;Ratnam et al.,2013;Nithya et al.,2017;Huang et al.,2019,2021),and considerable attention has been paid to the factors that contribute to these variations.For instance,it has been demonstrated that changes in the tropical sea surface temperature(SST)and stratospheric processes can modulate the TEJ’s strength interannually(Bi et al.,2013;Nithya et al.,2017;Rai and Dimri,2017;Li et al.,2022).El Ni?o–Southern Oscillation (ENSO),perceived as the most important tropical air–sea coupling phenomenon on the interannual time scale,could significantly change the TEJ’s strength via exciting a typical Gill response and changing the Walker circulation(Arkin,1982;Chen and van Loon,1987;Nithya et al.,2017;Huang et al.,2019;Hrudya et al.,2021).

Moreover,the Quasi-Biennial Oscillation (QBO) has been suggested to have a direct effect on the tropical UTLS,e.g.,the tropical tropospheric winds(Gray et al.,1992;Hitchman and Huesmann,2009).Meanwhile,Li et al.(2022) also proposed an indirect pathway that connects the QBO and the TEJ’s intensity in summer;that is,the easterly phase of the QBO favors strong convective activity over the west side of the Maritime Continent,which can significantly weaken the TEJ in situ.On the decadal time scale,Li et al.(2017)suggested that the warm phase of the Atlantic Multidecadal Oscillation (AMO) could result in the intensification of the TEJ.Given the limited understanding regarding the TEJ’s vertical structure,whether and how these factors modulate the TEJ’s vertical distribution remains unclear.Thus,the present study assessed the spatiotemporal characteristics of the TEJ’s vertical structure in summer and examined its associated factors.

The rest of this paper is structured as follows.Section 2 describes the data and methods.The variability of the TEJ’s vertical structure and its plausible causes are presented in Section 3.Section 4 gives a summary.

2.Data and methods

The following observational and reanalysis datasets were used in this study:(1)monthly atmospheric circulation and precipitation data from ERA5 (Hersbach et al.,2020),with a horizontal resolution of 2.5?×2.5?;and (2) monthly SST data from HadISST (Rayner et al.,2003),with a horizontal resolution of 1?× 1?.

The AMO index was obtained from the SST anomaly(SSTA)averaged in the North Atlantic(0?–60?N,0?–80?W)(for details,refer to Trenberth and Shea (2006)).The Pacific Decadal Oscillation (PDO) index was defined as the principal component corresponding to the leading empirical orthogonal function(EOF)mode of SSTA in the North Pacific(20?–70?N,110?E–100?W) (Zhang et al.,1997).The QBO index was defined as the normalized zonal wind at 50 hPa observed from three radiosonde stations (Canton Island,Gan/Maledive Islands,and Singapore) provided by Freie Universit?t,Berlin.

This study focused on the boreal summer (June–July–August,JJA)during 1959–2021.All variables were detrended,and the Butterworth filter with a 9-yr high (low)pass was applied to obtain the interannual(interdecadal) component.The Monte-Carlo test was used to estimate statistical significance.

3.Two distinct modes of the TEJ’s vertical distribution

The pressure–longitude profile of the climatological TEJ averaged over 5?S–20?N features a broad extent ranging from 300 to 70 hPa,with the core near 150 hPa (Fig.1(b)).Its standard deviation exceeding 1.0 basically locates above 400 hPa,which reaches its maximum at 150 hPa(Fig.1(b)).Interestingly,the zonal wind anomalies at 200 hPa are commonly used to measure the TEJ’s variability(Sathiyamoorthy et al.,2007;Huang et al.,2019,2020),whereas our result indicates that the most significant year-to-year variation of the TEJ is observed at 150 hPa.

3.1.Leading EOF mode associated with the TEJ’s intensity

To depict the spatiotemporal variation of the TEJ’s vertical structure,EOF analysis was applied to the normalized zonal wind anomalies from 300 to 70 hPa,averaged along 5?S–20?N.The first two modes are insensitive to the chosen analyzed domains.Furthermore,the results based on log-pressure weighting fields and the original ones are quite similar.

