1. Introduction
The upper troposphere–lower stratosphere (UTLS) is a unique region of the atmosphere with combined properties of both the troposphere and the stratosphere. The transitional region between the upper troposphere and lower stratosphere in the tropics [referred to as tropical tropopause layer (TTL)], acts as a major vertical pathway for tropospheric air to enter the stratosphere (due to very strong vertical gradients in many chemical constituents), affecting the composition of the entire stratosphere (Fueglistaler et al. 2009a). Furthermore, variations in the radiatively active trace gases (i.e., ozone and water vapor) in the UTLS have a large impact on radiative forcing, influencing surface climate (Lacis et al. 1990; Riese et al. 2012)
Atmospheric and oceanic oscillations have a significant impact on the dynamics and composition of the UTLS (Randel and Wu 2015; Fueglistaler et al. 2009a). These oscillations vary on multiple time scales ranging from a few days to many years. The intraseasonal (20–90 day) variability of the tropical UTLS is dominated by the Madden–Julian oscillation (MJO; Madden and Julian 1994, 1972). Since the discovery of the MJO, this oscillation has been a subject of much interest due to its strong connection between weather and climate variations. Despite the significant impacts of the MJO on the global climate system, general circulation models/chemistry climate models (GCMs/CCMs) generally struggle to accurately simulate MJO variability (Kim et al. 2009; Lin et al. 2006). This can affect how models simulate trace gas distribution and variability in the UTLS on intraseasonal time scales. Thus, precise knowledge of variations in temperature, circulation and composition due to natural processes like the MJO in this key region of the atmosphere is important for model evaluation and in understanding future projections.
The MJO is identified as a slow eastward propagation of anomalous tropical deep convection, winds, and surface pressure from a region over the equatorial Indian Ocean, across the Maritime Continent, and into the western Pacific warm pool with an average period of 45 days. These MJO induced anomalies span the equatorial region but are the strongest in the warm pool of the Indian and western Pacific oceans (Hendon and Salby 1994; Kiladis et al. 2001; Rui and Wang 1990). The slow eastward propagation of many tropospheric fields is more strongly exhibited during the boreal winter months [November–February (NDJF)] than during the boreal summer months [June–September (JJAS)]. The cause of the MJO is still uncertain but generally believed to be related to the coupling of convective heating and Kelvin and Rossby wave–like dynamical responses near the equator (Zhang 2005; Adames and Kim 2016; Chen and Wang 2019).
As the MJO propagates from a region over the equatorial Indian Ocean, across the Maritime Continent, and into the western Pacific warm pool, the perturbations in temperature and circulation, associated with vertically propagating equatorial Rossby and Kelvin waves, extend upward into the UTLS and propagate throughout the tropical belt (Kiladis et al. 2001; Virts and Wallace 2014; Adames and Wallace 2014). When discussing equatorial Rossby and Kelvin waves in the UTLS, it is important to emphasize the differences between “free” and “forced” modes of these waves (Andrews et al. 1987). While both modes are generated from time varying convective heating, vertically propagating forced Rossby and Kelvin waves are coupled to the convection whereas much faster free modes propagate independently of convective heating (Andrews et al. 1987; Randel and Wu 2005; Lavender and Matthews 2009). Previous studies often refer to the signal in temperature and circulation above convective heating as Rossby and/or Kelvin wave response. Note that this is a “forced” response since anomalies propagate along with the heating source. For instance, subtropical cyclones and anticyclones associated with MJO in the UTLS (also known as Rossby gyres) are a forced response to the heating because Rossby gyres move eastward along with the MJO convection.
Convectively coupled Kelvin waves were described in detail by Takayabu (1994), Wheeler et al. (2000), Kiladis et al. (2001), and Ryu et al. (2008). These forced waves have typical periods of 5–10 days, zonal wavenumbers 3–6, and eastward phase propagation of ~15 m s−1. Ryu et al. (2008) showed and provided an explanation for why the forced Kelvin wave response is greatest in the western and central Pacific, and why vertically propagating Kelvin waves play a pivotal role for tropical tropopause undulations.
As discussed in previous studies, free Kelvin waves in the UTLS are characterized by periods of 10–20 days, global longitudinal structure (zonal waves 1–2), and phase speeds of 20–30 m s−1 (Chang 1976; Randel and Wu 2005; Wallace and Kousky 1968). To some extent, a temperature response related to the free Kelvin wave can be seen as a narrow belt of equatorial cold anomalies near the tropical tropopause layer, “Kelvin wave front” (see section 3c for more details), that propagates rapidly eastward and encircles the globe by the time that MJO convection reaches the western Pacific Ocean (Virts and Wallace 2014; Heckley and Gill 1984; Ryu et al. 2008).
Previous studies showed that the MJO and the associated anomalies in circulation can significantly impact the composition of the UTLS. The Total Ozone Mapping Spectrometer (TOMS) total ozone was the first dataset to provide evidence for 30–50-day variability above the southeast Pacific and southern Indian Oceans (Sabutis et al. 1987). A few years later Gao and Stanford (1990) found 1–2 months signal in 8-yr TOMS ozone data. Using ozonesonde data over Indonesia, Fujiwara et al. (1998) suggested that the upper-tropospheric ozone changes were tied to the propagation of forced Kelvin waves and the MJO. Tian et al. (2007) used a merged ozone dataset (MOD) developed by Stolarski and Frith (2006) to show that large subtropical negative (positive) variations in total ozone, collocated with subtropical upper-troposphere anticyclones (cyclones), were generated by the MJO and flanked to the west of the equatorial enhanced (suppressed) convection. These MJO induced extratropical perturbations in total ozone were comparable in magnitude to those of annual and interannual time scales; however, they reported only small equatorial total ozone anomalies. Using vertical profiles of ozone from Aura Microwave Limb Sounder (MLS) and Tropospheric Emission Spectrometer (TES), Li et al. (2012) confirmed the hypothesis of Tian et al. (2007) that the subtropical total ozone variations associated with the MJO are mostly from the ozone anomalies in the lower stratosphere. More recently, Virts and Wallace (2014) used observational datasets to examine MJO-related variations in the temperatures, circulation, clouds, and trace gases at the TTL. They showed that as MJO convection dissipates over the central Pacific Ocean, a fast-moving free Kelvin wave flanked by Rossby waves propagate eastward across South America, and Africa into the western Indian Ocean. These give rise to nearly zonally symmetric equatorial cold temperature anomalies, increased upwelling and associated low ozone mixing ratios in the TTL.
While the MJO impact on the UTLS temperature, circulation, and composition has been well discussed in the literature, many of the previous studies focused either on tropospheric circulation (Knutson et al. 1986; Knutson and Weickmann 1987), boreal winter MJO (Kiladis et al. 2001; Li et al. 2013; Schwartz et al. 2008; Weickmann et al. 1985) or used full time series (Virts and Wallace 2010, 2014). This study compliments and extends the analyses of the referenced above literature by emphasizing and explaining the differences in the UTLS circulation and trace gas anomalies associated with the MJO during boreal winter (NDJF) and summer (JJAS) months separately. We demonstrate that strong seasonal differences in subseasonal variability of the UTLS temperature and ozone can be explained in terms of seasonal variations in vertically propagating Kelvin waves that strongly depend on the zonal structure of the climatological zonal winds. In the next section we briefly describe the data and methodology. Seasonal differences in the MJO propagation are described in section 3a. In section 3b we examine the eastward propagation of the MJO-induced anomalies in 100 hPa temperature, horizontal winds, and upward mass flux during boreal winter and summer separately. The mechanisms responsible for seasonal variations in the UTLS temperature and circulation are discussed in section 3c while section 3d shows a trace gas response to the anomalous circulation. Discussions and conclusions are presented in section 4.
