Abstract

This study investigates modulation of El Niño–Southern Oscillation (ENSO) on the Madden–Julian oscillation (MJO) propagation during boreal winter. Results show that the spatiotemporal evolution of MJO manifests as a fast equatorially symmetric propagation from the Indian Ocean to the equatorial western Pacific (EWP) during El Niño, whereas the MJO during La Niña is very slow and tends to frequently “detour” via the southern Maritime Continent (MC). The westward group velocity of the MJO is also more significant during El Niño. Based on the dynamics-oriented diagnostics, it is found that, during El Niño, the much stronger leading suppressed convection over the EWP excites a significant front Walker cell, which further triggers a larger Kelvin wave easterly wind anomaly and premoistening and heating effects to the east. However, the equatorial Rossby wave to the west tends to decouple with the MJO convection. Both effects can result in fast MJO propagation. The opposite holds during La Niña. A column-integrated moisture budget analysis reveals that the sea surface temperature anomaly driving both the eastward and equatorward gradients of the low-frequency moisture anomaly during El Niño, as opposed to the westward and poleward gradients during La Niña, induces moist advection over the equatorial eastern MC–EWP region due to the intraseasonal wind anomaly and therefore enhances the zonal asymmetry of the moisture tendency, supporting fast propagation. The role of nonlinear advection by synoptic-scale Kelvin waves is also nonnegligible in distinguishing fast and slow MJO modes. This study emphasizes the crucial roles of dynamical wave feedback and moisture–convection feedback in modulating the MJO propagation by ENSO.

1. Introduction

The Madden–Julian oscillation (MJO) (Madden and Julian 1971, 1972) is the dominant mode of tropical intraseasonal variability (20–100 days) (Zhang 2005) and exerts significant impacts on the global weather and climate (Zhang 2013), including tropical cyclones (Maloney and Hartmann 2000), westerly wind events (Hao et al. 2019), convectively coupled equatorial waves (Roundy 2008; Kiladis et al. 2009), extreme rainfall (Ren and Ren 2017; Ren et al. 2018), monsoons (Lau and Chan 1986a; Lorenz and Hartmann 2006), and El Niño–Southern Oscillation (ENSO) (McPhaden 1999; Hendon et al. 2007). Over the past several decades, even though large progress has been frequently reported with respect to understanding the dynamics and physics of the MJO (e.g., Madden and Julian 1994; Zhang 2005; Lau and Waliser 2012; Li 2014; DeMott et al. 2015; Wang et al. 2016), most state-of-the-art general circulation models (GCMs) still struggle to satisfactorily simulate some fundamental characteristics of the MJO, especially its systematic eastward propagation from the Indian Ocean (IO) to the equatorial western Pacific (EWP) (e.g., Kim et al. 2009; Neena et al. 2014; Jiang et al. 2015; Ahn et al. 2017). This strongly suggests the necessity of advancing our understanding of the MJO propagation mechanism.

Of the numerous theories, three representative kinds have been widely used to explain the eastward propagation of the MJO. The first theory emphasizes the first-order importance of the tropical atmospheric humidity under the weak temperature gradient approximation (Sobel et al. 2001), that is, the “moisture mode” theory (Sobel and Maloney 2012, 2013; Adames and Kim 2016). Based on diagnostic results from GCMs (Maloney 2009; Andersen and Kuang 2012), reanalysis (Kiranmayi and Maloney 2011), and observations (Sobel et al. 2014), the intraseasonal variation of the anomalous moisture is assumed to be determined by a linear combination of the frictionally driven boundary layer (BL) moisture convergence, the horizontal moisture advection, and the net moistening of evaporation and precipitation. This theory can appropriately simulate not only the eastward propagation phase speed but also the westward-moving group velocity of the MJO, both of which arise primarily from the meridional advection of the low-frequency background moisture due to the MJO-related wind anomalies (Adames and Kim 2016).

The second theory highlights the key role of dynamical wave feedbacks, such as the frictionally coupled Kelvin–Rossby wave theory (Wang 1988; Wang and Rui 1990). In this theory, the moisture tendency is neglected. This means that the diabatic precipitation heating is directly parameterized as a function of the moisture convergence, which is essentially a Kuo-type cumulus scheme (Kuo 1965). Therefore, unlike the so-called moisture mode theory, which retains the two-way interaction of moisture and convection, this theory can only resolve the wave feedback, regardless of the key role of moisture–convection feedback (Wang et al. 2016; Liu and Wang 2017b). The simulated eastward propagation of the MJO is completely determined by the dynamical wave feedbacks, which involve both the equatorially eastward-propagating Kelvin waves and westward-moving Rossby waves. For example, the easterly wind anomaly of a Kelvin wave can drive large-scale moisture convergence and therefore destabilization to the east of the convective center, contributing to the further eastward propagation of the MJO. This mechanism associated with Kelvin wave dynamics appears to be agreed upon (e.g., Wang et al. 2017; Chen and Wang 2018a). However, there are still contrasting views concerning the impact of Rossby wave, one of which (called the “drag effect”), as derived from observational diagnostics (Wang and Lee 2017; Chen and Wang 2018a), aquaplanet GCM experiments (Kang et al. 2013), and theoretical studies (Wang and Chen 2016; Wang et al. 2016), suggests that the tight coupling of the equatorially westward-propagating Rossby waves will naturally slow down the eastward-propagating phase speed of the MJO. However, the other view (called the “acceleration effect”) argues that a stronger Rossby wave to the west can enhance the east–west zonal asymmetry of the moisture tendency and therefore favors eastward propagation (e.g., Hsu and Li 2012; Wang et al. 2017, 2018b).

The third theory, namely the frictionally coupled dynamic moisture mode theory (Wang and Chen 2016; Liu and Wang 2017a), has been put forward as a coupling of the first two theories. The eastward propagation of the MJO can be well reproduced by the three-way interaction among convective heating, moisture, and wave–BL dynamics. In fact, Liu and Wang (2017b) have pointed out that both dynamical wave feedback and moisture–convection feedback are essential to shape the MJO. They showed that the former is helpful to slow down the eastward propagation and increase the growth rate of the planetary waves, while the latter is responsible for producing dispersive MJO modes. On the basis of previous work, in this study, we try to explain the distinctive propagation characteristics of the MJO under different ENSO background states by examining both dynamical wave feedback and moisture–convection feedback.

Studies of the modulation of the ENSO on the MJO can be traced back to Lau and Chan (1986b), who revealed that El Niño might reduce the frequency of the MJO via the air–sea interactions. Subsequently, a large number of studies tried to understand the MJO variation under ENSO (e.g., Weickmann 1991; Li and Smith 1995; Slingo et al. 1999; Zhang and Gottschalck 2002; Hendon et al. 2007; Moon et al. 2011; Pillai and Chowdary 2016; Wu and Song 2018). Even though some uncertainties still exist, the preferred finding may be summarized as follows: the MJO intensity should vary with the different ENSO phases. Typically, the MJO activity over the EWP is enhanced in the developing phase of El Niño and reduced in and after the mature and decaying phases. Recently, with the recognition of non-canonical-type El Niño events in the 1990s, namely El Niño Modoki (Ashok et al. 2007), central Pacific (CP) El Niño (Kao and Yu 2009), or the warm-pool ENSO (Kug et al. 2009; Ren and Jin 2011), the above findings have been argued to only be true during canonical eastern Pacific (EP) El Niño years (Yuan et al. 2015; Chen et al. 2016). However, the MJO activity is always enhanced during CP El Niño years regardless of the phase (Wang et al. 2018a; Hsu and Xiao 2017).

As seen from the above studies, a large amount of effort has been made to explore the modulations of the ENSO on the MJO intensity. However, there are only a few studies mentioning variations in the MJO propagation under the ENSO background state. For example, Pohl and Matthews (2007) found significantly shorter (longer) lifetimes of the MJO during warm (cold) conditions in the EP, consistent with Gray (1988) and Goulet and Duvel (2000). Therefore, they concluded that the phase speed of the MJO is significantly higher during El Niño years. In an observational analysis combined with a theoretical validation, Liu et al. (2016) demonstrated that the boreal summer intraseasonal oscillation over the western North Pacific is dominated by a relatively high-frequency (low-frequency) oscillation during El Niño (La Niña) summers. More recently, Suematsu and Miura (2018) showed that the low-frequency (LF; period > 60 days) basic-state sea surface temperature (SST) anomaly pattern has a robust connection with the eastward propagation of the MJO. Composites of the LF SST anomalies showed contrasting patterns for the eastward propagating MJO from the IO to the EWP and those short-lived, stagnant, and convective events, which were associated with positive and negative SST anomalies over the EWP, respectively.

Enlightened by these studies, we conjecture that the MJO propagation tends to be faster during El Niño years and slower during La Niña years. As an example, Figs. 1a and 1b provide a comparison of time–longitude sections of the equatorial (15°S–15°N), MJO-filtered (Wheeler and Kiladis 1999; Kiladis et al. 2009) outgoing longwave radiation (OLR) anomalies during the 1997/98 El Niño and 2007/08 La Niña winters, respectively. We can see that the eastward propagation of the MJO convection is very fast during the entire winter season of 1997/98. Meanwhile, during the boreal winter of 2007/08, especially prior to around 30 January, the eastward movement is very slow, even though the convection becomes fast again after early February. How does the complexity of the ENSO (Timmermann et al. 2018) modulate the phase speed of the MJO? How do the moisture–convection feedback and the dynamical wave feedback work under different ENSO background states? These questions are the primary focus of this study. The rest of this paper is arranged as follows. In section 2, we introduce the data and methodology. The basic characteristics under different ENSO background states are shown in section 3. In section 4, we explore the mechanisms through which the ENSO modulates the fast and slow MJO modes. Section 5 discusses possible roles of other LF variabilities over the Indo-Pacific warm pool, tropical–extratropical interactions, and dynamically and thermodynamically intraseasonal air–sea interactions in modulating the MJO eastward propagation. A summary and concluding remarks are given in section 6.

Fig. 1.

Hovmöller diagrams of equatorial (15°S–15°N) OLR anomalies (W m−2) during the (a) 1997/98 and (b) 2007/08 boreal winters. The thick black contours denote the MJO-filtered OLR anomalies. The contours are drawn from −8 W m−2 with intervals of −10 W m−2. The thin magenta (blue) contours of −5 W m−2 represent the equatorial Kelvin wave (Rossby wave) filtered OLR anomalies.

Fig. 1.

Hovmöller diagrams of equatorial (15°S–15°N) OLR anomalies (W m−2) during the (a) 1997/98 and (b) 2007/08 boreal winters. The thick black contours denote the MJO-filtered OLR anomalies. The contours are drawn from −8 W m−2 with intervals of −10 W m−2. The thin magenta (blue) contours of −5 W m−2 represent the equatorial Kelvin wave (Rossby wave) filtered OLR anomalies.

2. Data and methodology

a. Data

The daily Advanced Very High Resolution Radiometer (AVHRR) interpolated OLR data from the National Oceanic and Atmospheric Administration (NOAA) satellite (Liebmann and Smith 1996) are used as an indicator of the convective activity. The daily atmospheric reanalysis data are from the European Centre for Medium-Range Weather Forecasts (ECMWF) interim reanalysis (ERA-Interim, hereinafter ERA-I; Dee et al. 2011). The three-dimensional fields consist of the zonal and meridional winds (u and υ), the vertical pressure velocity (ω), the specific humidity (q), and the geopotential (ϕ) at 19 levels from the surface to 10 hPa. The all-season real-time multivariate MJO (RMM) index of Wheeler and Hendon (2004) is also used here. We identify the El Niño and La Niña events via the monthly global 2° × 2° Extended Reconstructed SST version 5 (ERSSTv5) (Huang et al. 2017), which incorporates a new release of the International Comprehensive Ocean–Atmosphere Dataset 3.0 (Freeman et al. 2017), a decade of near-surface data from Argo floats, and a new estimate of centennial sea ice from the Hadley Centre Sea Ice and Sea Surface Temperature dataset version 2 (Titchner and Rayner 2014). Huang et al. (2017) showed that, compared to the ERSST version 4 (Huang et al. 2015, 2016), this newest version of the SST dataset has improved the magnitudes of El Niño and La Niña events. Both the NOAA OLR and the ERA-I data have a horizontal resolution of 2.5° × 2.5°. The 37-yr data from 1980 to 2016 are analyzed in this study, and we focus on the boreal winter from November to April when both the MJO and ENSO peak.

b. Methodology

1) Data processing

The daily (monthly) anomalies were obtained by removing the daily (monthly) climatology and the first three harmonics from 1980 to 2016. A 20–100-day bandpass filtering using the Lanczos filter (Duchon 1979) with 201 weights was further performed on the daily anomalies to extract variations on the intraseasonal time scale. The daily synoptic-scale high-frequency (HF) activity and LF basic state were obtained using a 20-day high-pass and a 100-day low-pass filtering, respectively. In Fig. 1, to highlight the eastward propagation of the wave envelope associated with the large-scale MJO signals, we performed a zonal wavenumber–frequency filtering (Wheeler and Kiladis 1999). The retained range is 1–5 for zonal wavenumber and 1/100–1/20 cpd for the frequency. A filtering with a broad frequency range can adequately simulate the fast (slow) eastward propagation during El Niño (La Niña) winters. Following Wheeler and Kiladis (1999), the equatorial Kelvin wave was derived based on a period of 3–20 days and wavenumbers 2–14, while the equatorial Rossby wave was based on a period of 10–40 days and wavenumbers 2–10.

2) Selection of El Niño and La Niña events

Following the definition of the NOAA Climate Prediction Center, we chose the warm and cold periods based on a threshold of ±0.5°C for the Oceanic Niño Index (ONI), which is calculated as the 3-month running mean of the ERSSTv5 anomalies in the Niño-3.4 region (120°–170°W, 5°N–5°S). The El Niño (La Niña) events were identified as the periods above (below) normal SSTs when the threshold was met for a minimum of five consecutive overlapping seasons. In this way, a total of 8 La Niña (1984/85, 1988/89, 1995/96, 1999/00, 2000/01, 2007/08, 2010/11, and 2011/12), and 10 El Niño (1982/83, 1986/87, 1991/92, 1994/95, 1997/98, 2002/03, 2004/05, 2006/07, 2009/10, and 2015/16) years were selected.