The first EOF mode (EOF1),accounting for 46.22% of the total variance,exhibits large negative values west of 130?E and at 300–70 hPa,with its minimum in tropical Africa(near 20?E)at 150 hPa(Fig.1(c)).Bonding to a positive principal component(PC1),the main body of the TEJ tends to be strengthened from 300 to 70 hPa,and its core shifts slightly westward.Note that the westerly anomaly east of 130?E may weaken the TEJ over its inflow region owing to weak positive values observed in EOF1(Fig.1(c)).Based on the abovementioned features,we argue that EOF1 primarily represents a consistent intensification or weakening of the TEJ’s main body in the vertical direction.The strong relation between PC1 and the TEJ intensity index suggested in previous research (Chen et al.,2007;Sreekala et al.,2014;Huang et al.,2019)provides further evidence for it as their correlation coefficient reaches up to 0.8,significant at the 0.05 level(Table 1).Also,our results suggest that the use in previous studies of a single level to measure the TEJ was reasonable if their focus was on changes in the TEJ’s intensity.Furthermore,EOF1 exhibits both interannual and interdecadal variabilities.The interannual component of PC1 makes up 72%of the total variance,which is estimated by the variance ratio of the interannual component to the total.Similarly,the interdecadal component occupies 23%.The 13-yr running averaged PC1 also implies the phase of EOF1 shifted in the early-1970s and mid-1990s (Fig.1(d)).Therefore,the interannual and interdecadal variations of EOF1 are separately examined in the following subsections,and the corresponding PC1 is marked as PC1_iav and PC1_idv,respectively.

3.2.Interannual component

A large-scale,both horizontally and vertically,easterly anomaly associated with PC1_iav locates between 300 and 70 hPa and 60?W–120?E (Fig.2(a,b)).Its westernmost point,represented by the composite isotach of -20 m s-1(-10 m s-1),has a longitudinal difference of 15?(27.5?) between the strong and weak TEJ years.By contrast,the easternmost point of the TEJ’s main body shows relatively subtle differences owing to weak eigenvector loads east of 130?E(Fig.1(c)).The TEJ’s upper and lower boundary extends accordingly in strong TEJ years(Fig.2(a)).

Given the high correlation (R=-0.72) between PC1_iav and the Ni?o3.4 index in summer(Fig.1(d)),the ENSO-related SSTA may have an effect on the interannual variations of EOF1.The regression map of the SSTA onto PC1_iav shows a canonical La Ni?a-like pattern over the tropical Pacific(Fig.2(d)).The La Ni?a-like SSTA can have an influence on the intensification of the TEJ through several physical processes.Firstly,it can induce an anomalous Walker circulation with the ascending motion anomaly over the South Asian summer monsoon region and Maritime Continent and the descending motion anomaly mainly lying over the tropical western-central Pacific (Chen and van Loon,1987;Hrudya et al.,2021),which can also be observed in Fig.2(c).This anomalous Walker circulation indicates an intensification of the tropical divergent circulations over southern Asia and Africa in the upper troposphere,which can modulate the local kinetic energy flux and then accelerate the western part of the TEJ(Chen and van Loon,1987).Secondly,the suppressed precipitation over the tropical eastern Pacific induced by the La Ni?a-like SSTA can trigger a typical Gill response(Gill,1980) with a pair of upper-level cyclones to its west that are equatorially symmetric,and a zonally elongated Kelvin wave response to its east.This easterly response in the upper troposphere prevails over the Atlantic Ocean and east towards the western Indian Ocean(Fig.2(b,c)),which could enhance the main body of TEJ.Furthermore,a significantly positive geopotential height anomaly across the subtropical Afro-Eurasia region,which has been proven to be related to La Ni?a events (Seager et al.,2003;Shaman and Tziperman,2007),could increase the poleward pressure gradient over North Africa and South Asia and thus the TEJ’s intensity.