2. Data and methodology
We examine the response of the tropical and subtropical UTLS region to the MJO using temperature and trace gases [ozone (O3), carbon monoxide (CO), and water vapor (H2O)] from MLS observations on board the Aura satellite and meteorological fields from the Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2).
Version 4.2 MLS level 2 profiles (from 261 to 30 hPa) were used to construct a 5-day-averaged temperature and trace gas dataset from August 2004 through December 2018 (Livesey et al. 2015). We used the recommended quality and convergence threshold to bin the data into a 5° latitude × 5° longitude grid. Between 250 and 30 hPa, the MLS data are outputed on 12 pressure levels (261, 215, 178, 147, 121, 100, 83, 68, 56, 46, 38, and 32 hPa) and the MLS vertical resolution was retained. Sufficient vertical resolution is crucial for resolving vertically propagating Kelvin waves due to their short vertical wavelength (which is about 6–10 km). The vertical resolutions of temperature and O3 is ~3 km, while for CO is ~4 km in the pressure range used in this analysis. The vertical resolution of H2O ranges from ~1.5 km in the upper troposphere to ~3 km in the lower stratosphere. Although data are reported on the same pressure grids, their vertical resolution is not the same because the number of independent measurements that the instrument makes varies in the different species (Livesey et al. 2015). 100-hPa vertical level was used to show temperature and trace gas response to the MJO near the TTL.
Fields of other atmospheric variables (zonal and meridional winds U and V, and total diabatic heating rates Q) are from MERRA-2. MERRA-2 is the latest atmospheric reanalysis of the modern satellite era produced by NASA’s Global Modeling and Assimilation Office (Gelaro et al. 2017; Molod et al. 2015). The grid of MERRA-2 dataset is 0.625° longitude × 0.5° latitude × 72 model hybrid-sigma layers. Data, interpolated to standard pressure levels, is used in this study. MERRA-2 is outputted on fewer vertical levels than MLS between 250 and 30 hPa (250, 200, 150, 100, 70, 50, 40, and 30 hPa). MERRA-2 data is available at 3-hourly resolution but the analysis is done using 5-day (pentad) averages to be consistent with MLS.
Total diabatic heating rates are used as a proxy for vertical velocity. Ascent in the TTL and lower stratosphere induces adiabatic cooling that is balanced by radiative heating, as the air undergoes relaxation toward its radiative equilibrium temperature (Andrews et al. 1987; Fueglistaler et al. 2009b). Previous studies by Fueglistaler et al. (2009b) and Yang et al. (2008) recognized the uncertainties in the dynamically simulated vertical velocities derived from the reanalysis while Virts and Wallace (2014) showed a much stronger correspondence of trace gases and temperature with diabatic heating than with vertical velocity in the TTL.
This study also uses NOAA’s daily values of outgoing longwave radiation (OLR) as a proxy for deep convection, interpolated into 2.5° latitude × 2.5° longitude global grid (Liebmann and Smith 1996; https://www.esrl.noaa.gov/psd/).
The indices for examining MJO influence on circulation and composition is based on multivariate EOF analysis of daily 850-hPa zonal winds, 200-hPa zonal winds, and OLR (15°N–15°S) during the MLS period (Wheeler and Hendon 2004; http://www.bom.gov.au/climate/mjo/). Real-time multivariate MJO (RMM) indices by Wheeler and Hendon (2004) has been widely used to quantify the evolution and strength of the MJO. The leading two EOFs generally appear as a pair, which taken together describe the large-scale eastward-propagating signal attributed to the MJO (See Fig. 1 and section 3a). To match the temporal resolution of MLS dataset, daily RMM indices were 5-day averaged.
Phase space defined by two components of real-time multivariate MJO index (RMM1 and RMM2) for pentads (5-day averages) in (a) NDJF and (b) JJAS from January 2005 to December 2018. Regions inside the red unit circle represent weak MJO activity. Labeled are the approximate locations of the enhanced convection associated with the MJO (e.g., the “Indian Ocean” for phases 2 and 3).
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
Phase space defined by two components of real-time multivariate MJO index (RMM1 and RMM2) for pentads (5-day averages) in (a) NDJF and (b) JJAS from January 2005 to December 2018. Regions inside the red unit circle represent weak MJO activity. Labeled are the approximate locations of the enhanced convection associated with the MJO (e.g., the “Indian Ocean” for phases 2 and 3).
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
Phase space defined by two components of real-time multivariate MJO index (RMM1 and RMM2) for pentads (5-day averages) in (a) NDJF and (b) JJAS from January 2005 to December 2018. Regions inside the red unit circle represent weak MJO activity. Labeled are the approximate locations of the enhanced convection associated with the MJO (e.g., the “Indian Ocean” for phases 2 and 3).
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
All variables were bandpass-filtered at 20–90 days to isolate intraseasonal component and remove seasonal cycle, ENSO, and short-term fluctuations prior to linearly regressing them against RMM1, RMM2, and their linear combinations (i.e., “RMM1 − RMM2” and “RMM1 + RMM2”). This approach allows to observe time evolution of the MJO influence on the UTLS as deep convection propagates eastward. The opposite phase combinations of RMM index are exactly the same except of the opposite sign. As an alternative to this approach, previous studies performed composites around eight MJO phases, which captures better nonlinear behavior between positive and negative perturbations. In spite of a small loss of nonlinearity, regression analysis shows very similar results to a compositing technique and also allows taking the strength of the MJO events into consideration.
The statistical significance of regression is estimated using the Student’s t test at 95% confidence level with temporal autocorrelation taken into account when computing degrees of freedom. In this study we perform analysis for boreal summer months (JJAS), when Asian summer monsoon circulation dominates background flow, and boreal winter (NDJF) months separately.
3. Results
a. The MJO propagation during boreal winter and summer
A well-known and documented characteristic of the MJO is a strong seasonal variation in the strength, frequency of occurrence and geographical locations of associated deep convection (Gutzler and Madden 1989; Madden 1986; Salby and Hendon 1994; Zhang and Dong 2004). This can be seen in Fig. 1 where RMM1 and RMM2 define the phase space for boreal winter and summer separately from 2005 to 2018. An evolution of the MJO appears as a counterclockwise rotation when represented on this chart, propagating eastward around the globe [e.g., the MJO is initiated during phase 1 over the eastern coast of Africa and then moves across Indian Ocean (phases 2 and 3), Maritime continents (phases 4 and 5), and western Pacific Ocean (phases 6 and 7)]. Although, MJO tends to be stronger and more frequent during boreal winters (as there are more points outside of the unit circle), summertime MJO also has plenty of strong or moderate in strength events.
Eastward migration of the MJO activity and its geographical location can be better seen in Fig. 2 where filtered OLR anomalies are regressed onto MJO indices (RMM1, RMM2, and their linear combinations; RMM1 − RMM2, RMM1, RMM1 + RMM2, and RMM2 are separated by 1/8 of the cycle). During both summer and winter seasons, regions of deep convection (negative values of OLR anomalies) propagate eastward from the Indian Ocean to the western Pacific Ocean. For instance, during the peak in RMM1 (Fig. 2b) enhanced convection is observed over the Indian Ocean and suppressed convection is over the western Pacific, while by the time of the peak in RMM2 (Fig. 2d) enhanced convection is over the western Pacific. Figure 2 also shows differences in the location of the MJO-related deep convection between summer and winter seasons. During NDJF the Indo-Pacific warm pool is centered near the equator and the MJO convection is more symmetric with respect to the equator. The eastward propagation of deep convection is accompanied by a southward shift toward the South Pacific convergence zone. During JJAS, when the large-scale circulation during JJAS is dominated by the Asian summer monsoon, the MJO related features in OLR are more elongated and displaced northward. Previous studies related seasonal displacements of the deep convection toward the summer hemisphere to the intraseasonal variations in the Asian summer monsoon and the seasonal migration of the ITCZ (Zhang and Dong 2004; Lawrence and Webster 2002).