To distinguish the EP and CP El Niño events, we used the cold tongue index (CTI) and the warm pool index (WPI), which were proposed by Ren and Jin (2011). These two indices are calculated as follows:

 
{CTI=N3αN4WPI=N4αN3,α={2/5,N4N3>00,otherwise.
(1)

Here, N3 and N4 denote the Niño-3 and Niño-4 indices, which are defined as the ERSSTv5 anomalies over the regions 150°–90°W, 5°S–5°N and 160°E–150°W, 5°S–5°N, respectively. For specified El Niño events selected by the ONI, when the CTI was larger than the WPI, we regarded it as an EP-type event; otherwise, we regarded it a CP-type event (Ren et al. 2018). Motivated by the significant influence of the contrasting intensities of the different types of El Niño events on the background states of the wind, temperature, and moisture (e.g., Latif et al. 2015; Rao and Ren 2016; Geng et al. 2017) and therefore the different propagation characteristics of the MJO, in this study we also partitioned the EP-type El Niño events into super and regular El Niño cases. Because the amplitude of EP El Niño events is generally stronger than that of CP El Niño events (Kug et al. 2009), we used a Niño-3.4 index warmer than 2.5°C to select the super El Niño events. In the end, we selected three super El Niño years (1982/83, 1997/98, and 2015/16), three regular EP El Niño years (1986/87, 1991/92, and 2006/07), and four CP El Niño years (1994/95, 2002/03, 2004/05, and 2009/10), which is consistent with previous studies (e.g., Kug et al. 2009; Ren et al. 2019).

3) Compositing technique

In this subsection, we explain the methodology used to select the existing MJO events. To observe the entire life cycle of the MJO, a lagged composite technique was introduced.

The second component of the RMM (i.e., the RMM2 index) was used here to identify the possible convection peaks occurring over the IO. This is because the dominant spatial pattern of convection associated with the negative RMM2 displays a seesaw structure over the Indo-Pacific warm pool, in which enhanced convection is located in the IO, while suppressed convection prevails over the WP (Wheeler and Hendon 2004). We first selected possible MJO events based on the following procedure: 1) we identified the “day 0” (or MJO phase 2) when the RMM2 index reached a local minimum smaller than one negative standard deviation; 2) after day 0, the RMM amplitude should increase above 1.0 and remain high for at least 7 days; and 3) subsequently, the RMM index should progress through at least two phases, that is, reach the MJO phase 4, in the counterclockwise direction. The first two steps guarantee that it is a realistic MJO initiation (Straub 2013; Kiladis et al. 2014), and we do not require that the RMM amplitude be larger than 1.0 preceding phase 2; in this way, both the primary and successive events (Matthews 2008; Wei et al. 2019) are included here. The third step implies that IO-initiated MJO events should, at the very least, propagate into the western Maritime Continent (MC); therefore, both the propagating and nonpropagating MJO cases (Kim et al. 2014; Feng et al. 2015) were considered here. Finally, the Hovmöller diagram of the meridionally (10°S–10°N) averaged, 20–100-day, bandpass-filtered OLR anomaly overlaid by the MJO-filtered, wave-envelope pattern was also examined. This was motivated by the fact that the RMM index may exaggerate the role of the upper-level circulation signals (Straub 2013). More specifically, we observed if the OLR anomaly showed an organized (the convective center evolved continuously for at least 10 days following day 0), large-scale (maximal zonal range > 50°), and eastward-propagating (phase speed > 2.0 m s−1) pattern. We also tested the sensitivity of our results to some key parameters by modulating the duration of high (>1) RMM amplitude since day 0 ranging from 7 to 9 days, propagating phases ranging from 2 to 3, and maximal zonal distances ranging from 50° to 60°. As a consequence, the total number of MJO events only varied from 29 to 37 under the ENSO background state. Nevertheless, the composite results remain unchanged whether using 29 or 37 cases.

Once the events were isolated, we obtained the composite 91-day (from −45 days to 45 days) time series S¯=S¯(x,t0τ) associated with any meteorological field S with any dimensionality, where t0 is the composite lag 0, and τ is the time lag ranging from −45 days to 45 days. To test the significance, we used the following statistic that obeys the Student’s t distribution with a degree of freedom (DOF) of N − 1:

 
T=μσ2N1,
(2)

where μ and σ2 are the population mean and variance, respectively. Using the sampling technique, the T statistic can be evaluated as follows:

 
T^=μ^σ^2n1,
(3)

where μ^ and σ^2 denote the sample mean and variance, respectively, and n is the DOF, that is, the total number of MJO cases. The sample mean μ^ is significant when the T^ value is larger than the threshold at the 90% confidence level.

4) Evaluating the MJO phase speed

To quantitatively examine the modulation of the ENSO on the eastward propagation speed of the MJO, we developed a new algorithm to objectively calculate the phase speed. The methodology is introduced step by step as follows.

  • Step 1: For a given time (from −30 days to 30 days)–longitude (30°E–180°) pattern of the composite equatorial (15°S–5°N) OLR anomalies, we first identified the longitudinal location of the OLR minimum at lag 0 and stored the result as (x0, t0). However, there may be multiple equatorial minima at lag 0. Under such scenarios, we only considered those that finally evolved into a large-scale, eastward propagating MJO signal.

  • Step 2: We drew two reference lines passing through the point (x0, t0) with a slope of υmin = 2 m s−1 and υmax = 10 m s−1, respectively. The longitudinal locations of the OLR minimum at lags 1 and −1 are chosen from 
    {x|x0+υmindtxx0+υmaxdt}
    (4)
    and 
    {x|x0υmaxdtxx0υmindt},
    (5)
     respectively, where dt = 1 day. A tacit assumption is that a reasonable MJO phase speed should range from 2.0 to 10.0 m s−1 (Ling et al. 2013; Zhang and Ling 2017). The results were stored as (x1, t1) and (x−1, t−1), respectively.
  • Step 3: Step 2 was repeated but the referenced point (x0, t0) was updated to (xi, ti) (i = 1, …, 9) for negative time lags and to (xj, tj) (j = 1, …, 19) for positive time lags. Then, the next newly tracked points of the OLR minimum at lags of j + 1 and i + 1 were respectively chosen from 
    {x|xj+υmindtxxj+υmaxdt}
    (6)
    and 
    {x|xiυmaxdtxxiυmindt}.
    (7)
  • Step 4: The phase speed of this space–time field was evaluated as the regression coefficient of X (in meters) versus T (in seconds), where X = (x−10, … , x0, … , x20), and T = (t−10, …, t0, … , t20).

We also carried out a series of sensitivity experiments with continuously varying parameter values for υmin and υmax. The results showed that the evaluated phase speed remains unchanged when specifying any combination of υmin and υmax from the following sets:

 
{υmin|0.2υmin3.0},
(8)
 
{υmax|9.6υmax12.6}.
(9)

When choosing the values of these two parameters beyond sets (8) and (9), the evaluated phase speed becomes unrealistic. For example, the MJO propagation speed is overestimated when using a value of υmin larger than 3.0 m s−1 or a value of υmax larger than 12.6 m s−1. Conversely, the phase speed is underestimated with a parameter value of υmin smaller than 0.2 m s−1 or a value of υmax smaller than 9.6 m s−1. Zhang and Ling (2017) also showed that the propagation speed of MJO can exist in the range of 2.0–10.0 m s−1, which is just contained within the sets (8) and (9).

3. Modulation of ENSO on the MJO

a. Power spectrum analysis

Figures 2a and 2b show the longitudinal distribution (40°E–120°W) of the power spectrum of the daily, equatorial (10°S–10°N) OLR anomaly for the 10 El Niño and 8 La Niña winters during the period of 1980–2016. The data connection method is the same as that used by Liu et al. (2016). The significance of the power spectra was obtained by comparing the power spectral variance of the input time series with that of the background red noise at the significance level of 0.1. It is interesting to note that the MJO in the IO during El Niño winters is dominated by a much weaker HF oscillation with a significant spectral peak of 40 days. However, over the IO during La Niña winters, the MJO spectrum variance is very strong and exhibits a significant double-peak spectrum with both a 40-day HF oscillation and an 80-day LF oscillation, which implies that the La Niña background state may breed two types of intraseasonal modes characterized by distinctive frequencies (e.g., Saji et al. 2006; Izumo et al. 2010). The spatial distribution of the 20–100-day variance over the Indo-Pacific region was also examined (figure not shown), and it was found that the MJO intensity of El Niño winters over the IO is weaker than that of La Niña winters, consistent with Wu and Song (2018) and Li and Mao (2019).

Fig. 2.

Longitudinal distributions of the power spectral variance of the equatorially (10°S–10°N) averaged OLR anomalies (W m−2) during (a) El Niño, (b) La Niña, (c) super El Niño, (d) regular EP El Niño, and (e) CP El Niño conditions. The black contour outlines those anomalies passing the Student’s t test at the 90% confidence level. The two horizontal dashed lines indicate periods of 50 and 100 days. The horizontal axis indicates the longitude (°), while the vertical axis indicates the period (days).

Fig. 2.

Longitudinal distributions of the power spectral variance of the equatorially (10°S–10°N) averaged OLR anomalies (W m−2) during (a) El Niño, (b) La Niña, (c) super El Niño, (d) regular EP El Niño, and (e) CP El Niño conditions. The black contour outlines those anomalies passing the Student’s t test at the 90% confidence level. The two horizontal dashed lines indicate periods of 50 and 100 days. The horizontal axis indicates the longitude (°), while the vertical axis indicates the period (days).

Over the MC–Pacific region, as seen in Figs. 2a and 2b, the frequency distribution of the spectrum in El Niño is much narrower than that in La Niña. For example, the former displays a very regular and much stronger HF oscillation with a period of 50 days, whereas in the latter, the oscillating period displays a mixture with the variance peak occurring at 40–60 days over the MC and 60–70 days over the western CP. In addition, the El Niño condition also displays a 50-day oscillation at 150°W, which indicates that the MJO may propagate farther east during the warm phase of the ENSO cycle. This spectral analysis may imply that the MJO eastward propagation should be faster in the warm phase and slower in the cold phase of the ENSO cycle (Gray 1988; Goulet and Duvel 2000; Pohl and Matthews 2007).

Previous studies have demonstrated that the MJO intensity in the Indo-Pacific warm pool undergoes a strong sensitivity to the diversity of the El Niño (e.g., Hsu and Xiao 2017; Wang et al. 2018a). Is this still true for the main periodicity of the MJO? To answer this question, we calculated the power spectrum for the four CP El Niño, three super El Niño, and three regular EP El Niño winters during the period from 1980 to 2016 (Figs. 2c–e). As can be seen, over the IO–MC region, the variance in the 20–100-day frequency range is very strong during the CP El Niño events (Fig. 2e). For the EP El Niño events, however, the variance is very weak, especially for the super El Niño events (Fig. 2a), whereas over the equatorial Pacific both the regular EP and CP El Niño conditions support strong MJO activity. More interestingly, the predominate period of the winter MJO over the IO also displays a clear difference with 30 days in EP El Niño and 45 days in CP El Niño. Over the MC and equatorial Pacific, the spectrum of the super El Niño events exhibits a broad frequency ranging from 22 to 80 days and the EP El Niño events also exhibit a broad-frequency oscillation (35–80 days). For CP El Niño, the oscillation is very regular with a narrow period range of 40–60 days.

b. Hovmöller diagrams

Figures 3a and 3b shows the propagation patterns of the composite meridionally averaged (15°S–5°N) OLR anomalies during the El Niño and La Niña winters, respectively. As expected from Fig. 2a, the MJO over the IO during the warm phase of the ENSO cycle (Fig. 3a) is weak while quickly propagating toward the east (~5.8 m s−1). In the cold phase of the ENSO cycle (Fig. 3b), however, the MJO moves eastward slowly (~4.1 m s−1) with a much stronger amplitude, especially over the IO. Over the EWP, both the enhanced and suppressed phases of the MJO are slightly stronger during El Niño, implying that a warm (cold) ENSO event may support the propagating (nonpropagating) MJO events (Kim et al. 2014; Feng et al. 2015; Wang et al. 2017; Zhang and Ling 2017; Fu et al. 2018). Representations of the MJO properties in Figs. 3a and 3b, including the phase speed, amplitude, and propagation, are similar to those shown in Fig. 4 of Yadav and Straus (2017) and in Fig. 4 of Henderson and Maloney (2018), even though they primarily focus on the North Atlantic Oscillation and the high-latitude blocking response to the MJO, respectively. Our results, together with previous studies (e.g., Pohl and Matthews 2007; Henderson and Maloney 2018; Suematsu and Miura 2018), strongly suggest that the LF SST anomalies exert a fundamental influence on the eastward propagation speed of the boreal winter MJO. In addition, the dispersive property is also modulated by the ENSO cycle; as seen here: the eastward propagation pattern of the OLR anomalies displays a clear westward group velocity (Adames and Kim 2016; Chen and Wang 2018b) during the El Niño winters (Fig. 3a), whereas during the La Niña winters (Fig. 3b) the westward dispersion is weak.

Fig. 3.

Lagged composites of the equatorial (15°S–5°N) OLR anomalies (W m−2) for the (a) 10 El Niño, (b) 8 La Niña, (c) 3 Super El Niño, (d) 3 regular EP El Niño, and 4 CP El Niño winters. The blue (red) shading with interval of 3 W m−2 indicates the enhanced (suppressed) convection. The dotted area represents those anomalies passing the two-tailed Student’s t test at the confidence level of 90%. The dashed magenta lines show the regression lines for the tracked longitude locations. The evaluated phase speed of the enhanced convection is 5.8 m s−1 for El Niño, 4.1 m s−1 for La Niña, 7.1 m s−1 for super El Niño, 5.3 m s−1 for regular EP El Niño, and 5.2 m s−1 for CP El Niño winters.

Fig. 3.