3.3.Interdecadal component

EOF1 also exhibits significant interdecadal variation,as the 13-yr running average of PC1 shifts from a positive to negative phase in the early-1970s and then returns back around the mid-1990s(Fig.1(d)).The vertical structure of EOF1 at the interdecadal time scale (Fig.3(a)) is similar to that at the interannual time scale,except for a relatively weaker easterly anomaly above 100 hPa and a weaker and westwardshifted westerly anomaly east of 90?E above 400 hPa.We note that the EOF1-associated atmospheric circulations at the interdecadal and interannual time scales are distinct.Alternative centers of the associated 150 hPa geopotential height anomalies appear over the northern Atlantic and midlatitudes of Eurasia (Fig.3(b)),resembling the Silk Road pattern (Lu et al.,2002;Wang et al.,2017).The associated wave activity flux (Takaya and Nakamura,2001) shows a significant northward propagation of the Rossby wave energy from the tropical Atlantic Ocean and then reflects eastward near the Mediterranean Sea(Fig.3(c)).A strong easterly wind anomaly over the tropical Atlantic Ocean and Africa,linked with the increased poleward pressure gradient induced by the positive centers of action embedded in this Silk Road–like teleconnection pattern,may result in the TEJ’s intensification.

The PC1_idv-related SSTA shows significant warming anomalies in the North Atlantic and a PDO-like distribution in North Pacific,which motivates us to examine the effects of the AMO and PDO.We find that the PC1_idv–AMO (PC1_idv–PDO) correlation coefficient is 0.65(-0.49),and the regression map of anomalous geopotential height and wind anomalies against the AMO (PDO) index (Fig.3(d) (Fig.3(e)))shows high similarity to that against PC1_idv (Fig.3(b)),mainly characterized by a Silk Road–like teleconnection.Meanwhile,a strong easterly anomaly spans from the tropical Atlantic Ocean to the western Indian Ocean,leading to a strengthening of the TEJ’s main body west of 60?E.Thus,one of the possible processes by which the AMO and PDO modulate EOF1 at the interdecadal scale is via changing the Silk Road–like teleconnection.Although the correlation coefficient between PC1_idv and the AMO is higher than that between PC1_idv and the PDO,the relative importance cannot easily be stated,which calls for further exploration.

It is also worth mentioning that significant cold SSTAs appear in the southwestern Indian Ocean (Fig.3(c)).Upon examination of the associated vertical velocity and surface heat flux anomalies,the SSTA over there may act in a passive role(not shown),which cannot actively affect the atmospheric circulation.

3.4.Second EOF mode revealing the TEJ’s vertical movement

In contrast to the leading mode characterized by a vertical consistency of zonal wind anomalies between 300 and 70 hPa,EOF2 shows a significant negative relationship between the upper troposphere and lower stratosphere(Fig.1(e)).A strong easterly anomaly belt lies firmly in tropics at 70 hPa and there is a significant westerly anomaly located in the upper troposphere over the Maritime Continent corresponding to a positive PC2,leading to the upward movement and extension of the TEJ with its core at 125 hPa (Fig.4(a)).Conversely,the TEJ would move downward with its upper boundary shrinking vertically below 80 hPa and its core moving down to 150 hPa due to the opposite wind anomalies corresponding to a negative PC2 (Fig.4(a)).Wavelet analysis of PC2 indicates a quasi-biennial variation (not shown),and the interannual component can explain 93% of the total variance.Thus,we argue that EOF2 captures the interannual variability of the TEJ’s vertical movement.The following analyses of EOF2 were extracted from the interannual component with a 9-yr high pass filter.