The 20–90-day bandpass-filtered OLR for 2005–18 regressed onto (a) RMM1 − RMM2, (b) RMM1, (c) RMM1 + RMM2, and (d) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Hatching indicates sensitivities (in uits of W m−2 per standard deviation of the MJO index) that are statistically significant at the 95% confidence level using the Student’s t test. The contour interval (CI) is 2 W m−2. The zero contour is not shown.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
The 20–90-day bandpass-filtered OLR for 2005–18 regressed onto (a) RMM1 − RMM2, (b) RMM1, (c) RMM1 + RMM2, and (d) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Hatching indicates sensitivities (in uits of W m−2 per standard deviation of the MJO index) that are statistically significant at the 95% confidence level using the Student’s t test. The contour interval (CI) is 2 W m−2. The zero contour is not shown.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
The 20–90-day bandpass-filtered OLR for 2005–18 regressed onto (a) RMM1 − RMM2, (b) RMM1, (c) RMM1 + RMM2, and (d) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Hatching indicates sensitivities (in uits of W m−2 per standard deviation of the MJO index) that are statistically significant at the 95% confidence level using the Student’s t test. The contour interval (CI) is 2 W m−2. The zero contour is not shown.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
b. Seasonality of UTLS temperature and circulation associated with the MJO
Virts and Wallace (2014) demonstrated that the MJO can significantly impact the circulation of the TTL and the lowermost stratosphere. However, seasonal variations of the MJO impact were not examined since the authors used full time series (all seasons) in their analysis. In a similar manner to Fig. 2, we now examine the eastward propagation of MJO anomalies in 100-hPa temperature and horizontal winds during NDJF (Fig. 3, left) and JJAS (Fig. 3, right).
The 20–90-day bandpass-filtered temperature from MLS (shaded) and horizontal winds from MERRA-2 (arrows) at 100 hPa, regressed onto (a) RMM1 − RMM2, (b) RMM1, (c) RMM1 + RMM2, and (d) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Contour intervals are 0.2 K for temperature. Hatching indicates regions that are statistically significant at the 95% confidence level using the Student’s t test. Centers of high and low pressure systems are indicated by H and L, respectively.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
The 20–90-day bandpass-filtered temperature from MLS (shaded) and horizontal winds from MERRA-2 (arrows) at 100 hPa, regressed onto (a) RMM1 − RMM2, (b) RMM1, (c) RMM1 + RMM2, and (d) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Contour intervals are 0.2 K for temperature. Hatching indicates regions that are statistically significant at the 95% confidence level using the Student’s t test. Centers of high and low pressure systems are indicated by H and L, respectively.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
The 20–90-day bandpass-filtered temperature from MLS (shaded) and horizontal winds from MERRA-2 (arrows) at 100 hPa, regressed onto (a) RMM1 − RMM2, (b) RMM1, (c) RMM1 + RMM2, and (d) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Contour intervals are 0.2 K for temperature. Hatching indicates regions that are statistically significant at the 95% confidence level using the Student’s t test. Centers of high and low pressure systems are indicated by H and L, respectively.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
Temperature and wind anomalies during NDJF (Fig. 3, left) show a well-known response to the MJO (Weare 2010). When the MJO convection is above the Indian Ocean (prior to a peak in RMM1), equatorial easterlies above the center of deep convection are flanked by anticyclonic centers north and south of the equator. These equatorial easterlies and high pressure systems (labeled “H”) are associated with positive and negative temperature perturbations, respectively. To the east of the enhanced convection region, significant westerly anomalies are associated with two subtropical cyclones (labeled “L”) with warm temperature perturbations centered near the lows. The observed pattern of wind and temperature anomalies propagate eastward and is consistent with the upper-level flow of the forced response, modeled by Gill (1980): a combination of equatorially trapped Kelvin waves to the east and Rossby waves to the west of the heating.
The response of the temperature and horizontal circulation during boreal summer months is quite different from the one during boreal winter months. During JJAS, a dipole structure of wind and temperature perturbations due to the MJO is not very obvious (e.g., twin high and low pressure extratropical systems are weaker during the boreal summer months). Instead, planetary-scale region of low temperature anomalies develop at the equator to the east of the MJO convection and propagate eastward ahead of it from the central Pacific across South America and Africa, reaching the western Indian Ocean by the time of the peak in RMM2. During RMM2, two symmetric about the equator cold anticyclonic systems form between 120°E and 180°.
Total diabatic heating rates at 100 hPa, which serves as a proxy for vertical motion, are in a good agreement with temperature and circulation anomalies during both NDJF and JJAS (Fig. 4). Regions of upward (downward) mass flux (positive and negative heating rates, respectively) coincide with decreasing (increasing) temperature by adiabatic expansion (compression); see the thick solid (dashed) contours. In NDJF, subtropical upward mass flux is associated with high and low pressure regions seen above as a dipole structure of wind and temperature perturbations. During the JJAS anomalous equatorial upwelling, linked to cold temperature perturbations, encircles the globe by the time of the peak in RMM2. One may argue that within the subtropical cyclones and anticyclones in the UTLS, large-scale vertical motion may not be directly associated with any diabatic heating but rather remotely linked to the equatorial diabatic heating. However, given the observational and theoretical expectations that UTLS ozone is mostly controlled by transport processes and that the temperature and vertical velocity are linked through the thermal wind balance, we are confident that changes in vertical motion can be inferred, at least qualitatively, from total diabatic heating rates.
The 20–90-day bandpass-filtered MERRA-2 heating rate (K day−1) at 100 hPa regressed onto (a) RMM1 and (b) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Hatching indicates sensitivities that are statistically significant at the 95% confidence level using the Student's t test. Solid contours correspond to the 100-hPa temperature sensitivities with values of −0.2 and −0.4 K while dashed contours are 0.2 and 0.4 K.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
The 20–90-day bandpass-filtered MERRA-2 heating rate (K day−1) at 100 hPa regressed onto (a) RMM1 and (b) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Hatching indicates sensitivities that are statistically significant at the 95% confidence level using the Student's t test. Solid contours correspond to the 100-hPa temperature sensitivities with values of −0.2 and −0.4 K while dashed contours are 0.2 and 0.4 K.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
The 20–90-day bandpass-filtered MERRA-2 heating rate (K day−1) at 100 hPa regressed onto (a) RMM1 and (b) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Hatching indicates sensitivities that are statistically significant at the 95% confidence level using the Student's t test. Solid contours correspond to the 100-hPa temperature sensitivities with values of −0.2 and −0.4 K while dashed contours are 0.2 and 0.4 K.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
c. Dynamical controls
Previous studies showed that the temperature near or at the tropical tropopause can be controlled by both free and forced (convectively coupled) Kelvin waves, as well as by hydrostatic adjustment above the heating source (Kim et al. 2018; Randel and Wu 2005). While the Kelvin wave response in the TTL is more global, a response to the hydrostatic adjustment is localized above deep convection. Since observed perturbations in temperature, horizontal winds and upwelling rates (see Figs. 3 and 4) are not strictly confined to the heating source, it is most likely that a Kelvin wave response is a major factor controlling the differences in the tropical temperature and circulation between summer and winter seasons.
Cooling of the TTL in the presence of Kelvin waves forced by convection has a strong connection to the MJO (Zhou and Holton 2002; Ryu et al. 2008). To examine seasonal differences in the vertical propagation of forced Kelvin waves and their influence on UTLS temperature, Fig. 5 shows the equatorial (5°N–5°S) longitude–height cross sections of temperature anomalies regressed with the MJO indices, again, for JJAS and NDJF separately. The vertical cross sections of the correlation coefficients are included in Fig. S1 in the online supplemental material for a direct comparison with Virts and Wallace (2014) study. During both seasons we observe a wavelike features in the MJO-related temperature perturbations titling eastward with height. This is in agreement with the downward phase propagation and upward energy dispersion in equatorially trapped Kelvin waves (Andrews et al. 1987). However, there are considerable differences in the equatorial wave structure between the two seasons. Prior to the peak in RMM1, NDJF temperature perturbations (Fig. 5a, left) have a stronger tilt (steeper slope) than JJAS anomalies (Fig. 5a, right), but JJAS tilting temperature perturbations reach much farther east. Note that hydrostatic adjustment mechanism also contributes to the tropopause cooling leading to the strongest temperature sensitivities immediately above deep convection (Kim et al. 2018). However, the vertical structure of the temperature perturbations in the UTLS is dominated by the Kelvin wave response.