Lagged composites of the equatorial (15°S–5°N) OLR anomalies (W m−2) for the (a) 10 El Niño, (b) 8 La Niña, (c) 3 Super El Niño, (d) 3 regular EP El Niño, and 4 CP El Niño winters. The blue (red) shading with interval of 3 W m−2 indicates the enhanced (suppressed) convection. The dotted area represents those anomalies passing the two-tailed Student’s t test at the confidence level of 90%. The dashed magenta lines show the regression lines for the tracked longitude locations. The evaluated phase speed of the enhanced convection is 5.8 m s−1 for El Niño, 4.1 m s−1 for La Niña, 7.1 m s−1 for super El Niño, 5.3 m s−1 for regular EP El Niño, and 5.2 m s−1 for CP El Niño winters.

To further examine the modulation of the different types of El Niño background states on the MJO propagation, the composite equatorial (15°S–5°N) OLR anomalies during the CP, regular EP, and super El Niño events were also investigated (Figs. 3c–e). As demonstrated in previous studies (e.g., Yuan et al. 2015; Chen et al. 2015) and implied in Figs. 2c–e, the much stronger convective anomalies of the MJO during the CP El Niño winter (Fig. 3e) can propagate farther eastward. In the EP El Niño winters, especially the super El Niño winters (Fig. 3c), however, the MJO is much weaker. With the small number of MJO samples occurring under EP El Niño conditions, especially under regular EP El Niño conditions, the dotted areas that can pass the Student’s t significance test, even at the 90% confidence level, in the composite results are very sparse. Regardless, the strong evolution of the dry phase from day −20 to day 0 is still significant. In addition, the phase speed of the MJO also manifests diversity. In regular EP (CP) El Niño events, the evaluated speed is approximately 5.3 (5.2) m s−1. The fastest propagation occurs during super El Niño (~7.1 m s−1). It is interesting to observe that the westward group velocity is primarily contributed to by the EP El Niño background state (Fig. 3d); however, the group velocity is nearly zero during CP El Niño.

c. Evolution patterns of MJO convection and circulation

Before studying the mechanism of the ENSO modulating the fast and slow MJO modes, we first examine the evolution patterns of the MJO convection and circulation under the El Niño and La Niña winters (Fig. 4). There are four primary differences that can be observed from the contrasting evolutions: 1) the propagation speed of the MJO is much faster in El Niño winters than in La Niña winters, especially the eastward progression of the enhanced phases from day −5 to day 15; 2) the overall amplitude of the suppressed convections over the equatorial western CP during El Niño is well organized and stronger than that during La Niña; 3) as a major response to the more strongly suppressed convection over the equatorial western CP, the coupled equatorial easterly wind anomalies are also stronger in El Niño winters; and 4) the convective phase of the MJO tends to have a meridionally symmetric structure with respect to the equator and can smoothly propagate across the MC along the equator during El Niño winters, whereas the MJO evolution during La Niña winters tends to support a “detour” (Kim et al. 2017) via the southern MC.

Fig. 4.

Evolution patterns of the composite OLR (W m−2; shading) and 850-hPa wind (m s−1; vectors) anomalies from day −15 to day 15 every 5 days during (a) El Niño and (b) La Niña winters. The contour interval is 3 W m−2. The dotted areas, as well as the color vectors, denote those anomalies passing the two-tailed Student’s t test at the 90% confidence level.

Fig. 4.

Evolution patterns of the composite OLR (W m−2; shading) and 850-hPa wind (m s−1; vectors) anomalies from day −15 to day 15 every 5 days during (a) El Niño and (b) La Niña winters. The contour interval is 3 W m−2. The dotted areas, as well as the color vectors, denote those anomalies passing the two-tailed Student’s t test at the 90% confidence level.

Can these differences in shape, especially the leading suppressed convection and the associated easterly wind anomalies over the western CP, contribute to the fast and slow MJO propagation modes? We try to answer this question in the following section.

4. How does the ENSO modulate the fast and slow MJO modes?

In this section, we try to unravel the mechanism through which the ENSO modulates the propagation speed of the MJO. First, the aforementioned dynamical wave feedback is examined based on a dynamics-oriented diagnostic proposed by Wang et al. (2018), including the coupling between the equatorial waves and MJO convection, and the vertical structures of the MJO associated circulation, moisture, and precipitation heating. Second, to evaluate the moisture–convection feedback, a column-integrated moisture budget analysis from the surface to 100 hPa is performed for El Niño and La Niña winters, respectively.

a. Dynamical wave feedback

We start with the diagnosis of the horizontal structures of the MJO circulation in the lower troposphere. Figure 5 shows the composite 850-hPa wind anomaly overlaid with its zonal component at lag 0 for the fast MJO mode during El Niño and for the slow mode during La Niña. A clear zonal asymmetry can be observed in this figure in terms of the MJO circulation between the fast and slow MJO modes. First, for the fast MJO mode (Fig. 5a), the organized easterly wind anomaly over the EWP is very strong and even reaches an amplitude of 6.0 m s−1, whereas it is much weaker (~2.5 m s−1) and nonorganized in the slow MJO mode (Fig. 5b), suggesting that the relatively stronger easterly Kelvin wind anomaly located over the EWP may favor a faster eastward propagation of the MJO (Chen and Wang 2018a). Second, to the west of the convective center, the zonal range of the westerly wind anomaly in the fast MJO mode is much larger than that in the slow mode. However, in the slow mode, the equatorial westerly wind anomaly is fairly localized and only appears over the western-central IO. In addition, the maximal equatorial westerly wind anomaly is located closer to the convective center for the slow mode. These evident contrasts may imply that the equatorial Rossby wave has a weak coupling with the MJO during El Niño; therefore, it can move farther toward the west, which is well reflected in the significantly widespread westerly wind anomalies over Africa and the EP. In the La Niña condition, conversely, the Rossby wave may have a tight coupling with the MJO as can be seen from the very limited westward propagation of the equatorial westerly wind anomaly. One could argue that the equatorially westerly wind anomaly for the fast MJO mode may be a dry Kelvin wave response to the strong negative convection over the EWP; however, its much wider longitudinal range may offer evidence that it is also rooted, possibly more so, in the moist Rossby wave response to the positive convection over the IO.

Fig. 5.

Horizontal structures of the composite lower-tropospheric (850 hPa) wind (m s−1; vector) anomalies and their zonal components (m s−1; shading) at lag 0 during (a) El Niño and (b) La Niña winters. The yellow (blue) shading shows the westerly (easterly) wind anomaly, and the contour interval is 0.5 m s−1. The dotted areas denote those anomalies passing the two-tailed Student’s t test at the 90% confidence level. The red stars mark the deep convective center of the MJO.

Fig. 5.

Horizontal structures of the composite lower-tropospheric (850 hPa) wind (m s−1; vector) anomalies and their zonal components (m s−1; shading) at lag 0 during (a) El Niño and (b) La Niña winters. The yellow (blue) shading shows the westerly (easterly) wind anomaly, and the contour interval is 0.5 m s−1. The dotted areas denote those anomalies passing the two-tailed Student’s t test at the 90% confidence level. The red stars mark the deep convective center of the MJO.

To further explain of the above dynamical wave feedback, Fig. 6 shows the composite 850-hPa geopotential anomalies at lag 0 for the fast and slow MJO modes. As we can see in Fig. 6a, the geopotential anomaly for the fast MJO mode displays a well-defined zonal wavenumber-1 structure. It is interesting to see that the two low pressure centers (~30°S and 40°N) to the west are located very far from the convective center, which confirms that the Rossby wave gyres have a weak coupling with the fast MJO mode during the El Niño condition. For the slow mode during La Niña (Fig. 6b), the negative geopotential anomaly penetrates into the entire equatorial region (10°S–10°N). In addition, it can be clearly seen that the two off-equatorial Rossby wave gyres are well coupled with the convective center. This implies that a tight coupling of the westward-moving Rossby wave may effectively drag on the MJO convection, therefore clearly slowing the eastward propagation (Kang et al. 2013; Wang et al. 2018b).

Fig. 6.

As in Fig. 5, but for the geopotential anomaly (gpm).

Fig. 6.

As in Fig. 5, but for the geopotential anomaly (gpm).

In particular, the circulation response of the slow MJO mode manifests as a clearer Gill-like response (Gill 1980) and a narrower meridional scale (see the negative geopotential anomalies over 40°–150°E), whereas for the fast MJO mode, the meridional scale of the anomalous circulation is much larger. For the steady atmospheric response to the idealized equatorial diabatic heating, the Kelvin wave exists to the east, while to the west two Rossby gyres straddle the north and south sides of the equator. Close to the equator, the easterly Kelvin (westerly Rossby) wave leads (trails) the convective center. Because of the 3 times larger eastward-propagating phase speed of the Kelvin wave (~15 m s−1) compared to the westward-moving Rossby wave (~5 m s−1), the ratio of the zonal extent between the easterly equatorial Kelvin wind and the westerly Rossby wind is just 3.0 (Gill 1980). Based on this discussion and following Wang and Chen (2016), a shape parameter for the MJO can be defined as the ratio of the zonal extent of the equatorial easterly “Kelvin” (below −0.2 in terms of the normalized value) versus the westerly “Rossby” (above 0.2 in terms of the normalized value) averaged between 5°S and 5°N. We find that the slow MJO mode has a more Gill-like response with a shape parameter of 1.7, which is much larger than that of the fast MJO mode (0.5). This suggests a dramatic dependence of the MJO’s eastward propagation speed on the structure of the MJO itself (Wang and Chen 2016; Wang et al. 2016; Wang and Lee 2017; Wang et al. 2018).

Moreover, the propagation speed of the MJO acts to interactively influence the structure of itself. For example, the fast MJO mode corresponds to a small shape parameter, primarily due to the Doppler effect induced by the fast moving diabatic heating. Conversely, the more standing-like heating source of the slow MJO mode just excites a more Gill-like response with a larger shape parameter. These arguments imply that the MJO may be regarded as a structure-propagation nexus (Wang et al. 2019).

To represent the intensity of the equatorial waves, we also calculated the ratio of the maximum equatorial Rossby westerly and Kelvin easterly wind anomalies, namely, the so-called westerly intensity index proposed by Wang and Chen (2016). The results showed that the westerly intensity index is 0.73 for the slow MJO mode during La Niña but only 0.48 for the fast MJO mode during El Niño. This further proves that a stronger Kelvin wave component versus Rossby wave component is associated with a faster propagation of the MJO.

The above discussion focused on the horizontal structures of the MJO; however, some more interesting phenomena can be detected when examining the vertical distributions. Accordingly, we further investigated the vertical structures of the MJO-associated circulation, moisture, and atmospheric apparent moisture sink Q2 anomalies (Fig. 7). The Q2, which represents the precipitation heating to some extent, is estimated from the residual between the moisture tendency and the adiabatic advection processes (Yanai et al. 1973). As we can see in Figs. 7a and 7b, for the fast MJO mode during the El Niño winters, there is an obviously stronger east–west zonal circulation cell leading a strong updraft over the IO. Chen and Wang (2018a) called this the “front Walker cell” (FWC), and it serves as a new mechanism to understand the eastward propagation of the MJO. The strong leading suppressed convection over the EWP supports a stronger downdraft there. Meanwhile, mass continuity drives a stronger easterly wind anomaly associated with the Kelvin wave, which initiates an extended premoistening perturbation to the east (120°–150°E). The corresponding lower-level precipitation heating anomalies are also significantly stronger and even extend upward above 700 hPa, suggesting massive energetic congestus clouds developing to the east of the deep convection. The congestus convection further predestabilizes the lower-level atmosphere likely via moistening and heating due to the subgrid-scale processes of entrainment and condensation, respectively, and therefore speeds up the eastward propagation.

Fig. 7.

Composite vertical structures of the equatorial (10°S–5°N) specific humidity (g kg−1; shading) and (u, ω × 300) (vector) anomalies at lag 0 during (a) El Niño and (c) La Niña winters. To clearly reflect the updraft and downdraft, the vertical pressure velocity ω has been amplified by a factor of 300. The contour interval is 0.05 g kg−1. The dotted areas, as well as the color vectors, denote those anomalies passing the Student’s t test at the 90% confidence level. (b),(d) As in (a) and (c), respectively, but for the precipitation heating (shading) and vertical circulation (vector). The significance test for the vectors in (a) and (c) is based on the vertical pressure velocity anomaly and for those in (b) and (d) on the zonal wind anomaly.

Fig. 7.

Composite vertical structures of the equatorial (10°S–5°N) specific humidity (g kg−1; shading) and (u, ω × 300) (vector) anomalies at lag 0 during (a) El Niño and (c) La Niña winters. To clearly reflect the updraft and downdraft, the vertical pressure velocity ω has been amplified by a factor of 300. The contour interval is 0.05 g kg−1. The dotted areas, as well as the color vectors, denote those anomalies passing the Student’s t test at the 90% confidence level. (b),(d) As in (a) and (c), respectively, but for the precipitation heating (shading) and vertical circulation (vector). The significance test for the vectors in (a) and (c) is based on the vertical pressure velocity anomaly and for those in (b) and (d) on the zonal wind anomaly.

Conversely, for the slow MJO mode during the La Niña winters (Figs. 7c,d), the leading suppressed convection over the EWP is much weaker and nonorganized (refer to Fig. 4), resulting in a weaker subsidence and therefore a weaker FWC. These weaker easterly Kelvin wind anomalies only produce regionally confined BL moistening, and these weaker shallow congestus clouds tend to trigger a relatively weaker lower-level predestabilization effect and therefore a slower eastward propagation of the MJO. These details confirm the critical role of the leading suppressed convection in contributing to the eastward MJO propagation (Kim et al. 2014; Chen and Wang 2018a), and significantly support the distinct responses of the MJO propagation during different ENSO phases.

b. Moisture budget analysis

Before performing a budget analysis, we should first examine the variations of the LF background states in the different types of El Niño and La Niña events (Fig. 8). For the El Niño background state characterized by a warm SST anomaly over the equatorial EP (Figs. 8a,f,k), there are strong westerly wind and positive moisture anomalies over the central EP as well as a negative moisture anomaly off the equator. When the MJO convection is located over the eastern IO (see day 0 in Figs. 4 and 5), the intraseasonal easterly wind anomaly prevails over the MC–EWP region. The eastward gradient of the LF background moisture may induce positive zonal advection due to the easterly wind anomaly (Liu and Wang 2016). Meanwhile, both the strong negative meridional gradient of the specific humidity over the MC–EWP region along 4°N and the positive moisture gradient along 7°S may also produce a net moistening effect due to the intraseasonal southerly and northerly wind anomalies, respectively (Kim et al. 2014; Wang et al. 2017). This scenario is also true for the super (Figs. 8c,h,m), regular EP (Figs. 8d,i,n), and CP El Niño (Figs. 8e,j,o) conditions, in which the moistening effect may decrease progressively due to the weakening atmospheric response to the decreasing LF SST anomaly. During the La Niña winters (Figs. 8b,g,l), however, the central-EP displays nearly opposite patterns including the LF background easterly wind anomaly and a poleward meridional and westward moisture gradient anomalies, all of which may cause a net drying tendency over the MC–EWP region. This suggests that the zonal gradient of LF SST from the eastern to western hemisphere may serve as a new indication of the MJO propagation speed. We prove the above qualitative descriptions in the following.