Fig.4.(a) Regression map of the JJAmean zonal wind anomalies (shading;units: m s–1) averaged over 5?S–20?N against PC2_iav.(b) Lead–lagged regression of zonal wind(5?S–20?N,30?–150?E)from JJA(-2) to JJA(1) against PC2_iav.(c,d)Regression maps of(c)precipitation(shading;units: mm) and (d) 150 hPa horizontal wind(vectors;units:m s–1)and the 150 hPa vertical gradient of potential temperature (?θ/?z;shading;units: K km–1) against PC2_iav.Solid (dashed)contours in (a) denote the total zonal wind by compositing the years with PC2 greater than 0.8 (smaller than -0.8).The green dots in (a) denote anomalies significant at the 0.05 level based on the Monte-Carlo test.The orange dashed lines in (a,b) separate the regions with different contour intervals.The stippling indicates anomalies significant at the 0.05 level,and only wind anomalies above the 95% confidence level are shown.The yellow frame in (a) denotes the chosen region for the EOF analysis in Fig.1.

Owing to the banded zonal wind anomalies in the tropics at 70 hPa and the quasi-biennial variation of EOF2,we speculate that the QBO may hold the key to EOF2.A significant correlation(R=-0.64)between the QBO index and PC2 (Fig.1(f)) provides statistical evidence for this conjecture.We further examined the lead–lag relationship between PC2 and the zonal wind in the UTLS averaged over the region of interest(5?S–20?N,30?–150?E) (Fig.4(b)).In the present JJA,an out-of-phase relationship between the zonal wind anomalies below and above 125 hPa is observed.The upper tropospheric westerly anomalies between 300 and 150 hPa in JJA can be traced back to the westerly phase of the QBO peaking in JJA(-2).The lower stratospheric easterly anomalies above 125 hPa in JJA originate from the easterly phase of the QBO peaking in DJF(-1).The alternating occurrence of the easterly and westerly QBO events with a prevailing 2-yr period partly determines the out-of-phase relationship of EOF2.Moreover,the QBO leads to a mean meridional circulation (Gray et al.,1992;Collimore et al.,2003) characterized by an out-of-phase temperature response between the tropical and subtropical UTLS(not shown),which can modulate the TEJ-related zonal wind via the thermal wind balance.

Li et al.(2022) proposed that the QBO can indirectly affect the intensity of the TEJ over the Maritime Continent by changing the upper-tropospheric stability near the Maritime Continent.The PC2-associated tropospheric stability(measured by the vertical gradient of potential temperature,?θ/?z) and convection show that the tropospheric stability decreases significantly over the tropical Indian Ocean and Maritime Continent (Fig.4(f)).Since the climatological ascending motion lying on the west side of the Maritime Continent reaches a higher level than on the east side,the weakened tropospheric stability can induce convective activity west of the Maritime Continent more efficiently than that to its east (Fig.4(e)).The convective anomalies over there would exert the zonal wind anomaly over the Maritime Continent as a Kelvin wave response.Hence,there may be two pathways—a direct and indirect one—via which the QBO modulates the vertical movement of the TEJ.

4.Summary

The spatiotemporal characteristics of the TEJ’s vertical structure and the associated factors have been examined in this study.EOF1 features vertically consistent anomalies of zonal wind between 300 and 70 hPa,which oscillate on both interannual and interdecadal time scales.On the interannual time scale,ENSO has a crucial influence on the vertical change in the TEJ’s intensity via tropical and extratropical processes.On the interdecadal time scale,the AMO and PDO can together result in vertically consistent enhancement or weakening of the western TEJ by exciting a Silk Road–like teleconnection pattern.

EOF2,exhibiting an out-of-phase relationship between the zonal wind below and above 125 hPa,reveals the year-to-year variation of the TEJ’s vertical movement.We propose that the alternating westerly and easterly phases of the QBO can directly influence the TEJ at different pressure levels.Furthermore,another indirect mechanism involving changes in the tropospheric stability and convection over the tropical Indian Ocean and Maritime Continent was also verified.

Funding

This research was supported by the National Natural Science Foundation of China [grant numbers 42030601 and 42205019],and S.Huang was supported by a project funded by the China Postdoctoral Science Foundation [grant number 2022M710715].

Acknowledgments

We would like to acknowledge the inspiring discussion with Dr.Fei Xie,and we are also grateful for the valuable suggestions provided by Dr.Yanke Tan,Dr.Ruifen Zhan,Dr.Jiacan Yuan,and Dr.Xiaodan Chen during a group seminar.