Longitude–height profiles of the 20–90-day bandpass-filtered temperature from MLS, averaged over 5°N–5°S and regressed with (a) RMM1 − RMM2, (b) RMM1, (c) RMM1 + RMM2, and (d) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Contour intervals are 0.2 K and hatching indicates sensitivities that are statistically significant at the 95% confidence level using the Student’s t test.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
Longitude–height profiles of the 20–90-day bandpass-filtered temperature from MLS, averaged over 5°N–5°S and regressed with (a) RMM1 − RMM2, (b) RMM1, (c) RMM1 + RMM2, and (d) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Contour intervals are 0.2 K and hatching indicates sensitivities that are statistically significant at the 95% confidence level using the Student’s t test.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
Longitude–height profiles of the 20–90-day bandpass-filtered temperature from MLS, averaged over 5°N–5°S and regressed with (a) RMM1 − RMM2, (b) RMM1, (c) RMM1 + RMM2, and (d) RMM2 during boreal (left) winter (NDJF) and (right) summer (JJAS) months. Contour intervals are 0.2 K and hatching indicates sensitivities that are statistically significant at the 95% confidence level using the Student’s t test.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
As MJO convection rapidly propagates across the Maritime Continent and the western and central Pacific, a narrow band of cold equatorial temperature anomalies forms around 100 hPa and encircles the globe. The band of negative temperature anomalies develops earlier and is much stronger and more zonally uniform during JJAS than NDJF. Previous studies associated this temperature anomalies with a fast-moving Kelvin wave front (Heckley and Gill 1984). However, Sakaeda and Roundy (2015, 2016) showed that eastern Pacific circumnavigating signal is tightly coupled with the midlatitude circulation and part of the observed response is due to midlatitude waves, approaching the equator and playing an important role in the modulation of the tropical circulation and associated with it temperature.
The faster-moving Kelvin wave front during JJAS than NDJF can also be seen in Fig. 3 as a belt of equatorial negative temperature sensitivities at 100 hPa, propagating rapidly eastward from the central Pacific across South America and Africa. While “Kelvin wave front” cannot be completely interpreted as a Gill Kelvin response, the circulation in the Western Hemisphere starts to resemble a Kelvin wave structure when anomalies reach Africa and the Indian Ocean (Sakaeda and Roundy 2015, 2016). The Kelvin wave complex is flanked by subtropical warm anomalies, suggestive of a Rossby wave signature. This is in agreement with Sardeshmukh and Hoskins (1988), who showed a Rossby response to the east of a heat source in a barotropic model that does not support Kelvin waves (see their Figs. 1 and 7).
(a) The vertical Kelvin wave group velocity Cgz (m s−1), (b) the zonal Kelvin wave group velocity Cgx (m s−1), and (c) climatological zonal wind U (the easterlies are in green and the westerlies are in red; m s−1) for boreal (left) winter (NDJF) and (right) summer (JJAS) months. The thick black contour in (b) indicates the region with (Cgz/Cgx)(L/D) ≥ 1.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
(a) The vertical Kelvin wave group velocity Cgz (m s−1), (b) the zonal Kelvin wave group velocity Cgx (m s−1), and (c) climatological zonal wind U (the easterlies are in green and the westerlies are in red; m s−1) for boreal (left) winter (NDJF) and (right) summer (JJAS) months. The thick black contour in (b) indicates the region with (Cgz/Cgx)(L/D) ≥ 1.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
(a) The vertical Kelvin wave group velocity Cgz (m s−1), (b) the zonal Kelvin wave group velocity Cgx (m s−1), and (c) climatological zonal wind U (the easterlies are in green and the westerlies are in red; m s−1) for boreal (left) winter (NDJF) and (right) summer (JJAS) months. The thick black contour in (b) indicates the region with (Cgz/Cgx)(L/D) ≥ 1.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
Zonal and seasonal fluctuations in Cgx can be explained by differences in the background state. According to Eq. (2), Cgx depends on the equatorial zonal wind U(X, Z, T) and static stability N(X, Z, T). Since N does not change much with longitude, zonal variations in Cgx are mainly due to the background climatological zonal wind (Fig. 6c). During NDJF (Fig. 6c, left), the Eastern Hemisphere is characterized by persistent easterlies (negative climatological zonal winds) and the Western Hemisphere by westerlies (positive values of climatological zonal winds). Maximum in easterlies is located around 120°E and ~15 km in altitude. Two maxima in westerlies are centered around 13 km (or 150 hPa) and 130° and 30°W. The influence of the Eastern Hemisphere easterlies and double peak in the Western Hemisphere westerlies is reflected in the zonal structure of Cgx during the winter season. The Eastern Hemisphere easterlies (Western Hemisphere westerlies) are much stronger (weaker) during JJAS (Fig. 6c, right) than during NDJF (Fig. 6c, left). Furthermore, a peak in JJAS climatological easterlies is located more than 30° westward (between 60° and 90°E), while the double peak in westerlies does not reach as high an amplitude as during winter. Smaller zonal variations in zonal winds during JJAS in the Western Hemisphere are consistent with smaller zonal variations in Cgx.
Equation (5) suggests that larger values of Cgx can lead to a smaller Kelvin wave amplitude and vice versa. This relationship holds everywhere outside of thick contour in Fig. 6b, where horizontal component of vertically–propagating Kelvin waves are sufficiently large [i.e., (Cgz/Cgx)(L/D) < 1, where L and D are horizontal and vertical scales]. In agreement with Ryu et al. (2008), relatively large temperature perturbations in the western Pacific during boreal winter months are observed where the U < 0 and Cgx is small. Likewise, the Kelvin wave amplitude is smaller in the eastern Pacific. According to the simplified wave-action conservation equation, JJAS Kelvin waves amplify farther westward and propagate farther east due to 1) stronger equatorial easterlies that are also located more westward in the Eastern Hemisphere and 2) weaker westerlies in the Western Hemisphere.
In the vertical direction, the Kelvin wave temperature is modulated by the background buoyancy frequency N(z) to the higher power than c − U(z). Thus, Kelvin wave temperature fluctuations depend only weakly on the background wind speed U(z). However, differences in static stability between summer and winter seasons are very small. Thus, differences in the UTLS temperature perturbation between JJAS and NDJF are mainly due to seasonal differences in the zonal structure of the climatological zonal winds.
Despite the very simplified nature of the performed analysis, the above conclusions are in overall agreement with more extensive Kelvin wave ray-tracing analyses by Flannaghan and Fueglistaler (2013). Their analysis of Kelvin wave seasonality indicated that wave propagation is inhibited by strong westerlies associated with Walker circulation during boreal winter but propagating over most longitudes during boreal summer. As in our study, wave action and wave amplitude in the tropical UTLS, computed by integrating all terms in Eq. (1), is more zonally uniform during boreal summer than winter. While zonal winds in the TTL influence Kelvin wave propagation, they emphasize that the exact relationship between two is more complex. Flannaghan and Fueglistaler (2013) demonstrate an important contribution of the ray convergence term in the direction perpendicular to the rays, which was eliminated in Ryu et al. (2008) study. This implies that not only Kelvin waves in the TTL are modulated locally by the background conditions, but they are also affected nonlocally by the ray convergence.
d. Seasonal variations in trace gases in the UTLS
Ozone variations in the UTLS (below about 20–22 km) are mostly controlled by transport processes (i.e., atmospheric circulation acting on the strong background horizontal and vertical gradients) due to much longer chemical lifetime in relationship to transport time scales. Thus, it is important to examine the trace gas response to the anomalous changes in the circulation due to the MJO. In previous observational studies of intraseasonal ozone variability in the tropical UTLS, the authors either used full time series (all seasons) or focused on the boreal winter months when the MJO is stronger and more frequent (Fujiwara et al. 1998; Virts and Wallace 2014). Observations of a large suite of trace gases from the MLS now provide long and continuous coverage of high-quality data, suitable for the investigation of zonal and height variations in ozone and other trace gases as well as for performing analysis for boreal winter and summer months separately.