Fig. 8.

(left) Composite monthly SST anomaly (°C) for the (a) 10 El Niño, (b) 8 La Niña, (c) 3 super El Niño, (d) 3 regular EP El Niño, and (e) 4 CP El Niño winters. (center),(right) As in the left panels, but for the 100-day, low-pass-filtered, lower-tropospheric (850 hPa) (f)–(j) zonal wind (m s−1) and (k)–(o) specific humidity (g kg−1) anomalies. The dotted areas denote those anomalies passing the Student’s t test at the 90% confidence level.

Fig. 8.

(left) Composite monthly SST anomaly (°C) for the (a) 10 El Niño, (b) 8 La Niña, (c) 3 super El Niño, (d) 3 regular EP El Niño, and (e) 4 CP El Niño winters. (center),(right) As in the left panels, but for the 100-day, low-pass-filtered, lower-tropospheric (850 hPa) (f)–(j) zonal wind (m s−1) and (k)–(o) specific humidity (g kg−1) anomalies. The dotted areas denote those anomalies passing the Student’s t test at the 90% confidence level.

The moisture or moist static energy budget analysis has been widely recognized as a powerful tool to understand both the eastward propagation of the winter MJO (e.g., Maloney 2009; Hsu and Li 2012; Kim et al. 2014, 2017; Jiang 2017) and the northward propagation of the boreal summer intraseasonal oscillation (e.g., Adames et al. 2016; Jiang et al. 2018). Here to quantitatively prove the dependence of the fast and slow eastward propagations of the MJO on the horizontal advection-induced net moistening over the EWP, a column-integrated moisture budget is further analyzed.

Figures 9a and 9b show the spatial distributions of the composite column-integrated (from the surface to 100 hPa) specific humidity and its temporal tendency anomalies at lag 0 for the fast MJO mode during El Niño winters and the slow mode during La Niña winters, respectively. We can see that there is a strong net moistening over the equatorial eastern MC–EWP region, the Bay of Bengal, and the southern MC, and a strong net drying in the equatorial western Indian Ocean (WIO) and the northwestern Pacific for the fast MJO mode during El Niño (Fig. 9a). Conversely, for the slow MJO mode during La Niña (Fig. 9b), both the positive moisture tendency over the equatorial eastern MC–EWP and the negative tendency over the equatorial WIO and the northwestern Pacific are fairly weak. Therefore, along the equator, the zonal asymmetry of the moisture tendency for the fast MJO mode during El Niño is stronger than that for the slow mode during La Niña (refer to the dashed lines shown in Fig. 9c). This suggests that a strong (weak) zonal asymmetry of the moisture tendency should support a fast (slow) eastward propagation of the MJO (Hsu and Li 2012; Hsu et al. 2014; Wang et al. 2017, 2018b). As seen in Figs. 9a and 9b, the positive anomalies of the moisture tendency over the southern MC in the slow MJO mode are much stronger than those in the fast mode (refer to the dashed lines shown in Fig. 9d), implying that the “detour” feature of the MJO (Kim et al. 2017) tends to be more frequent during La Niña conditions.

Fig. 9.

(a) Spatial distribution of the composite column-integrated specific humidity (q; contour; g kg−1) and its temporal tendency (q_t; shading; 1 × 10−10 g kg−1 s−1) anomalies for the fast MJO mode during El Niño. The contour interval is 0.1 g kg−1 and the zero contours are omitted. The dotted areas denote those anomalies passing the Student’s t test at the 90% confidence level. The two gray rectangles denote the west (WBOX; 10°S–10°N, 40°–80°E) and east (EBOX; 4°S–14°N, 115°–150°E) boxes used for the budget analysis. (b) As in (a), but for the slow MJO mode during La Niña. (c) Zonal distributions of q (solid) and q_t (dashed) over 10°S–10°N for El Niño (red) and La Niña (blue). The left (right) axis indicates q (q_t). (d) As in (c), but for the meridional distribution over 60°–90°E (100°–140°E) associated with q (q_t).

Fig. 9.

(a) Spatial distribution of the composite column-integrated specific humidity (q; contour; g kg−1) and its temporal tendency (q_t; shading; 1 × 10−10 g kg−1 s−1) anomalies for the fast MJO mode during El Niño. The contour interval is 0.1 g kg−1 and the zero contours are omitted. The dotted areas denote those anomalies passing the Student’s t test at the 90% confidence level. The two gray rectangles denote the west (WBOX; 10°S–10°N, 40°–80°E) and east (EBOX; 4°S–14°N, 115°–150°E) boxes used for the budget analysis. (b) As in (a), but for the slow MJO mode during La Niña. (c) Zonal distributions of q (solid) and q_t (dashed) over 10°S–10°N for El Niño (red) and La Niña (blue). The left (right) axis indicates q (q_t). (d) As in (c), but for the meridional distribution over 60°–90°E (100°–140°E) associated with q (q_t).

In addition, the moisture anomaly itself, which serves as a good indicator of the MJO intensity (Sobel and Maloney 2012, 2013; Adames and Kim 2016), is weaker for the fast MJO mode in El Niño than for the slow mode in La Niña (refer to the solid lines shown in Figs. 9c and 9d). Two implications may shed light on this interesting phenomenon. First, using only the moisture advection may be insufficient to satisfactorily explain the unstable MJO growth. Second, a weaker moisture anomaly during El Niño implies that a moisture tendency of the same size or larger can propagate a weaker moisture anomaly eastward more quickly. In summary, we can make the conjecture that a stronger zonal asymmetry of the moisture tendency does not necessarily result in a larger growth of the MJO, but instead can contribute to a faster eastward propagation. The theoretical study of Adames and Kim (2016) also showed that it is the eastward propagation rather than the selection of the planetary-scale instability that results from the moisture advection. The following analysis demonstrates that such a zonal asymmetry is largely determined by the horizontal advection, especially its meridional component, associated with the ENSO background variations.

To examine the dominant source of the zonal asymmetry of the moisture tendency, we use the following equation (Yanai et al. 1973):

 
qt=VhhqωqpQ2Lυ,
(10)

where Vh = (u, υ) denotes the horizontal velocity vector, ∇h = [(∂/∂x), (∂/∂y)] is the horizontal gradient operator, Q2 is the atmospheric apparent moisture sink, ⟨⋅⟩ indicates column integration from the surface to 100 hPa, and the prime symbol denotes the anomaly at the 20–100-day intraseasonal time scale. The first term on the right-hand side of (10) is the horizontal moisture advection. The sum of the second and third terms represents the net moistening associated with the column processes, including the large-scale adiabatic vertical motion and the subgrid-scale evaporation and condensation. Following Yanai et al. (1973), the net moistening associated with the subgrid-scale processes can be approximated as the residual term between the moisture tendency and the adiabatic advective processes.

According to the descriptions of Fig. 9, the most significant difference associated with the zonal asymmetry in terms of the moisture tendency between the fast and slow MJO modes is primarily reflected in the net drying over the equatorial IO and the net moistening over the equatorial eastern MC–EWP region. Therefore, two boxes (refer to the gray rectangles shown in Fig. 9) are specified here: a west box (WBOX: 10°S–10°N, 40°–80°E) and an east box (EBOX: 15°S–5°N, 115°–150°E). Figure 10 shows the composite column-integrated total moisture budget results averaged in WBOX and EBOX at the intraseasonal time scale (20–100 days) for the fast and slow MJO modes and the difference between them. To represent the uncertainty range of composite results, the error bars generated based on the standard deviation of sampling members are also drawn. To make the results under different amplitude MJO events comparable, each budget term was normalized by the domain-averaged (10°S–15°N, 50°–110°E) moisture anomaly. For the fast MJO mode, as seen in Figs. 10a and 10d, both the zonal and meridional advection contribute to the net drying over WBOX (i.e., over the IO) and to the net moistening over EBOX (i.e., equatorial eastern MC–EWP). For the slow MJO mode (Figs. 10b,e), even though the role of the horizontal advection is qualitatively similar to that for the fast mode, their amplitudes are very weak, especially the zonal component over WBOX and the meridional component over EBOX. From the plot of the differences in the horizontal advection between the fast and slow MJO modes (Figs. 9c,f), we can see that the dominant process distinguishing these two MJO modes is the meridional moisture advection over EBOX. The zonal moisture advection also enhances the net drying over WBOX and the net moistening over EBOX, therefore contributing to the zonal asymmetry of the moisture tendency and the fast eastward propagation of the MJO.

Fig. 10.

The west box (WBOX; 10°S–10°N, 40°–80°E)-averaged moisture tendency (black bar), zonal advection (red bar), meridional advection (blue bar), and net moistening associated with column processes (green bar) for (a) the fast MJO mode during El Niño and (b) the slow MJO mode during La Niña. (c) The difference between the fast and slow MJO modes. (d)–(f) As in (a)–(c), but for the east box (EBOX; 4°S–14°N, 115°–150°E). The error bars denote the uncertainty range of composite results generated based on the standard deviation of sampling members.

Fig. 10.

The west box (WBOX; 10°S–10°N, 40°–80°E)-averaged moisture tendency (black bar), zonal advection (red bar), meridional advection (blue bar), and net moistening associated with column processes (green bar) for (a) the fast MJO mode during El Niño and (b) the slow MJO mode during La Niña. (c) The difference between the fast and slow MJO modes. (d)–(f) As in (a)–(c), but for the east box (EBOX; 4°S–14°N, 115°–150°E). The error bars denote the uncertainty range of composite results generated based on the standard deviation of sampling members.

Revisiting Fig. 10f, we find that the column processes play a comparable role as to zonal advection in contributing to the net moistening over EBOX. To examine the role of the column processes, Fig. 11 shows the composite vertical profiles of the EBOX-averaged vertical advection −Q2/Lυ and the net effect of the column processes for the fast and slow MJO modes and the differences between them. As seen from the results for the fast MJO (Fig. 11a), both the vertical advection and −Q2/Lυ manifest as a vertical diploe structure, in which the node is found at 700 hPa. Above this node, the atmosphere shows a net moistening, while below it drying dominates. Because the amplitude of −Q2/Lυ is generally stronger than that of vertical advection, both the net drying and the net moistening are contributed to by the latter, consistent with previous studies (e.g., Hung and Sui 2018). In the slow mode (Fig. 11b), however, all the budget terms are very weak, especially −Q2/Lυ, below 700 hPa. Therefore, with respect to the difference between the fast and slow modes (Fig. 11c), the net moistening effect of the column processes displays a very similar distribution to −Q2/Lυ. In summary, the moistening effect of the column processes shown in Fig. 10f is primarily located above 700 hPa and contributed to by −Q2/Lυ. However, below 700 hPa, the column processes primarily offer a net heating effect, therefore further destabilizing the column and supporting a fast eastward propagation. As seen in Fig. 7b, this heating effect is largely contributed to by the strong congestus clouds below 700 hPa during El Niño.

Fig. 11.

(a) Composite vertical profiles of the east box (EBOX; 4°S–14°N, 115°–150°E)-averaged vertical moisture advection (Wadv; blue), the subgrid-scale moistening effect associated with the atmospheric apparent moisture sink Q2 (green), and the column processes (red) for the fast MJO mode during El Niño. (b) As in (a), but for the slow MJO mode during La Niña. (c) The difference between the fast and slow MJO modes. The positive values on the x axis denote moistening, while the negative values denote drying; both have units of 10−10 g kg−1 s−1. The y axis indicates the pressure level (hPa).

Fig. 11.

(a) Composite vertical profiles of the east box (EBOX; 4°S–14°N, 115°–150°E)-averaged vertical moisture advection (Wadv; blue), the subgrid-scale moistening effect associated with the atmospheric apparent moisture sink Q2 (green), and the column processes (red) for the fast MJO mode during El Niño. (b) As in (a), but for the slow MJO mode during La Niña. (c) The difference between the fast and slow MJO modes. The positive values on the x axis denote moistening, while the negative values denote drying; both have units of 10−10 g kg−1 s−1. The y axis indicates the pressure level (hPa).

To quantitatively distinguish the modulation of the ENSO on the fast and slow MJO modes, following Wang et al. (2017), we decomposed each variable P (u, υ, or q) into three time scales: the HF synoptic scale P* (<20 days), the intraseasonal time scale P′ (20–100 days), and the LF background scale P¯ (>100 days); accordingly, both the zonal and meridional advections can be divided into nine terms as follows:

 
(u¯q¯x)T1(u¯qx)T2(u¯qx*)T3(uq¯x)T4(uqx)T5(uqx*)T6(u*q¯x)T7(u*qx)T8(u*qx*)T9,
(11)
 
(υq¯y)T1(υ¯qy)T2(υ¯qy*)T3(υq¯y)T4(υqy)T5(υqy*)T6(υ*q¯y)T7(υ*qy)T8(υ*qy*)T9,
(12)

For ease of writing, we call the nine terms T1, T2, …, T9 from left to right in (11) and (12), respectively. As in the case of (10), each term in (11) and (12) is filtered to the intraseasonal (20–100-day) time scale prior to the budget analysis.