Seasonal differences in ozone anomalies associated with the eastward propagation of the MJO during RMM1 and RMM2 are shown in Fig. 7 (full eight-panel figure is included in Fig. S2 in the supplemental materials). Similar to the UTLS temperature and circulation, the ozone response to the MJO during the boreal winter months was previously observed and documented (Weare 2010). The dipole pattern of statistically significant increases in ozone at the equator and decreases in the extratropics propagate eastward from Indian Ocean into the western Pacific.
As in Fig. 4, but for filtered 100-hPa ozone. Ozone sensitivities are shown as a percentage relative to the ozone climatological annual mean. The zero contour is not shown.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
As in Fig. 4, but for filtered 100-hPa ozone. Ozone sensitivities are shown as a percentage relative to the ozone climatological annual mean. The zero contour is not shown.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
As in Fig. 4, but for filtered 100-hPa ozone. Ozone sensitivities are shown as a percentage relative to the ozone climatological annual mean. The zero contour is not shown.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
The summertime response to the MJO is quite different from the wintertime response and is more zonally symmetric. During JJAS, positive ozone sensitivities above the equatorial Africa and Indian Ocean prior to RMM1 (see Fig. S2a), are replaced with reductions in ozone above the equatorial Pacific Ocean during RMM1 (Fig. 7a, right). As the MJO continues its eastward propagation across the western and central Pacific, negative ozone anomalies encircle the globe reaching western Indian Ocean by the time of RMM2 (Fig. 7b, right). The strongest JJAS ozone sensitivities are observed during RMM2 where values are more than 9% lower relative to the JJAS climatology above the tropical Indian Ocean and extratropical western Pacific.
During both seasons, ozone response to the MJO is largely coherent in space and time with the temperature (black thick contours in Fig. 7) and upwelling anomalies (see Fig. 4). The reduction of equatorial ozone during JJAS resembles the cold equatorial temperature anomalies due to a fast-moving Kelvin wave front, while ozone response during NDJF is in good agreement with temperature and circulation changes modeled by Gill (1980). As result, enhanced (reduced) vertical motion, associated with planetary-scale perturbations, that causes temperature to decrease (increase) is also pumping more (less) ozone-poor air from below. Furthermore, similar to the equatorial temperature perturbation, Kelvin wave features are also clearly observed vertically in filtered ozone anomalies regressed with MJO1 and MJO2 (Fig. 8; also see eight-panel version of this figure in the supplemental materials, Fig. S4). The vertical structure of ozone regressions are also in a good agreements with MERRA-2 total diabatic heating rates (not shown), indicating a dominant role of Kelvin wave–induced vertical transport acting on the (positive) background ozone gradient. Thus, seasonal differences in the structure of vertically propagating equatorial Kelvin waves plays a key role in determining seasonal variations in ozone response to the MJO near the equator.
As in Fig. 5, but for filtered ozone anomalies regressed with RMM1 and RMM2. The contour interval is 2%. Percentages are relative to the ozone climatological annual mean.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
As in Fig. 5, but for filtered ozone anomalies regressed with RMM1 and RMM2. The contour interval is 2%. Percentages are relative to the ozone climatological annual mean.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
As in Fig. 5, but for filtered ozone anomalies regressed with RMM1 and RMM2. The contour interval is 2%. Percentages are relative to the ozone climatological annual mean.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
While the ozone response to the MJO from MLS observations is strongly coherent with temperature anomalies and large-scale vertical transport, it is interesting to examine the influence of circulation on other trace gases, such as carbon monoxide and water vapor. The majority of carbon monoxide and water vapor in the UTLS are of tropospheric origin and they enter tropical UTLS primarily through the TTL. While the amount of carbon monoxide mostly depends on the rates of upwelling (similar to ozone), water vapor is also heavily controlled by cold temperatures in the TTL (through a process of dehydration). Our analysis of carbon monoxide at 100 hPa shows similar to ozone response during NDJF (see Fig. S4). However, statistical significance of JJAS sensitivities with the MJO indices is weak.
Very strong water vapor dependence on the TTL temperatures makes its response to the MJO more complicated. Water vapor anomalies originate near the TTL (~100 hPa) and then propagate upward with time by the large-scale upwelling in the tropical stratosphere. Above the MJO deep convection, temperatures near 100 hPa become even cooler (see Fig. S5) resulting in a strong reduction of water vapor. To isolate the effect of dehydration from the effect of changes in upwelling rates associated with the MJO propagation, we show water vapor sensitivities at a higher level (83 hPa) in the UTLS (Fig. 9). During NDJF, water vapor sensitivities at 83 hPa are similar to ozone and temperature perturbations at 100 hPa. However, a JJAS response to the RMM1 is weak and insignificant. Perturbations in water vapor due to RMM2 show very little indication of the Kelvin wave front encircling the globe (Fig. 9, right) with statistically significant sensitivities above tropical Indian Ocean and western Pacific. Figure 10 (left) shows that Kelvin wave signal is observed in the height variations of water during the boreal winter months. This is in agreement with Virts and Wallace (2014) and Schwartz et al. (2008), who also showed that temperature and water vapor anomalies in the TTL and above the Indian Ocean and Maritime Continent are nearly in quadrature (i.e., low water vapor anomaly follows the cold temperature perturbation associated with downward and eastward propagation of Kelvin wave 1/4 cycle later). In contrast, only local impacts of the MJO-related cooling and associated reduction of water vapor through dehydration near the TTL are observed during JJAS (Fig. 10, right). The cause of horizontal layering above 100 hPa during JJAS is unclear and requires further investigation; however, water vapor sensitivities above 17 km are not statistically significant.
As in Fig. 4, but for 83-hPa water vapor anomalies. The contour interval is 20 ppbv. Solid contours correspond to the 100-hPa temperature sensitivities with values of −0.2 and −0.4 K while dashed contours are 0.2 and 0.4 K.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
As in Fig. 4, but for 83-hPa water vapor anomalies. The contour interval is 20 ppbv. Solid contours correspond to the 100-hPa temperature sensitivities with values of −0.2 and −0.4 K while dashed contours are 0.2 and 0.4 K.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
As in Fig. 4, but for 83-hPa water vapor anomalies. The contour interval is 20 ppbv. Solid contours correspond to the 100-hPa temperature sensitivities with values of −0.2 and −0.4 K while dashed contours are 0.2 and 0.4 K.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
As in Fig. 5, but for filtered water vapor anomalies regressed with RMM1 and RMM2. The contour interval is 1%. Percentages are relative to the H2O climatological annual mean.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
As in Fig. 5, but for filtered water vapor anomalies regressed with RMM1 and RMM2. The contour interval is 1%. Percentages are relative to the H2O climatological annual mean.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
As in Fig. 5, but for filtered water vapor anomalies regressed with RMM1 and RMM2. The contour interval is 1%. Percentages are relative to the H2O climatological annual mean.
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
The results of similar temperature, circulation and trace gas analysis by Virts and Wallace (2014) are somewhat in the between JJAS and NDJF responses to the MJO. As a reminder, they used the full time series of MLS ozone over 4 years. For instance, their ozone and temperature sensitivities in the extratropics to the RMM2 shows much closer resemblance to our boreal winter response to the RMM2. This is not surprising since the MJO is stronger and more frequent during the boreal winters than summers.