The time-scale decomposition results over the aforementioned two boxes (WBOX and EBOX) are shown in Fig. 12. Seen in Figs. 12a and 12c, the net drying associated with both the zonal and meridional advection over IO is largely contributed to by the linear zonal advections T2 and T4. This budget result holds for both the fast MJO mode during El Niño and the slow mode during La Niña, which confirms the previous studies to some extent (e.g., Maloney 2009; Andersen and Kuang 2012; Adames and Kim 2016). Influences of ENSO on the MJO can be detected from T2 and T4 (refer to the green bars in Figs. 12a and 12c) in the zonal component, which are respectively modulated by the LF winds and the moisture gradient anomalies associated with the ENSO background state variations, even though the amplitudes are relatively weaker. In addition, the nonlinear advection, including T5 and T6 of the zonal advection and T9 of the meridional advection, also plays a role in making the IO drier for the fast MJO mode. Over EBOX (i.e., over the equatorial eastern MC–EWP region), the difference between the fast MJO mode of El Niño and slow mode of La Niña has reached its maximum. For example, for the fast mode during the warm phase of ENSO cycle, the amplitude of the linear zonal advection term T4 is more than 3 times larger than that for the slow mode (Fig. 10b), leaving a net moistening effect of T4 to distinguish the two MJO modes. For the meridional components (Fig. 10f), in the fast MJO mode during El Niño, the linear advection term T4 contributes nearly all of the total net moistening over the equatorial eastern MC–EWP region. During La Niña, however, all the budget terms become very small or even negative. Because the magnitude of the meridional component is generally larger than that of the zonal component (see Figs. 10f and 12d), the modulation of the ENSO on the fast and slow MJO modes during the boreal winter is primarily rooted in the linear meridional advection term T4, that is, the moisture advection of the ENSO-related LF moisture anomaly by the MJO-scale meridional wind.

Fig. 12.

Scale decomposition of the (left) zonal advection and (right) meridional advection. (a),(c) The west box (WBOX; 10°S–10°N, 40°–80°E)-averaged nine advection terms (T1–T9) for the fast MJO modes during El Niño (red bars), the slow MJO modes during La Niña (blue bars), and their difference (green bars). (b),(d) As in (a) and (c), but for the east box (EBOX; 4°S–14°N, 115°–150°E).

Fig. 12.

Scale decomposition of the (left) zonal advection and (right) meridional advection. (a),(c) The west box (WBOX; 10°S–10°N, 40°–80°E)-averaged nine advection terms (T1–T9) for the fast MJO modes during El Niño (red bars), the slow MJO modes during La Niña (blue bars), and their difference (green bars). (b),(d) As in (a) and (c), but for the east box (EBOX; 4°S–14°N, 115°–150°E).

Another interesting phenomenon seen in Fig. 12b is that the ENSO background state variation also has a nonnegligible modulating effect on the nonlinear HF zonal advection term T9, which can be further detected in Fig. 1. As can be seen, the activity of the convectively coupled Kelvin waves (CCKWs) over MC–EWP in the 1997/98 El Niño is much stronger than that in the 2007/08 La Niña, which led to a much faster (~5.8 m s−1) MJO event initiating on 13 December 1997 and a much slower (~3.0 m s−1) MJO event initiated on 23 November 2007, respectively. This case study offers some evidence that supports the key role of the strong nonlinear upscale moisture feedback of CCKW on the fast MJO mode during El Niño. Conversely, La Niña winter corresponds to a slow MJO mode as a result of the weak upscale moisture feedback of the CCKWs. Note that our results here are somewhat similar to those of Guo et al. (2015), who showed that the reliability of GCMs in simulating the systematic eastward propagation of the MJO always corresponds to a well-simulated CCKW.

The above moisture budget analyses suggests that the modulation of the ENSO on the fast and slow MJO modes is primarily manifested as a net moistening over the equatorial eastern MC–EWP region, which then enhances the zonal asymmetry of the moisture tendency and therefore leads to a faster eastward propagation of the MJO during El Niño than during La Niña. As suggested by Fig. 8, the dominant physical processes associated with this net moistening effect have been diagnosed as the linearly horizontal advection (both meridional and zonal) of the LF background moisture anomaly by the intraseasonal (20–100 days) wind anomalies. The nonlinear advection by the HF zonal winds also plays a role in speeding up the east propagation of the MJO during El Niño conditions. The budget analysis results also imply the close relationship between the eastward propagation of the MJO and the background state variations (e.g., Kim et al. 2009; Jiang et al. 2015; Liu et al. 2016; Gonzalez and Jiang 2017; Ling et al. 2017).

5. Discussion of other factors in modulating the eastward MJO propagation

From section 4, the modulation effect of the ENSO on the fast and slow MJO modes can be ultimately rooted in the influences of LF (>100 days) circulation and convection. In this section, we discuss the possible roles of 1) other LF variabilities including the Indian Ocean dipole (IOD; Saji et al. 1999) and the quasi-biennial oscillation (QBO; Baldwin et al. 2001), 2) the tropical–extratropical interaction, and 3) the dynamically and thermodynamically intraseasonal air–sea interactions in modulating the eastward propagation of the MJO.

a. IOD

Several previous studies have focused on the year-to-year variability of the MJO under the IOD-related LF background state. For example, Shinoda and Han (2005) found that the interannual variation of the intraseasonal variability in the central and eastern equatorial IO is highly correlated with the large-scale zonal SST gradient of the IOD. Motivated by a hypothesis proposed by Saji et al. (2006), Izumo et al. (2010) proved that there exist two types of MJO modes: a higher-frequency mode with a period of 30–50 days and a lower-frequency mode with a period of 55–100 days, which show equatorially symmetric and asymmetric structures, respectively. Modulated by the changes in the background atmospheric circulation after an IOD, the HF MJO modes propagate eastward faster than the LF modes. Wilson et al. (2013) further demonstrated that the MJO becomes weaker and faster during the positive IOD condition, and the opposite is true during the negative IOD condition. However, because many of the SST anomalies associated with the IOD could be produced remotely by the ENSO in the tropical Pacific (Baquero-Bernal et al. 2002; Shinoda et al. 2004), it is interesting and necessary to distinguish the relative contributions of ENSO and the IOD to the MJO propagation, which is an important part of our ongoing and future research.

b. QBO

Despite being a leading mode of the interannual variability in the tropical stratosphere, the impacts of the QBO, which feature the changing of flow direction of the zonal mean zonal wind at 50 hPa every other year (Baldwin et al. 2001), on the MJO have not been well documented until very recently (e.g., Liu et al. 2014; Yoo and Son 2016; Marshall et al. 2017; Sun et al. 2019). In summary, it has been found that the composite OLR anomaly shows a larger negative value and a slower eastward propagation with a prolonged period of active convection in the QBO easterly (QBOE) phase than in the QBO westerly (QBOW) phase (e.g., Nishimoto and Yoden 2017; Son et al. 2017). The reasons for the fast eastward propagation of the MJO under QBOW conditions may result from two mechanisms: 1) the deep convection of the MJO itself in the QBOW phase is weaker, possibly leading to a much faster phase speed of the planetary-scale Kelvin waves (Chang 1977) and therefore the MJO (Seo and Kumar 2008; Son et al. 2017); and 2) the dry convection preceding the eastward propagation of deep convection under QBOW conditions is generally stronger than that under QBOE conditions, especially over the MC–EWP region (see Fig. 3 of Zhang and Zhang 2018). According to our findings in this study, the much stronger leading suppressed convection may speed up the eastward propagation of the MJO with a weaker intensity.

c. Tropical–extratropical interaction

Our findings indicate that ENSO modulation of the MJO propagation is indirectly caused by enhancing of the leading suppressed convection rather than by accelerating the process of deep convective clouds. However, it remains unclear what causes this leading suppressed convection. Yong and Mao (2016) suggested that the anomalous descending motion associated with suppressed convection is dynamically forced by an anomalous convergence. This convergent wind anomaly results from changes in the subtropical circulation structure forced by extratropical disturbances associated with the equatorward advection of positive potential vorticity. Chen and Wang (2018a) showed that for the successive MJO cases (Matthews 2008), the anomaly comes from the eastward propagation of the preceding dry MJO phases, while for the primary cases it is brought about by a two-way interaction between the tropical heating of the MJO and the associated tropical–extratropical teleconnection, which evolves and forms an anomalous cyclone over the western North Pacific that generates upper-level convergence.

d. Intraseasonal air–sea interaction

The intraseasonal air–sea interaction has been shown to play a key role in the MJO dynamics [see the review of DeMott et al. (2015)]. During El Niño winters, the prevailing background westerly wind anomalies against the easterly wind of the MJO over the EWP region may lead to a net warming due to the reduced surface latent heat flux or the entrainment feedback beneath the mixed layer base. In addition, the strong leading suppressed convection over the EWP may also bring in more insolation and therefore warm the oceans directly. During La Niña winters, however, the amplified winds associated with the easterly wind of the MJO and the background easterly wind of ENSO cool the oceans via enhanced wind–evaporation–entrainment–SST feedback (Wang and Xie 1998; Li et al. 2008; Wei et al. 2018). The weak leading suppressed convection over the EWP may also induce cooling via the cloud–radiation–SST feedback (Wang et al. 1995; Liu and Wang 2013). Therefore, the intraseasonal warm (cool) SST anomalies over the EWP may also feed back to the atmosphere and support the fast (slow) MJO modes.

In addition to the above thermodynamical air–sea interactions, the oceanic responses to the atmospheric forcing may also cause a dynamic feedback to the MJO propagation. Han (2005) revealed that the observed 90-day spectral peak of the SST in the equatorial IO could be explained as a result of a selective response to the 90-day winds rather than to the 30–60-day winds. This implies that oceanic feedback is more important for the lower-frequency MJO modes (Saji et al. 2006; Izumo et al. 2010) because, at lower frequencies, both the Rossby and Kelvin waves of low-order baroclinic modes have longer wavelengths, which are more effectively excited by large-scale winds (Han 2005). A quantitative investigation concerning the roles of intraseasonal air–sea interaction in contributing to different propagations of the MJO, both dynamical and thermodynamical, requires a detailed observational diagnostic analysis and an auxiliary numerical modeling study.

6. Summary and concluding remarks

Using the observations and ECMWF reanalysis data since the 1980s, this study investigated the modulation of the ENSO on the fast and slow MJO modes during boreal winter. The main conclusions are summarized as follows.

The power spectrum analysis showed that the MJO variability during El Niño winters is very uniform along the entire Indo-Pacific region and is dominated by a relatively HF oscillation with a significant spectral peak of 40–50 days. However, during La Niña winters, the spectrum is very broad and is only visible over the IO and MC, and, more interesting, the MJO variability displays a double-peak spectrum with both a 40-day HF oscillation and an 80-day low-frequency oscillation. The diversity of El Niño events also evidently results in a modulation on the MJO periodicity. For example, over the IO, the period of the MJO manifests a clear difference between 30 days in EP El Niño winters and 45 days in CP El Niño winters.

The eastward propagation speed of the MJO varies strongly for different ENSO background states. For example, the evaluated phase speeds using an objective algorithm are 5.8 and 4.1 m s−1 for the fast MJO mode during El Niño and the slow one during La Niña, respectively. In addition, the phase speed and intensity of the fast MJO mode tends to vary as a function of the El Niño amplitude, where the strongest super El Niño produces the fastest propagation (~7.1 m s−1) and the weakest MJO, while regular EP and CP El Niño winters only generate a moderate phase speed but a somewhat stronger MJO. The dispersive feature of the MJO is also evidently modulated by the ENSO background state. Even though the westward group velocity is visible during La Niña, it is considerably weaker than during El Niño. The significantly westward dispersion of the MJO wave energy during the warm phase of the ENSO is primarily contributed by the EP-type El Niño events.

The fast MJO mode during El Niño winters can propagate smoothly across the MC into the EWP, and the corresponding evolution pattern manifests an obvious meridional symmetry with respect to the equator. However, the La Niña condition tends to support an MJO “detour” via the southern MC. The horizontal distribution of the large-scale convection of the MJO indicates that that the overall amplitude of the leading suppressed convection over the EWP for the fast MJO mode is much stronger than that for the slow MJO mode.

The mechanisms by which the different ENSO background states modulate the fast and slow MJO modes are illustrated schematically in Fig. 13. From a dynamical point of view (e.g., Wang and Rui 1990; Kang et al. 2013; Wang and Chen 2016; Liu and Wang 2017b; Wang et al. 2017, 2018; Wang et al. 2018), as we can see from Fig. 13a, during the El Niño winters the leading suppressed convection over the EWP is much stronger, therefore generating a large zonal pressure gradient over the Indo-Pacific warm pool and lower-level Kelvin easterly wind anomalies. Due to mass continuity, a strong FWC (Chen and Wang 2018a) is excited to initiate an extended premoistening to the east of the deep convection. The premoistening-induced shallow congestus clouds (below 700 hPa) can destabilize the lower troposphere to the east and contribute to the fast eastward propagation (refer to Fig. 11). However, the two off-equatorial Rossby wave low pressure centers tend to decouple from the major convection, which is reflected in the distant location of the maximal equatorial westerly from the convective center and the widespread westerly to the west covering Africa and the EP. This weak coupling reduces the “drag effect” of the westward-moving Rossby wave and therefore generates a faster MJO mode. For La Niña winter (Fig. 12b), because both the leading suppressed convection and the FWC are very weak, the easterly wind anomaly of Kelvin wave over the EWP becomes very small. In addition, the two low pressure centers of the Rossby wave to the west are tightly coupled to the major convection, which can be seen from the small localized equatorial westerly wind anomaly over the IO and the closer location of the maximal westerly to the convective center. In summary, unlike the fast MJO mode during El Niño, the La Niña condition, which supports the slow eastward propagation of the MJO, generates a more Gill-like circulation response to the diabatic heating. This implies the dramatic dependence of the eastward propagation speed of the MJO on its own structure (e.g., Wang and Chen 2016; Wang et al. 2016; Wang and Lee 2017; Wang et al. 2018).

Fig. 13.