4. Conclusions
In this study MLS observations and MERRA-2 were used to examine the zonal and seasonal variations in the response of the tropical and extratropical UTLS temperature, circulation, and trace gases to MJO forcing.
We have shown that the response of temperature, circulation, and trace gases during boreal summer months is different from the response during boreal winter months. As summarized in Fig. 11a, atmospheric fields in the UTLS during NDJF exhibit planetary-scale perturbations consistent with the upper-level flow of the model by Gill (1980) and is a combination of Rossby and Kelvin flows patterns. Anticyclonic centers (high pressure systems) north and south of the equator over the Indian Ocean (during a peak in RMM1) are associated with anomalous upwelling and cooling while cyclonic centers (low pressure systems) to the east of the deep convection coincide with anomalous downwelling and warming. In contrast, twin high and low pressure extratropical systems associated with the forced Rossby wave response are weaker during the boreal summer months with a much stronger equatorial Kelvin wave front. The Kelvin wave front produces enhanced upwelling leading to a strong cooling (Fig. 11b).
Schematic depiction of the 100-hPa temperature, circulation, and ozone anomalies associated with the MJO when the enhanced convection is centered across the Indian Ocean and the suppressed convective phase is centered over the west-central Pacific Ocean (RMM1) during (a) boreal winter and (b) boreal summer months. Blue regions correspond to areas of negative temperature and ozone perturbations while opposite is true for red regions. The circulation cells (white arrows) at the 100-hPa level highlight characteristic wind anomalies associated with the MJO. Upward and downward motions are shown by vertical black arrows. Green arrows show the tropospheric structure of the MJO with the surface convergence of the winds and upwelling above the region of enhanced convection and divergence of the surface winds and downwelling above the region of suppressed convection [adopted from Rui and Wang (1990)].
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
Schematic depiction of the 100-hPa temperature, circulation, and ozone anomalies associated with the MJO when the enhanced convection is centered across the Indian Ocean and the suppressed convective phase is centered over the west-central Pacific Ocean (RMM1) during (a) boreal winter and (b) boreal summer months. Blue regions correspond to areas of negative temperature and ozone perturbations while opposite is true for red regions. The circulation cells (white arrows) at the 100-hPa level highlight characteristic wind anomalies associated with the MJO. Upward and downward motions are shown by vertical black arrows. Green arrows show the tropospheric structure of the MJO with the surface convergence of the winds and upwelling above the region of enhanced convection and divergence of the surface winds and downwelling above the region of suppressed convection [adopted from Rui and Wang (1990)].
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
Schematic depiction of the 100-hPa temperature, circulation, and ozone anomalies associated with the MJO when the enhanced convection is centered across the Indian Ocean and the suppressed convective phase is centered over the west-central Pacific Ocean (RMM1) during (a) boreal winter and (b) boreal summer months. Blue regions correspond to areas of negative temperature and ozone perturbations while opposite is true for red regions. The circulation cells (white arrows) at the 100-hPa level highlight characteristic wind anomalies associated with the MJO. Upward and downward motions are shown by vertical black arrows. Green arrows show the tropospheric structure of the MJO with the surface convergence of the winds and upwelling above the region of enhanced convection and divergence of the surface winds and downwelling above the region of suppressed convection [adopted from Rui and Wang (1990)].
Citation: Journal of the Atmospheric Sciences 77, 4; 10.1175/JAS-D-19-0183.1
A pattern of temperature and circulation anomalies in the UTLS during NDJF follows deep convection as the MJO propagates eastward and reaches the western Pacific Ocean (at the time of a peak in RMM2). During the peak in RMM2, cold regions within extratropical high pressure systems are above the western Pacific Ocean and warm regions within low pressure systems are to the north and south of the Indian Ocean. In contrast, planetary-scale region of low-temperature anomalies associated with the Kelvin wave front during boreal summer encircles the globe by the time of RMM2.
As a key result of this study, we have demonstrated that the differences in UTLS temperature and circulation anomalies between boreal summer and winter months are mainly due to differences in the zonal structure of Kelvin wave group velocity at the equator, which strongly depends on the background zonal winds. Following the analysis of Kelvin wave amplitude by Ryu et al. (2008), we show that JJAS Kelvin waves amplify more westward and propagate farther east than NDJF Kelvin waves because of stronger summertime equatorial easterlies in the Eastern Hemisphere that are also more westward and weaker summertime westerlies in the Western Hemisphere.
Trace gas intraseasonal variability in the tropical and extratropical UTLS is also different between JJAS and NDJF. The response of ozone and carbon monoxide to the MJO is strongly coherent with large-scale vertical transport (e.g., enhanced upwelling leads to the reduction of ozone and increase in carbon monoxide). The very strong dependence of water vapor on the TTL temperature makes the response to the MJO more complicated. Above the TTL the water vapor anomalies during boreal winter months are similar to those of ozone and carbon monoxide and in an agreement with the vertical propagation of Kelvin waves. However, perturbations in water vapor during JJAS show almost no indication of Kelvin wave structure. Instead, enhanced local cooling above the MJO convection near the TTL dominates the region resulting in a strong reduction of water vapor through dehydration.
The higher vertical resolution of the dataset is crucial for resolving vertical structure of Kelvin waves, which has been shown to play a key role in the MJO-induced anomalies of temperature and tropical ozone (Kiladis et al. 2001). While we most likely cannot fully resolve the vertical wavelength and its changes (due to relatively small vertical wavelength of the Kelvin response), the MLS temperature fields indicate the location and basic structure of the equatorial planetary Kelvin waves.
In spite of notable improvements in some modeling systems (Kim et al. 2009), a realistic MJO has remained a challenging task for most GCMs/CCMs (Hung et al. 2013; Jiang et al. 2015; Kim et al. 2009; Lin et al. 2006; Slingo et al. 1996; Sperber et al. 2005) mostly due to the lack of understanding of MJO dynamics. In particular, many models struggle to realistically represent MJO variance, eastward propagation, and associated tilted vertical structure of the MJO (Jiang et al. 2015). The inability of the CCMs to generate shorter-time-scale variability such as that from the MJO, can potentially lead to the lack of or much weaker variability in tropical and extratropical composition of the troposphere and lower stratosphere as noted in previous studies (Butchart et al. 2011; Young et al. 2018; Ziemke et al. 2015). The analysis of MLS observations presented here may be useful for evaluation and validation of the MJO-related physical and dynamical processes in a hierarchy of models. For instance, it is highly desirable to examine a range of models in their ability to simulate seasonal differences in the UTLS temperature and circulation due to the MJO as it can have a profound impact on trace gas variability in the tropics and subtropics. A more realistic representation of the spectrum of variability in climate models will provide a better estimate of future projections.
Acknowledgments
OVT was supported by an appointment to the NASA Postdoctoral Program at the NASA’s Goddard Space Flight Center, administered by Universities Space Research Association under contract with NASA. DDW was supported by NASA ACMAP Grant NNX17AI31G. LDO was supported by the NASA MAP program. We would also like to acknowledge and thank all members of the data science and support teams. MERRA-2 and MLS data are available from the NASA Goddard Space Flight Center Earth Sciences Data and Information Services Center. Interpolated OLR data provided by the NOAA/OAR/ESRL PSD, Boulder, Colorado, from their website at https://www.esrl.noaa.gov/psd/. The all-season real-time multivariate (RMM) MJO index (Wheeler and Hendon 2004) is from the Centre for Australian Weather and Climate Research (http://www.bom.gov.au/climate/mjo/graphics/rmm.74toRealtime.txt.). We thank the three reviewers for their substantive comments, which helped to improve the manuscript.
REFERENCES
Adames, Á. F., and J. M. Wallace, 2014: Three-dimensional structure and evolution of the MJO and its relation to the mean flow. J. Atmos. Sci., 71, 2007–2026, https://doi.org/10.1175/JAS-D-13-0254.1.
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.
Andrews, D. G., and M. F. McIntyre, 1976: Planetary waves in horizontal and vertical shear: Asymptotic theory for equatorial waves in weak shear. J. Atmos. Sci., 33, 2049–2053, https://doi.org/10.1175/1520-0469(1976)033<2049:PWIHAV>2.0.CO;2.
Andrews, D. G., J. Holton, and C. B. Leovy, 1987: Middle Atmosphere Dynamics. Academic Press, 489 pp.
Butchart, N., and Coauthors, 2011: Multimodel climate and variability of the stratosphere. J. Geophys. Res., 116, D05102, https://doi.org/10.1029/2010JD014995.
Chang, C.-P., 1976: Forcing of stratospheric Kelvin waves by tropospheric heat sources. J. Atmos. Sci., 33, 740–744, https://doi.org/10.1175/1520-0469(1976)033<0740:FOSKWB>2.0.CO;2.
Chen, G., and B. Wang, 2019: Dynamic moisture mode versus moisture mode in MJO dynamics: Importance of the wave feedback and boundary layer convergence feedback. Climate Dyn., 52, 5127–5143, https://doi.org/10.1007/s00382-018-4433-7.
Flannaghan, T. J., and S. Fueglistaler, 2013: The importance of the tropical tropopause layer for equatorial Kelvin wave propagation. J. Geophys. Res. Atmos., 118, 5160–5175, https://doi.org/10.1002/JGRD.50418.
Fueglistaler, S., E. Dessler, T. J. Dunkerton, I. Folkins, Q. Fu, and P. W. Ote, 2009a: Tropical tropopause layer. Rev. Geophys., 47, RG1004, https://doi.org/10.1029/2008RG000267.
Fueglistaler, S., B. Legras, A. Beljaars, J.-J. Morcrette, A. Simmons, A. M. Tompkins, and S. Uppala, 2009b: The diabatic heat budget of the upper troposphere and lower/mid stratosphere in ECMWF reanalyses. Quart. J. Roy. Meteor. Soc., 135, 21–37, https://doi.org/10.1002/qj.361.
Fujiwara, M., K. Kita, and T. Ogawa, 1998: Stratosphere-troposphere exchange of ozone associated with the equatorial Kelvin wave as observed with ozonesondes and rawinsondes. J. Geophys. Res., 103, 19 173–19 182, https://doi.org/10.1029/98JD01419.
Gao, X. H., and J. L. Stanford, 1990: Low-frequency oscillations in total ozone measurements. J. Geophys. Res., 95, 13 797–13 806, https://doi.org/10.1029/JD095iD09p13797.
Gelaro, R., and Coauthors, 2017: The Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2). J. Climate, 30, 5419–5454, https://doi.org/10.1175/JCLI-D-16-0758.1.
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.
Gutzler, D. S., and R. A. Madden, 1989: Seasonal variations in the spatial structure of intraseasonal tropical wind fluctuations. J. Atmos. Sci., 46, 641–660, https://doi.org/10.1175/1520-0469(1989)046<0641:SVITSS>2.0.CO;2.
Heckley, W. A., and A. E. Gill, 1984: Some simple analytical solutions to the problem of forced equatorial long waves. Quart. J. Roy. Meteor. Soc., 110, 203–217, https://doi.org/10.1002/QJ.49711046314.
Hendon, H. H., and M. L. Salby, 1994: The life cycle of the Madden–Julian oscillation. J. Atmos. Sci., 51, 2225–2237, https://doi.org/10.1175/1520-0469(1994)051<2225:TLCOTM>2.0.CO;2.
Hung, M. P., J. L. Lin, W. Wang, D. Kim, T. Shinoda, and S. J. Weaver, 2013: MJO and convectively coupled equatorial waves simulated by CMIP5 climate models. J. Climate, 26, 6185–6214, https://doi.org/10.1175/JCLI-D-12-00541.1.
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. Atmos., 120, 4718–4748, https://doi.org/10.1002/2014JD022375.
Kiladis, G. N., K. H. Straub, G. C. Reid, and K. S. Gage, 2001: Aspects of interannual and intraseasonal variability of the tropopause and lower stratosphere. Quart. J. Roy. Meteor. Soc., 127, 1961–1983, https://doi.org/10.1002/QJ.49712757606.
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, J., W. J. Randel, and T. Birner, 2018: Convectively driven tropopause-level cooling and its influences on stratospheric moisture. J. Geophys. Res. Atmos., 123, 590–606, https://doi.org/10.1002/2017JD027080.
Knutson, T. R., and K. M. Weickmann, 1987: 30–60 day atmospheric oscillations: Composite life cycles of convection and circulation anomalies. Mon. Wea. Rev., 115, 1407–1436, https://doi.org/10.1175/1520-0493(1987)115<1407:DAOCLC>2.0.CO;2.
Knutson, T. R., K. M. Weickmann, and J. E. Kutzbach, 1986: Global-scale intraseasonal oscillations of outgoing longwave radiation and 250 mb zonal wind during Northern Hemisphere summer. Mon. Wea. Rev., 114, 605–623, https://doi.org/10.1175/1520-0493(1986)114<0605:GSIOOO>2.0.CO;2.
Lacis, A. A., D. J. Wuebbles, and J. A. Logan, 1990: Radiative forcing of climate by changes in the vertical distribution of ozone. J. Geophys. Res., 95, 9971–9981, https://doi.org/10.1029/JD095iD07p09971.
Lavender, S. L., and A. J. Matthews, 2009: Response of the West African monsoon to the Madden–Julian oscillation. J. Climate, 22, 4097–4116, https://doi.org/10.1175/2009JCLI2773.1.
Lawrence, D. M., and P. J. Webster, 2002: The boreal summer intraseasonal oscillation: Relationship between northward and eastward movement of convection. J. Atmos. Sci., 59, 1593–1606, https://doi.org/10.1175/1520-0469(2002)059<1593:TBSIOR>2.0.CO;2.
Li, K.-F., B. Tian, D. E. Waliser, M. J. Schwartz, J. L. Neu, J. R. Worden, and Y. L. Yung, 2012: Vertical structure of MJO-related subtropical ozone variations from MLS, TES, and SHADOZ data. Atmos. Chem. Phys., 12, 425–436, https://doi.org/10.5194/acp-12-425-2012.
Li, K.-F., B. Tian, K.-K. Tung, L. Kuai, J. R. Worden, Y. L. Yung, and B. L. Slawski, 2013: A link between tropical intraseasonal variability and Arctic stratospheric ozone. J. Geophys. Res. Atmos., 118, 4280–4289, https://doi.org/10.1002/JGRD.50391.
Liebmann, B., and C. A. Smith, 1996: Description of a complete (interpolated) outgoing longwave radiation dataset. Bull. Amer. Meteor. Soc., 77, 1275–1277, https://doi.org/10.1175/1520-0477-77.6.1274.
Lighthill, J., 1978: Waves in Fluids. Cambridge University Press, 504 pp.
Lin, J.-L., and Coauthors, 2006: Tropical intraseasonal variability in 14 IPCC AR4 climate models. Part I: Convective signals. J. Climate, 19, 2665–2690, https://doi.org/10.1175/JCLI3735.1.
Livesey, N., and Coauthors, 2015: EOS MLS version 4.2x level 2 data quality and description document. JPL Tech. Rep. JPL D-33509 Rev. C, 169 pp., https://mls.jpl.nasa.gov/data/v4-2_data_quality_document.pdf.
Madden, R. A., 1986: Seasonal variations of the 40–50 day oscillation in the tropics. J. Atmos. Sci., 43, 3138–3158, https://doi.org/10.1175/1520-0469(1986)043<3138:SVOTDO>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.