Schematic diagrams illustrating the mechanisms for how the ENSO modulates the (a) fast (~6.0 m s−1) and (b) slow (~4.0 m s−1) MJO modes during boreal winter. The cloud cluster denotes the MJO deep convection, while the preceding nonprecipitable cloud over the MC–EWP represents the net heating effect of shallow congestus clouds below 700 hPa. The FWC (purple ellipse), intraseasonal easterly and northerly wind anomalies (dashed blue arrows) and moisture anomalies (turquoise shading stands for moistening, brown shading for drying), and Kelvin–Rossby low pressure centers (green dashed ellipse) are overlaid. Along the equator, the eastward (westward) arrow indicates the westerly Rossby (easterly Kelvin) wave wind anomaly. For all the symbols, thicker lines or deeper color shadings indicate greater strengths. The thick red (blue) line outlining the central-eastern equatorial Pacific denotes the warming (cooling) associated with the El Niño (La Niña) background state. The dashed black arrows at the top of the diagrams denote the eastward propagation of the MJO circulation–convection coupled system.

Fig. 13.

Schematic diagrams illustrating the mechanisms for how the ENSO modulates the (a) fast (~6.0 m s−1) and (b) slow (~4.0 m s−1) MJO modes during boreal winter. The cloud cluster denotes the MJO deep convection, while the preceding nonprecipitable cloud over the MC–EWP represents the net heating effect of shallow congestus clouds below 700 hPa. The FWC (purple ellipse), intraseasonal easterly and northerly wind anomalies (dashed blue arrows) and moisture anomalies (turquoise shading stands for moistening, brown shading for drying), and Kelvin–Rossby low pressure centers (green dashed ellipse) are overlaid. Along the equator, the eastward (westward) arrow indicates the westerly Rossby (easterly Kelvin) wave wind anomaly. For all the symbols, thicker lines or deeper color shadings indicate greater strengths. The thick red (blue) line outlining the central-eastern equatorial Pacific denotes the warming (cooling) associated with the El Niño (La Niña) background state. The dashed black arrows at the top of the diagrams denote the eastward propagation of the MJO circulation–convection coupled system.

When focusing on the moisture–convection feedback (e.g., Sobel and Maloney 2012, 2013; Adames and Kim 2016; Jiang et al. 2018), the column-integrated moisture budget analysis showed that the modulation of the ENSO on the fast and slow MJO modes is primarily manifested as a net moistening over the equatorial eastern MC–EWP region, which can further enhance the zonal asymmetry of the moisture tendency and therefore lead to a faster eastward propagation (e.g., Hsu and Li 2012; Hsu et al. 2014; Wang et al. 2017). This net moistening is predominated by the horizontal advection, especially its meridional component. The scale decomposition revealed that the ENSO modulates the MJO propagation primarily by changing both the longitudinal and latitudinal gradients of the LF (>100 days) background moisture anomaly (refer to Figs. 8 and 12). The role of the nonlinear zonal moisture advection by the HF (<20 days) transients, possibly by the CCKWs, was also demonstrated to be nonnegligibly important in contributing to the net moistening over the equatorial eastern MC–EWP region.

This observational diagnostic analysis requires an in-depth validation from both an intercomparison of GCM modeling studies and a theoretical modeling work. As a first step, we have utilized a long-term run from a hybrid coupled GCM (HcGCM), that is, the fourth generation of ECHAM GCM (Roeckner et al. 1996) coupled with a mixed layer oceanic model (Fu and Wang 2001), to validate the observed findings represented in this study (figure not shown). The results showed that this HcGCM could well reproduce the fast MJO mode (~6.2 m s−1) during El Niño and slow MJO mode (~3.7 m s−1) during La Niña. In addition, the much stronger (weaker) leading suppressed convection and Kelvin wave easterly over the MC–EWP region for the fast (slow) MJO mode are also well simulated. These results show that the modulation of ENSO on the fast and slow MJO modes during boreal winter and the underlying mechanisms are also well captured and validated by the HcGCM.

The mechanisms proposed in this paper involve equatorial wave dynamics, moisture–convection feedback, diabatic precipitation heating, and the intraseasonal air–sea interaction. In future work, we plan to further 1) validate our conclusions by evaluating MJO simulations from the multimodel comparison project developed by the Working Group on Numerical Experimentation (WGNE) MJO Task Force and the GEWEX Atmosphere System Studies (Petch et al. 2011; Jiang et al. 2015; Wang et al. 2018) and 2) utilize the theoretical model developed by Wei et al. (2018), which can adequately consider these mechanisms, to validate the results presented in this paper.

Acknowledgments

This work was jointly supported by the National Key Research and Development Program on Monitoring, Early Warning and Prevention of Major Natural Disaster (2018YFC1506004), the National Basic Research (973) Program of China under Grant 2015CB453203, the China National Science Foundation under Grants 41775066 and 41375062, the NSFC Innovative Group Grant 41421005, and the NSFC-Shandong Joint Fund for Marine Science Research Centers U1406401. Daily interpolated OLRs were obtained from NOAA/OAR/ESRL PSD, Boulder, Colorado, https://www.esrl.noaa.gov/psd/.