Molod, A., L. Takacs, M. Suarez, and J. Bacmeister, 2015: Development of the GEOS-5 atmospheric general circulation model: Evolution from MERRA to MERRA2. Geosci. Model Dev., 8, 1339–1356, https://doi.org/10.5194/gmd-8-1339-2015.
Randel, W. J., and F. Wu, 2005: Kelvin wave variability near the equatorial tropopause observed in GPS radio occultation measurements. J. Geophys. Res., 110, D03102, https://doi.org/10.1029/2004JD005006.
Randel, W. J., and F. Wu, 2015: Variability of zonal mean tropical temperatures derived from a decade of GPS radio occultation data. J. Atmos. Sci., 72, 1261–1275, https://doi.org/10.1175/JAS-D-14-0216.1.
Riese, M., F. Ploeger, A. Rap, B. Vogel, P. Konopka, M. Dameris, and P. Forster, 2012: Impact of uncertainties in atmospheric mixing on simulated UTLS composition and related radiative effects. J. Geophys. Res., 117, D16305, https://doi.org/10.1029/2012JD017751.
Rui, H., and B. Wang, 1990: Development characteristics and dynamic structure of tropical intraseasonal convection anomalies. J. Atmos. Sci., 47, 357–379, https://doi.org/10.1175/1520-0469(1990)047<0357:DCADSO>2.0.CO;2.
Ryu, J.-H., S. Lee, and S.-W. Son, 2008: Vertically propagating Kelvin waves and tropical tropopause variability. J. Atmos. Sci., 65, 1817–1837, https://doi.org/10.1175/2007JAS2466.1.
Sabutis, J. L., J. L. Stanford, and K. P. Bowman, 1987: Evidence for 35-50 day low frequency oscillations in total ozone mapping spectrometer data. Geophys. Res. Lett., 14, 945–947, https://doi.org/10.1029/GL014i009p00945.
Sakaeda, N., and P. E. Roundy, 2015: The development of upper-tropospheric wind over the Western Hemisphere in association with MJO convective initiation. J. Atmos. Sci., 72, 3138–3160, https://doi.org/10.1175/JAS-D-14-0293.1.
Sakaeda, N., and P. E. Roundy, 2016: The development of upper-tropospheric geopotential height anomaly in the Western Hemisphere during MJO convective initiations. Quart. J. Roy. Meteor. Soc., 142, 942–956, https://doi.org/10.1002/qj.2696.
Salby, M. L., and H. H. Hendon, 1994: Intraseasonal behavior of clouds, temperature, and motion in the tropics. J. Atmos. Sci., 51, 2207–2224, https://doi.org/10.1175/1520-0469(1994)051<2207:IBOCTA>2.0.CO;2.
Sardeshmukh, P. D., and B. J. Hoskins, 1988: The generation of global rotational flow by steady idealized tropical divergence. J. Atmos. Sci., 45, 1228–1251, https://doi.org/10.1175/1520-0469(1988)045<1228:TGOGRF>2.0.CO;2.
Schwartz, M. J., D. E. Waliser, B. Tian, D. L. Wu, J. H. Jiang, and W. G. Read, 2008: Characterization of MJO-related upper tropospheric hydrological processes using MLS. Geophys. Res. Lett., 35, L08812, https://doi.org/10.1029/2008GL033675.
Slingo, J. M., and Coauthors, 1996: Intraseasonal oscillations in 15 atmospheric general circulation models: Results from an AMIP diagnostic subproject. Climate Dyn., 12, 325–357, https://doi.org/10.1007/BF00231106.
Sperber, K. R., S. Gualdi, S. Legutke, and V. Gayler, 2005: The Madden–Julian oscillation in ECHAM4 coupled and uncoupled general circulation models. Climate Dyn., 25, 117–140, https://doi.org/10.1007/s00382-005-0026-3.
Stolarski, R. S., and S. M. Frith, 2006: Search for evidence of trend slow-down in the long-term TOMS/SBUV total ozone data record: The importance of instrument drift uncertainty. Atmos. Chem. Phys., 6, 4057–4065, https://doi.org/10.5194/acp-6-4057-2006.
Takayabu, Y., 1994: Large-scale cloud disturbances associated with equatorial waves. J. Meteor. Soc. Japan, 72, 433–449, https://doi.org/10.2151/JMSJ1965.72.3_433.
Tian, B., Y. L. Yung, D. E. Waliser, T. Tyranowski, L. Kuai, E. J. Fetzer, and F. W. Irion, 2007: Intraseasonal variations of the tropical total ozone and their connection to the Madden-Julian oscillation. Geophys. Res. Lett., 34, L08704, https://doi.org/10.1029/2007GL029451.
Virts, K. S., and J. M. Wallace, 2010: Annual, interannual, and intraseasonal variability of tropical tropopause transition layer cirrus. J. Atmos. Sci., 67, 3097–3112, https://doi.org/10.1175/2010JAS3413.1.
Virts, K. S., and J. M. Wallace, 2014: Observations of temperature, wind, cirrus, and trace gases in the tropical tropopause transition layer during the MJO. J. Atmos. Sci., 71, 1143–1157, https://doi.org/10.1175/JAS-D-13-0178.1.
Wallace, J. M., and V. E. Kousky, 1968: Observational evidence of Kelvin waves in the tropical stratosphere. J. Atmos. Sci., 25, 900–907, https://doi.org/10.1175/1520-0469(1968)025<0900:oeokwi>2.0.co;2.
Weare, B. C., 2010: Madden-Julian oscillation in the tropical stratosphere. J. Geophys. Res., 115, D17113, https://doi.org/10.1029/2009JD013748.
Weickmann, K. M., G. R. Lussky, and J. E. Kutzbach, 1985: Intraseasonal (30–60 day) fluctuations of outgoing longwave radiation and 250 mb streamfunction during northern winter. Mon. Wea. Rev., 113, 941–961, https://doi.org/10.1175/1520-0493(1985)113<0941:IDFOOL>2.0.CO;2.
Wheeler, M. C., 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.
Wheeler, M. C., G. N. Kiladis, and P. J. Webster, 2000: Large-scale dynamical fields associated with convectively coupled equatorial waves. J. Atmos. Sci., 57, 613–640, https://doi.org/10.1175/1520-0469(2000)057<0613:LSDFAW>2.0.CO;2.
Yang, Q., Q. Fu, J. Austin, A. Gettelman, F. Li, and H. Vömel, 2008: Observationally derived and general circulation model simulated tropical stratospheric upward mass fluxes. J. Geophys. Res., 113, D00B07, https://doi.org/10.1029/2008JD009945.
Young, P. J., and Coauthors, 2018: Tropospheric Ozone Assessment Report: Assessment of global-scale model performance for global and regional ozone distributions, variability, and trends. Elem. Sci. Anth., 6, 10, https://doi.org/10.1525/elementa.265.
Zhang, C., 2005: Madden-Julian oscillation. Rev. Geophys., 43, RG2003, https://doi.org/10.1029/2004RG000158.
Zhang, C., and M. Dong, 2004: Seasonality in the Madden–Julian oscillation. J. Climate, 17, 3169–3180, https://doi.org/10.1175/1520-0442(2004)017<3169:SITMO>2.0.CO;2.
Zhou, X., and J. R. Holton, 2002: Intraseasonal variations of tropical cold-point tropopause temperatures. J. Climate, 15, 1460–1473, https://doi.org/10.1175/1520-0442(2002)015<1460:IVOTCP>2.0.CO;2.
Ziemke, J. R., A. R. Douglass, L. D. Oman, S. E. Strahan, and B. N. Duncan, 2015: Tropospheric ozone variability in the tropics from ENSO to MJO and shorter timescales. Atmos. Chem. Phys., 15, 8037–8049, https://doi.org/10.5194/acp-15-8037-2015.