REFERENCES

REFERENCES
Adames
,
Á. F.
, and
D.
Kim
,
2016
:
The MJO as a dispersive, convectively coupled moisture wave: Theory and observations
.
J. Atmos. Sci.
,
73
,
913
941
, https://doi.org/10.1175/JAS-D-15-0170.1.
Adames
,
Á. F.
,
J. M.
Wallace
, and
J. M.
Monteiro
,
2016
:
Seasonality of the structure and propagation characteristics of the MJO
.
J. Atmos. Sci.
,
73
,
3511
3526
, https://doi.org/10.1175/JAS-D-15-0232.1.
Ahn
,
M.-S.
,
D.
Kim
,
K. R.
Sperber
,
I.-S.
Kang
,
E.
Maloney
,
D.
Waliser
, and
H.
Hendon
,
2017
:
MJO simulation in CMIP5 climate models: MJO skill metrics and process-oriented diagnosis
.
Climate Dyn.
,
49
,
4023
4045
, https://doi.org/10.1007/s00382-017-3558-4.
Andersen
,
J. A.
, and
Z.
Kuang
,
2012
:
Moist static energy budget of MJO-like disturbances in the atmosphere of a zonally symmetric aquaplanet
.
J. Climate
,
25
,
2782
2804
, https://doi.org/10.1175/JCLI-D-11-00168.1.
Ashok
,
K.
,
S.
Behera
,
S.
Rao
,
H.
Weng
, and
T.
Yamagata
,
2007
:
El Niño Modoki and its possible teleconnection
.
J. Geophys. Res.
,
112
,
C11007
, https://doi.org/10.1029/2006JC003798.
Baldwin
,
M. P.
, and Coauthors
,
2001
:
The quasi-biennial oscillation
.
Rev. Geophys.
,
39
,
179
229
, https://doi.org/10.1029/1999RG000073.
Baquero-Bernal
,
A.
,
M.
Latif
, and
S.
Legutke
,
2002
:
On dipolelike variability of sea surface temperature in the tropical Indian Ocean
.
J. Climate
,
15
,
1358
1368
, https://doi.org/10.1175/1520-0442(2002)015<1358:ODVOSS>2.0.CO;2.
Chang
,
C. P.
,
1977
:
Viscous internal gravity waves and low-frequency oscillations in the tropics
.
J. Atmos. Sci.
,
34
,
901
910
, https://doi.org/10.1175/1520-0469(1977)034<0901:VIGWAL>2.0.CO;2.
Chen
,
G.
, and
B.
Wang
,
2018a
:
Effects of enhanced front Walker cell on the eastward propagation of the MJO
.
J. Climate
,
31
,
7719
7738
, https://doi.org/10.1175/JCLI-D-17-0383.1.
Chen
,
G.
, and
B.
Wang
,
2018b
:
Does the MJO have a westward group velocity?
J. Climate
,
31
,
2435
2443
, https://doi.org/10.1175/JCLI-D-17-0446.1.
Chen
,
X.
,
C.
Li
, and
Y.
Tan
,
2015
:
The influences of El Niño on MJO over the equatorial Pacific
.
J. Ocean Univ. China
,
14
,
1
8
, https://doi.org/10.1007/S11802-015-2381-Y.
Chen
,
X.
,
J.
Ling
, and
C.
Li
,
2016
:
Evolution of the Madden–Julian oscillation in two types of El Niño
.
J. Climate
,
29
,
1919
1934
, https://doi.org/10.1175/JCLI-D-15-0486.1.
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
, https://doi.org/10.1002/qj.828.
DeMott
,
C. A.
,
N. P.
Klingaman
, and
S. J.
Woolnough
,
2015
:
Atmosphere–ocean coupled processes in the Madden-Julian oscillation
.
Rev. Geophys.
,
53
,
1099
1154
, https://doi.org/10.1002/2014RG000478.
Duchon
,
C. E.
,
1979
:
Lanczos filtering in one and two dimensions
.
J. Appl. Meteor.
,
18
,
1016
1022
, https://doi.org/10.1175/1520-0450(1979)018<1016:LFIOAT>2.0.CO;2.
Feng
,
J.
,
T.
Li
, and
W.
Zhu
,
2015
:
Propagating and nonpropagating MJO events over the Maritime Continent
.
J. Climate
,
28
,
8430
8449
, https://doi.org/10.1175/JCLI-D-15-0085.1.
Freeman
,
E.
, and Coauthors
,
2017
:
ICOADS release 3.0: A major update to the historical marine climate record
.
Int. J. Climatol.
,
37
,
2211
2232
, https://doi.org/10.1002/joc.4775.
Fu
,
X.
, and
B.
Wang
,
2001
:
A coupled modeling study of the seasonal cycle of the Pacific cold tongue. Part I: Simulation and sensitivity experiments
.
J. Climate
,
14
,
765
779
, https://doi.org/10.1175/1520-0442(2001)014<0765:ACMSOT>2.0.CO;2.
Fu
,
X.
,
W.
Wang
,
H.-L.
Ren
,
X.
Jia
, and
T.
Shinoda
,
2018
:
Three different downstream fates of the boreal-summer MJOs on their passages over the Maritime Continent
.
Climate Dyn.
,
51
,
1841
1862
, https://doi.org/10.1007/s00382-017-3985-2.
Geng
,
X.
,
W.
Zhang
,
M. F.
Stuecker
, and
F. F.
Jin
,
2017
:
Strong sub-seasonal wintertime cooling over East Asia and northern Europe associated with super El Niño events
.
Sci. Rep.
,
7
,
3770
, https://doi.org/10.1038/s41598-017-03977-2.
Gill
,
A. E.
,
1980
:
Some simple solutions for heat-induced tropical circulation
.
Quart. J. Roy. Meteor. Soc.
,
106
,
447
462
, https://doi.org/10.1002/qj.49710644905.
Gonzalez
,
A. O.
, and
X.
Jiang
,
2017
:
Winter mean lower tropospheric moisture over the Maritime Continent as a climate model diagnostic metric for the propagation of the Madden-Julian oscillation
.
Geophys. Res. Lett.
,
44
,
2588
2596
, https://doi.org/10.1002/2016GL072430.
Goulet
,
L.
, and
J.-P.
Duvel
,
2000
:
A new approach to detect and characterize intermittent atmospheric oscillations: Application to the intraseasonal oscillation
.
J. Atmos. Sci.
,
57
,
2397
2416
, https://doi.org/10.1175/1520-0469(2000)057<2397:ANATDA>2.0.CO;2.
Gray
,
B. M.
,
1988
:
Seasonal frequency variation in the 40–50 day oscillation
.
J. Climatol.
,
8
,
511
519
, https://doi.org/10.1002/joc.3370080507.
Guo
,
Y.
,
D. E.
Waliser
, and
X.
Jiang
,
2015
:
A systematic relationship between the representations of convectively coupled equatorial wave activity and the Madden–Julian oscillation in climate model simulations
.
J. Climate
,
28
,
1881
1904
, https://doi.org/10.1175/JCLI-D-14-00485.1.
Han
,
W.
,
2005
:
Origins and dynamics of the 90-day and 30–60-day variations in the equatorial Indian Ocean
.
J. Phys. Oceanogr.
,
35
,
708
728
, https://doi.org/10.1175/JPO2725.1.
Hao
,
X.
,
H.-L.
Ren
,
W.
Zhang
,
M.
Liu
, and
Y.
Wei
,
2019
:
Diagnosing the spatiotemporal diversity of westerly wind events in the tropical Pacific
.
Dyn. Atmos. Oceans
,
86
,
90
103
, https://doi.org/10.1016/j.dynatmoce.2019.03.004.
Henderson
,
S. A.
, and
E. D.
Maloney
,
2018
:
The impact of the Madden-Julian oscillation on high-latitude winter blocking during El Niño–Southern Oscillation events
.
J. Climate
,
31
,
5293
5318
, https://doi.org/10.1175/JCLI-D-17-0721.1.
Hendon
,
H. H.
,
M. C.
Wheeler
, and
C.
Zhang
,
2007
:
Seasonal dependence of the MJO–ENSO relationship
.
J. Climate
,
20
,
531
543
, https://doi.org/10.1175/JCLI4003.1.
Hsu
,
P.-C.
, and
T.
Li
,
2012
:
Role of the boundary layer moisture asymmetry in causing the eastward propagation of the Madden–Julian oscillation
.
J. Climate
,
25
,
4914
4931
, https://doi.org/10.1175/JCLI-D-11-00310.1.
Hsu
,
P.-C.
, and
T.
Xiao
,
2017
:
Differences in the initiation and development of the Madden–Julian oscillation over the Indian Ocean associated with two types of El Niño
.
J. Climate
,
30
,
1397
1415
, https://doi.org/10.1175/JCLI-D-16-0336.1.
Hsu
,
P.-C.
,
T.
Li
, and
H.
Murakami
,
2014
:
Moisture asymmetry and MJO eastward propagation in an aquaplanet general circulation model
.
J. Climate
,
27
,
8747
8760
, https://doi.org/10.1175/JCLI-D-14-00148.1.
Huang
,
B.
, and Coauthors
,
2015
:
Extended reconstructed sea surface temperature version 4 (ERSST. v4). Part I: Upgrades and intercomparisons
.
J. Climate
,
28
,
911
930
, https://doi.org/10.1175/JCLI-D-14-00006.1.
Huang
,
B.
, and Coauthors
,
2016
:
Further exploring and quantifying uncertainties for extended reconstructed sea surface temperature (ERSST) version 4 (v4)
.
J. Climate
,
29
,
3119
3142
, https://doi.org/10.1175/JCLI-D-15-0430.1.
Huang
,
B.
, and Coauthors
,
2017
:
Extended reconstructed sea surface temperature, version 5 (ERSSTv5): Upgrades, validations, and intercomparisons
.
J. Climate
,
30
,
8179
8205
, https://doi.org/10.1175/JCLI-D-16-0836.1.
Hung
,
C.-S.
, and
C.-H.
Sui
,
2018
:
A diagnostic study of the evolution of the MJO from Indian Ocean to Maritime Continent: Wave dynamics versus advective moistening processes
.
J. Climate
,
31
,
4095
4115
, https://doi.org/10.1175/JCLI-D-17-0139.1.
Izumo
,
T.
, and Coauthors
,
2010
:
Low and high frequency Madden–Julian oscillations in austral summer: Interannual variations
.
Climate Dyn.
,
35
,
669
683
, https://doi.org/10.1007/s00382-009-0655-z.
Jiang
,
X.
,
2017
:
Key processes for the eastward propagation of the Madden–Julian oscillation based on multimodel simulations
.
J. Geophys. Res. Atmos.
,
122
,
755
770
, https://doi.org/10.1002/2016JD025955.
Jiang
,
X.
, and Coauthors
,
2015
:
Vertical structure and physical processes of the Madden–Julian oscillation: Exploring key model physics in climate simulations
.
J. Geophys. Res.
,
120
,
4718
4748
, https://doi.org/10.1002/2014JD022375.
Jiang
,
X.
,
Á. F.
Adames
,
M.
Zhao
, and
E.
Maloney
,
2018
:
A unified moisture mode framework for seasonality of the Madden–Julian oscillation
.
J. Climate
,
31
,
4215
4224
, https://doi.org/10.1175/JCLI-D-17-0671.1.
Kang
,
I.-S.
,
F.
Liu
,
M.-S.
Ahn
,
Y.-M.
Yang
, and
B.
Wang
,
2013
:
The role of SST structure in convectively coupled Kelvin–Rossby waves and its implications for MJO formation
.
J. Climate
,
26
,
5915
5930
, https://doi.org/10.1175/JCLI-D-12-00303.1.
Kao
,
H.-Y.
, and
J.-Y.
Yu
,
2009
:
Contrasting eastern-Pacific and central-Pacific types of ENSO
.
J. Climate
,
22
,
615
632
, https://doi.org/10.1175/2008JCLI2309.1.
Kiladis
,
G. N.
,
M. C.
Wheeler
,
P. T.
Haertel
,
K. H.
Straub
, and
P. E.
Roundy
,
2009
:
Convectively coupled equatorial waves
.
Rev. Geophys.
,
47
,
RG2003
, https://doi.org/10.1029/2008RG000266.
Kiladis
,
G. N.
, and Coauthors
,
2014
:
A comparison of OLR and circulation-based indices for tracking the MJO
.
Mon. Wea. Rev.
,
142
,
1697
1715
, https://doi.org/10.1175/MWR-D-13-00301.1.
Kim
,
D.
, and Coauthors
,
2009
:
Application of MJO simulation diagnostics to climate models
.
J. Climate
,
22
,
6413
6436
, https://doi.org/10.1175/2009JCLI3063.1.
Kim
,
D.
,
J.-S.
Kug
, and
A. H.
Sobel
,
2014
:
Propagating versus nonpropagating Madden–Julian oscillation events
.
J. Climate
,
27
,
111
125
, https://doi.org/10.1175/JCLI-D-13-00084.1.
Kim
,
D.
,
H.
Kim
, and
M. I.
Lee
,
2017
:
Why does the MJO detour the Maritime Continent during austral summer?
Geophys. Res. Lett.
,
44
,
2579
2587
, https://doi.org/10.1002/2017GL072643.
Kiranmayi
,
L.
, and
E. D.
Maloney
,
2011
:
Intraseasonal moist static energy budget in reanalysis data
.
J. Geophys. Res.
,
116
,
D21117
, https://doi.org/10.1029/2011JD016031.
Kug
,
J.-S.
,
F.-F.
Jin
, and
S.-I.
An
,
2009
:
Two types of El Niño events: Cold tongue El Niño and warm pool El Niño
.
J. Climate
,
22
,
1499
1515
, https://doi.org/10.1175/2008JCLI2624.1.
Kuo
,
H. L.
,
1965
:
On formation and intensification of tropical cyclones through latent heat release by cumulus convection
.
J. Atmos. Sci.
,
22
,
40
63
, https://doi.org/10.1175/1520-0469(1965)022<0040:OFAIOT>2.0.CO;2.
Latif
,
M.
,
V. A.
Semenov
, and
W.
Park
,
2015
:
Super El Niños in response to global warming in a climate model
.
Climatic Change
,
132
,
489
500
, https://doi.org/10.1007/s10584-015-1439-6.
Lau
,
K.-M.
, and
P.
Chan
,
1986a
:
Aspects of the 40–50 day oscillation during the northern summer as inferred from outgoing longwave radiation
.
Mon. Wea. Rev.
,
114
,
1354
1367
, https://doi.org/10.1175/1520-0493(1986)114<1354:AOTDOD>2.0.CO;2.
Lau
,
K.-M.
, and
P.
Chan
,
1986b
:
The 40–50 day oscillation and the El Niño/Southern Oscillation: A new perspective
.
Bull. Amer. Meteor. Soc.
,
67
,
533
534
, https://doi.org/10.1175/1520-0477(1986)067<0533:TDOATE>2.0.CO;2.
Lau
,
K.-M.
, and
D. E.
Waliser
,
2012
: Intraseasonal Variability in the Atmosphere–Ocean Climate System. 2nd ed. Springer, 613 pp.
Li
,
C.
, and
I.
Smith
,
1995
:
Numerical simulation of the tropical intraseasonal oscillation and the effect of warm SST
.
Acta Meteor. Sin.
,
9
,
1
12
.
Li
,
J.
, and
J.
Mao
,
2019
:
Factors controlling the interannual variation of 30-60-day boreal summer intraseasonal oscillation over the Asian summer monsoon region
.
Climate Dyn.
,
52
,
1651
1672
, https://doi.org/10.1007/s00382-018-4216-1.
Li
,
T.
,
2014
:
Recent advance in understanding the dynamics of the Madden–Julian Oscillation
.
J. Meteor. Res.
,
28
,
1
33
, https://doi.org/10.1007/S13351-014-3087-6.
Li
,
T.
,
F.
Tam
,
X.
Fu
,
T.
Zhou
, and
W.
Zhu
,
2008
:
Causes of the intraseasonal SST variability in the tropical Indian Ocean
.
Atmos. Oceanic Sci. Lett.
,
1
,
18
23
, https://doi.org/10.1080/16742834.2008.11446758.
Liebmann
,
B.
, and
C. A.
Smith
,
1996
:
Description of a complete (interpolated) outgoing longwave radiation dataset
.
Bull. Amer. Meteor. Soc.
,
77
,
1275
1277
.
Ling
,
J.
,
C.
Zhang
, and
P.
Bechtold
,
2013
:
Large-scale distinctions between MJO and non-MJO convective initiation over the tropical Indian Ocean
.
J. Atmos. Sci.
,
70
,
2696
2712
, https://doi.org/10.1175/JAS-D-13-029.1.
Ling
,
J.
,
C.
Zhang
,
S.
Wang
, and
C.
Li
,
2017
:
A new interpretation of the ability of global models to simulate the MJO
.
Geophys. Res. Lett.
,
44
,
5798
5806
, https://doi.org/10.1002/2017GL073891.
Liu
,
C.
,
B.
Tian
,
K.-F.
Li
,
G. L.
Manney
,
N. J.
Livesey
Y. L.
Yung
, and
D. E.
Waliser
,
2014
:
Northern Hemisphere mid-winter vortex-displacement and vortex-split stratospheric sudden warmings: Influence of the Madden–Julian oscillation and quasi-biennial oscillation
.
J. Geophys. Res.
,
119
,
12 599
12 620
, https://doi.org/10.1002/2014JD021876.
Liu
,
F.
, and
B.
Wang
,
2013
:
An air–sea coupled skeleton model for the Madden–Julian Oscillation
.
J. Atmos. Sci.
,
70
,
3147
3156
, https://doi.org/10.1175/JAS-D-12-0348.1.
Liu
,
F.
, and
B.
Wang
,
2016
:
Role of horizontal advection of seasonal-mean moisture in in Madden–Julian oscillation: A theoretical model analysis
.
J. Climate
,
29
,
6277
6293
, https://doi.org/10.1175/JCLI-D-16-0078.1.
Liu
,
F.
, and
B.
Wang
,
2017a
:
Effects of moisture feedback in a frictional coupled Kelvin-Rossby wave model and implication in the Madden–Julian oscillation dynamics
.
Climate Dyn.
,
48
,
513
522
, https://doi.org/10.1007/s00382-016-3090-y.
Liu
,
F.
, and
B.
Wang
,
2017b
:
Roles of the moisture and wave feedbacks in shaping the Madden–Julian oscillation
.
J. Climate
,
30
,
10 275
10 291
, https://doi.org/10.1175/JCLI-D-17-0003.1.
Liu
,
F.
,
T.
Li
,
H.
Wang
,
L.
Deng
, and
Y.
Zhang
,
2016
:
Modulation of boreal summer intraseasonal oscillations over the western North Pacific by ENSO
.
J. Climate
,
29
,
7189
7201
, https://doi.org/10.1175/JCLI-D-15-0831.1.
Lorenz
,
D. J.
, and
D. L.
Hartmann
,
2006
:
The effect of the MJO on the North American monsoon
.
J. Climate
,
19
,
333
343
, https://doi.org/10.1175/JCLI3684.1.
Madden
,
R. A.
, and
P. R.
Julian
,
1971
:
Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific
.
J. Atmos. Sci.
,
28
,
702
708
, https://doi.org/10.1175/1520-0469(1971)028<0702:DOADOI>2.0.CO;2.
Madden
,
R. A.
, and
P. R.
Julian
,
1972
:
Description of global-scale circulation cells in the tropics with a 40–50 day period
.
J. Atmos. Sci.
,
29
,
1109
1123
, https://doi.org/10.1175/1520-0469(1972)029<1109:DOGSCC>2.0.CO;2.
Madden
,
R. A.
, and
P. R.
Julian
,
1994
:
Observations of the 40–50-day tropical oscillation—A review
.
Mon. Wea. Rev.
,
122
,
814
837
, https://doi.org/10.1175/1520-0493(1994)122<0814:OOTDTO>2.0.CO;2.
Maloney
,
E. D.
,
2009
:
The moist static energy budget of a composite tropical intraseasonal oscillation in a climate model
.
J. Climate
,
22
,
711
729
, https://doi.org/10.1175/2008JCLI2542.1.
Maloney
,
E. D.
, and
D. L.
Hartmann
,
2000
:
Modulation of hurricane activity in the Gulf of Mexico by the Madden–Julian oscillation
.
Science
,
287
,
2002
2004
, https://doi.org/10.1126/science.287.5460.2002.
Marshall
,
A. G.
,
H. H.
Hendon
,
S. W.
Son
, and
Y.
Lim
,
2017
:
Impact of the quasi-biennial oscillation on predictability of the Madden–Julian oscillation
.
Climate Dyn.
,
49
,
1365
1377
, https://doi.org/10.1007/s00382-016-3392-0.
Matthews
,
A. J.
,
2008
:
Primary and successive events in the Madden–Julian Oscillation
.
Quart. J. Roy. Meteor. Soc.
,
134
,
439
453
, https://doi.org/10.1002/qj.224.
McPhaden
,
M. J.
,
1999
:
Genesis and evolution of the 1997-98 El Niño
.
Science
,
283
,
950
954
, https://doi.org/10.1126/science.283.5404.950.
Moon
,
J.-Y.
,
B.
Wang
, and
K.-J.
Ha
,
2011
:
ENSO regulation of MJO teleconnection
.
Climate Dyn.
,
37
,
1133
1149
, https://doi.org/10.1007/s00382-010-0902-3.
Neena
,
J. M.
,
J. Y.
Lee
,
D.
Waliser
,
B.
Wang
, and
X.
Jiang
,
2014
:
Predictability of the Madden–Julian oscillation in the Intraseasonal Variability Hindcast Experiment (ISVHE)
.
J. Climate
,
27
,
4531
4543
, https://doi.org/10.1175/JCLI-D-13-00624.1.
Nishimoto
,
E.
, and
S.
Yoden
,
2017
:
Influence of the stratospheric quasi-biennial oscillation on the Madden–Julian oscillation during austral summer
.
J. Atmos. Sci.
,
74
,
1105
1125
, https://doi.org/10.1175/JAS-D-16-0205.1.
Petch
,
J.
,
D.
Waliser
,
X.
Jiang
,
P.
Xavier
, and
S.
Woolnough
,
2011
: A global model intercomparison of the physical processes associated with the Madden–Julian oscillation. GEWEX News, No. 21, International GEWEX Project Office, Silver Spring, MD, 3–5.
Pillai
,
P. A.
, and
J. S.
Chowdary
,
2016
:
Indian summer monsoon intra-seasonal oscillation associated with the developing and decaying phase of El Niño
.
Int. J. Climatol.
,
36
,
1846
1862
, https://doi.org/10.1002/joc.4464.
Pohl
,
B.
, and
A. J.
Matthews
,
2007
:
Observed changes in the lifetime and amplitude of the Madden–Julian oscillation associated with interannual ENSO sea surface temperature anomalies
.
J. Climate
,
20
,
2659
2674
, https://doi.org/10.1175/JCLI4230.1.
Rao
,
J.
, and
R.
Ren
,
2016
:
Asymmetry and nonlinearity of the influence of ENSO on the northern winter stratosphere: 1. Observations
.
J. Geophys. Res. Atmos.
,
121
,
9000
9016
, https://doi.org/10.1002/2015JD024520.
Ren
,
H.-L.
, and
F.-F.
Jin
,
2011
:
Niño indices for two types of ENSO
.
Geophys. Res. Lett.
,
38
,
L04704
, https://doi.org/10.1029/2010GL046031.
Ren
,
H.-L.
, and
P.
Ren
,
2017
:
Impact of Madden–Julian oscillation upon winter extreme rainfall in southern China: Observations and predictability in CFSv2
.
Atmosphere
,
8
,
192
, https://doi.org/10.3390/atmos8100192.
Ren
,
H.-L.
, and Coauthors
,
2019
:
Seasonal predictability of winter ENSO types in operational dynamical model predictions
.
Climate Dyn.
,
52
,
3869
3890
, https://doi.org/10.1007/S00382-018-4366-1.
Ren
,
P.
,
H.-L.
Ren
,
X.
Fu
,
J.
Wu
, and
L.
Du
,
2018
:
Impact of boreal summer intraseasonal oscillation on rainfall extremes in southeastern China and its predictability in CFSv2
.
J. Geophys. Res. Atmos.
,
123
,
4423
4442
, https://doi.org/10.1029/2017JD028043.
Roeckner
,
E.
, and Coauthors
,
1996
: The atmospheric general circulation model ECHAM-4: Model description and simulation of present-day climate. Max Planck Institute for Meteorology Rep. 218, 90 pp.
Roundy
,
P. E.
,
2008
:
Analysis of convectively coupled Kelvin waves in the Indian Ocean MJO
.
J. Atmos. Sci.
,
65
,
1342
1359
, https://doi.org/10.1175/2007JAS2345.1.
Saji
,
N. H.
,
B. N.
Goswami
,
P. N.
Vinayachandran
, and
T.
Yamagata
,
1999
:
A dipole mode in the tropical Indian Ocean
.
Nature
,
401
,
360
363
, https://doi.org/10.1038/43854.
Saji
,
N. H.
,
S. P.
Xie
, and
C. Y.
Tam
,
2006
:
Satellite observations of intense intraseasonal cooling events in the tropical south Indian Ocean
.
Geophys. Res. Lett.
,
33
,
L14704
, https://doi.org/10.1029/2006GL026525.
Seo
,
K. H.
, and
A.
Kumar
,
2008
:
The onset and life span of the Madden–Julian oscillation
.
Theor. Appl. Climatol.
,
94
,
13
24
, https://doi.org/10.1007/s00704-007-0340-2.
Shinoda
,
T.
, and
W.
Han
,
2005
:
Influence of the Indian Ocean dipole on atmospheric subseasonal variability
.
J. Climate
,
18
,
3891
3909
, https://doi.org/10.1175/JCLI3510.1.
Shinoda
,
T.
,
M. A.
Alexander
, and
H. H.
Hendon
,
2004
:
Remote response of the Indian Ocean to interannual SST variations in the tropical Pacific
.
J. Climate
,
17
,
362
372
, https://doi.org/10.1175/1520-0442(2004)017<0362:RROTIO>2.0.CO;2.
Slingo
,
J.
,
D.
Rowell
,
K.
Sperber
, and
F.
Nortley
,
1999
:
On the predictability of the interannual behaviour of the Madden–Julian oscillation and its relationship with El Niño
.
Quart. J. Roy. Meteor. Soc.
,
125
,
583
609
, https://doi.org/10.1002/qj.49712555411.
Sobel
,
A.
, and
E.
Maloney
,
2012
:
An idealized semi-empirical framework for modeling the Madden–Julian oscillation
.
J. Atmos. Sci.
,
69
,
1691
1705
, https://doi.org/10.1175/JAS-D-11-0118.1.
Sobel
,
A.
, and
E.
Maloney
,
2013
:
Moisture modes and the eastward propagation of the MJO
.
J. Atmos. Sci.
,
70
,
187
192
, https://doi.org/10.1175/JAS-D-12-0189.1.
Sobel
,
A.
,
J.
Nilsson
, and
L. M.
Polvani
,
2001
:
The weak temperature gradient approximation and balanced tropical moisture waves
.
J. Atmos. Sci.
,
58
,
3650
3665
, https://doi.org/10.1175/1520-0469(2001)058<3650:TWTGAA>2.0.CO;2.
Sobel
,
A.
,
S.
Wang
, and
D.
Kim
,
2014
:
Moist static energy budget of the MJO during DYNAMO
.
J. Atmos. Sci.
,
71
,
4276
4291
, https://doi.org/10.1175/JAS-D-14-0052.1.
Son
,
S. W.
, and Coauthors
,
2017
:
Stratospheric control of the Madden–Julian oscillation
.
J. Climate
,
30
,
1909
1922
, https://doi.org/10.1175/JCLI-D-16-0620.1.
Straub
,
K. H.
,
2013
:
MJO initiation in the real-time multivariate MJO index
.
J. Climate
,
26
,
1130
1151
, https://doi.org/10.1175/JCLI-D-12-00074.1.
Suematsu
,
T.
, and
H.
Miura
,
2018
:
Zonal SST differences as a potential environmental factor supporting the longevity of the Madden–Julian oscillation
.
J. Climate
,
31
,
7549
7564
, https://doi.org/10.1175/JCLI-D-17-0822.1.
Sun
,
L.
,
H.
Wang
, and
F.
Liu
,
2019
:
Combined effect of the QBO and ENSO on the MJO
.
Atmos. Oceanic Sci. Lett.
,
12
,
170
176
, https://doi.org/10.1080/16742834.2019.1588064.
Timmermann
,
A.
, and Coauthors
,
2018
:
El Niño–Southern Oscillation complexity
.
Nature
,
559
,
535
545
, https://doi.org/10.1038/s41586-018-0252-6.
Titchner
,
H. A.
, and
N. A.
Rayner
,
2014
:
The Met Office Hadley Centre sea ice and sea surface temperature data set, version 2: 1. Sea ice concentrations
.
J. Geophys. Res.
,
119
,
2864
2889
, https://doi.org/10.1002/2013JD020316.
Wang
,
B.
,
1988
:
Dynamics of tropical low-frequency waves: An analysis of the moist Kelvin wave
.
J. Atmos. Sci.
,
45
,
2051
2065
, https://doi.org/10.1175/1520-0469(1988)045<2051:DOTLFW>2.0.CO;2.
Wang
,
B.
, and
H.
Rui
,
1990
:
Dynamics of the coupled moist Kelvin–Rossby wave on an equatorial β-plane
.
J. Atmos. Sci.
,
47
,
397
413
, https://doi.org/10.1175/1520-0469(1990)047<0397:DOTCMK>2.0.CO;2.
Wang
,
B.
, and
X.
Xie
,
1998
:
Coupled modes of the warm pool climate system. Part I: The role of air–sea interaction in maintaining Madden–Julian oscillation
.
J. Climate
,
11
,
2116
2135
, https://doi.org/10.1175/1520-0442-11.8.2116.
Wang
,
B.
, and
G.
Chen
,
2016
:
A general theoretical framework for understanding essential dynamics of Madden–Julian oscillation
.
Climate Dyn.
,
49
,
2309
2238
, https://doi.org/10.1007/s00382-016-3448-1.
Wang
,
B.
, and
S.-S.
Lee
,
2017
:
MJO propagation shaped by zonal asymmetric structures: Results from 24 GCM simulations
.
J. Climate
,
30
,
7933
7952
, https://doi.org/10.1175/JCLI-D-16-0873.1.
Wang
,
B.
,
T.
Li
, and
P.
Chang
,
1995
:
An intermediate model of the tropical Pacific Ocean
.
J. Phys. Oceanogr.
,
25
,
1599
1616
, https://doi.org/10.1175/1520-0485(1995)025<1599:AIMOTT>2.0.CO;2.
Wang
,
B.
,
F.
Liu
, and
G.
Chen
,
2016
:
A trio-interaction theory for Madden–Julian oscillation
.
Geosci. Lett.
,
3
,
34
, https://doi.org/10.1186/s40562-016-0066-z.
Wang
,
B.
, and Coauthors
,
2018
:
Dynamics-oriented diagnostics for the Madden–Julian oscillation
.
J. Climate
,
31
,
3117
3135
, https://doi.org/10.1175/JCLI-D-17-0332.1.
Wang
,
B.
,
G.
Chen
, and
F.
Liu
,
2019
:
Diversity of the Madden–Julian oscillation
.
Sci. Adv.
,
5
,
eaax0220
,https://doi.org/10.1126/SCIADV.AAX0220.
Wang
,
L.
,
T.
Li
,
E.
Maloney
, and
B.
Wang
,
2017
:
Fundamental causes of propagating and nonpropagating MJOs in MJOTF/GASS models
.
J. Climate
,
30
,
3743
3769
, https://doi.org/10.1175/JCLI-D-16-0765.1.
Wang
,
L.
,
T.
Li
,
L.
Chen
,
S. K.
Behera
, and
T.
Nasuno
,
2018a
:
Modulation of the MJO intensity over the equatorial western Pacific by two types of El Niño
.
Climate Dyn.
,
51
,
687
700
, https://doi.org/10.1007/s00382-017-3949-6.
Wang
,
L.
,
T.
Li
, and
T.
Nasuno
,
2018b
:
Impact of Rossby and Kelvin wave components on MJO eastward propagation
.
J. Climate
,
31
,
6913
6931
, https://doi.org/10.1175/JCLI-D-17-0749.1.
Wei
,
Y.
,
F.
Liu
,
H.-L.
Ren
, and
M.
Mu
,
2018
:
Planetary-scale selection of the Madden–Julian oscillation in an air–sea coupled dynamical moisture model
.
Climate Dyn.
,
50
,
3441
3456
, https://doi.org/10.1007/s00382-017-3816-5.
Wei
,
Y.
,
M.
Mu
,
H.-L.
Ren
, and
J.-X.
Fu
,
2019
:
Conditional nonlinear optimal perturbations of moisture triggering primary MJO initiation
.
Geophys. Res. Lett.
,
46
,
3492
3501
, https://doi.org/10.1029/2018GL081755.
Weickmann
,
K.
,
1991
:
El Niño/Southern Oscillation and Madden–Julian (30–60 day) oscillations during 1981–1982
.
J. Geophys. Res.
,
96
,
3187
3195
, https://doi.org/10.1029/90JD01832.
Wheeler
,
M.
, and
G. N.
Kiladis
,
1999
:
Convectively coupled equatorial waves: Analysis of clouds and temperature in the wavenumber–frequency domain
.
J. Atmos. Sci.
,
56
,
374
399
, https://doi.org/10.1175/1520-0469(1999)056<0374:CCEWAO>2.0.CO;2.
Wheeler
,
M.
, and
H. H.
Hendon
,
2004
:
An all-season real-time multivariate MJO index: Development of an index for monitoring and prediction
.
Mon. Wea. Rev.
,
132
,
1917
1932
, https://doi.org/10.1175/1520-0493(2004)132<1917:AARMMI>2.0.CO;2.
Wilson
,
E. A.
,
A. L.
Gordon
, and
D.
Kim
,
2013
:
Observations of the Madden–Julian oscillation during Indian Ocean dipole events
.
J. Geophys. Res. Atmos.
,
118
,
2588
2599
, https://doi.org/10.1002/JGRD.50241.
Wu
,
R.
, and
L.
Song
,
2018
:
Spatiotemporal change of intraseasonal oscillation intensity over the tropical Indo-Pacific Ocean associated with El Niño and La Niña events
.
Climate Dyn.
,
50
,
1221
1242
, https://doi.org/10.1007/s00382-017-3675-0.
Yadav
,
P.
, and
D. M.
Straus
,
2017
:
Circulation response to fast and slow MJO episodes
.
Mon. Wea. Rev.
,
145
,
1577
1596
, https://doi.org/10.1175/MWR-D-16-0352.1.
Yanai
,
M.
,
S.
Esbensen
, and
J.-H.
Chu
,
1973
:
Determination of bulk properties of tropical cloud cluster from large-scale heat and moisture budgets
.
J. Atmos. Sci.
,
30
,
611
627
, https://doi.org/10.1175/1520-0469(1973)030<0611:DOBPOT>2.0.CO;2.
Yong
,
Y.
, and
J.
Mao
,
2016
:
Mechanistic analysis of the suppressed convective anomaly precursor associated with the initiation of primary MJO events over the tropical Indian Ocean
.
Climate Dyn.
,
46
,
779
795
, https://doi.org/10.1007/s00382-015-2612-3.
Yoo
,
C.
, and
S. W.
Son
,
2016
:
Modulation of the boreal wintertime Madden–Julian oscillation by the stratospheric quasi-biennial oscillation
.
Geophys. Res. Lett.
,
43
,
1392
1398
, https://doi.org/10.1002/2016GL067762.
Yuan
,
Y.
,
C.
Li
, and
J.
Ling
,
2015
:
Different MJO activities between EP El Niño and CP El Niño (in Chinese)
.
Sci. Sin. Terr.
,
45
,
318
334
.
Zhang
,
C.
,
2005
:
Madden–Julian Oscillation
.
Rev. Geophys.
,
43
,
RG2003
, https://doi.org/10.1029/2004RG000158.
Zhang
,
C.
,
2013
:
Madden–Julian oscillation: Bridging weather and climate
.
Bull. Amer. Meteor. Soc.
,
94
,
1849
1870
, https://doi.org/10.1175/BAMS-D-12-00026.1.
Zhang
,
C.
, and
J.
Gottschalck
,
2002
:
SST anomalies of ENSO and the Madden–Julian oscillation in the equatorial Pacific
.
J. Climate
,
15
,
2429
2445
, https://doi.org/10.1175/1520-0442(2002)015<2429:SAOEAT>2.0.CO;2.
Zhang
,
C.
, and
J.
Ling
,
2017
:
Barrier effect of the Indo-Pacific Maritime Continent on the MJO: Perspectives from tracking MJO precipitation
.
J. Climate
,
30
,
3439
3459
, https://doi.org/10.1175/JCLI-D-16-0614.1.
Zhang
,
C.
, and
B.
Zhang
,
2018
:
QBO–MJO connection
.
J. Geophys. Res.
,
123
,
2957
2967
, https://doi.org/10.1002/2017JD028171.

Footnotes

© 2019 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).