• Albignat, J. P., and R. J. Reed, 1980: The origin of African wave disturbances during Phase III of GATE. Mon. Wea. Rev., 108, 18271839.

    • Search Google Scholar
    • Export Citation
  • Annamalai, H., and J. M. Slingo, 2001: Active/break cycles: Diagnosis of the intraseasonal variability of the Asian Summer Monsoon. Climate Dyn., 18, 85102.

    • Search Google Scholar
    • Export Citation
  • Bantzer, C. H., and J. M. Wallace, 1996: Intraseasonal variability in tropical mean temperature and precipitation and their relation to the tropical 40–50 day oscillation. J. Atmos. Sci., 53, 30323045.

    • Search Google Scholar
    • Export Citation
  • Berry, G., and C. D. Thorncroft, 2005: Case study of an intense African easterly wave. Mon. Wea. Rev., 133, 752766.

  • Bretherton, C. S., M. E. Peters, and L. E. Back, 2004: Relationships between water vapor path and precipitation over the tropical oceans. J. Climate, 17, 15171528.

    • Search Google Scholar
    • Export Citation
  • Burpee, R. W., 1972: The origin and structure of easterly waves in the lower troposphere of North Africa. J. Atmos. Sci., 29, 7790.

  • Burpee, R. W., 1974: Characteristics of North African easterly waves during the summers of 1968 and 1969. J. Atmos. Sci., 31, 15561570.

    • Search Google Scholar
    • Export Citation
  • Camargo, S. J., M. C. Wheeler, and A. H. Sobel, 2009: Diagnosis of the MJO modulation of tropical cyclogenesis using an empirical index. J. Atmos. Sci., 66, 30613074.

    • Search Google Scholar
    • Export Citation
  • Carlson, T. N., 1969a: Some remarks on African disturbances and their progress over the tropical Atlantic. Mon. Wea. Rev., 97, 716726.

    • Search Google Scholar
    • Export Citation
  • Carlson, T. N., 1969b: Synoptic histories of three African disturbances that developed into Atlantic hurricanes. Mon. Wea. Rev., 97, 256276.

    • Search Google Scholar
    • Export Citation
  • Chauvin, F., R. Roehrig, and J.-P. Lafore, 2010: Intraseasonal variability of the Saharan heat low and its link with midlatitudes. J. Climate, 23, 25442561.

    • Search Google Scholar
    • Export Citation
  • de Coëtlogon, G., S. Janicot, and A. Lazar, 2010: Intraseasonal variability of the ocean–atmosphere coupling in the Gulf of Guinea during boreal spring and summer. Quart. J. Roy. Meteor. Soc., 136, 426441.

    • Search Google Scholar
    • Export Citation
  • Gill, A. E., 1980: Some simple solutions for heat-induced tropical circulation. Quart. J. Roy. Meteor. Soc., 106, 447462.

  • Hall, N. M. J., G. N. Kiladis, and C. D. Thorncroft, 2006: Three-dimensional structure of African easterly waves. Part II: Dynamical modes. J. Atmos. Sci., 63, 22312245.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., and M. L. Salby, 1994: The life cycle of the Madden–Julian Oscillation. J. Atmos. Sci., 51, 22252237.

  • Hodges, K. I., D. W. Chappell, G. J. Robinson, and G. Yang, 2000: An improved algorithm for generating global window brightness temperatures from multiple satellite infrared imagery. J. Atmos. Oceanic Technol., 17, 12961312.

    • Search Google Scholar
    • Export Citation
  • Hsieh, J.-S., and K. H. Cook, 2005: Generation of African easterly wave disturbances: Relationship to the African easterly jet. Mon. Wea. Rev., 133, 13111327.

    • Search Google Scholar
    • Export Citation
  • Janicot, S., F. Mounier, N. M. J. Hall, S. Leroux, B. Sultan, and G. N. Kiladis, 2009: Dynamics of the West African monsoon. Part IV: Analysis of 25–90-day variability of convection and the role of the Indian monsoon. J. Climate, 22, 15411565.

    • Search Google Scholar
    • Export Citation
  • Janicot, S., and Coauthors, 2011: Intraseasonal variability of the West African monsoon. Atmos. Sci. Lett., 12, 5866.

  • Kiladis, G. N., C. D. Thorncroft, and N. M. J. Hall, 2006: Three-dimensional structure and dynamics of African easterly waves. Part I: Observations. J. Atmos. Sci., 63, 22122230.

    • Search Google Scholar
    • Export Citation
  • Knutson, D. R., and K. M. Weickmann, 1987: 30–60 day atmospheric oscillations: Composite life cycles of convection and circulation anomalies. Mon. Wea. Rev., 115, 14071436.

    • Search Google Scholar
    • Export Citation
  • Knutson, D. 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, 605623.

    • Search Google Scholar
    • Export Citation
  • Lebel, T., and Coauthors, 2010: The AMMA field campaigns: Multiscale and multidisciplinary observations in the West African region. Quart. J. Roy. Meteor. Soc., 136, 833.

    • Search Google Scholar
    • Export Citation
  • Leroux, S., and N. M. J. Hall, 2009: On the relationship between African Easterly waves and the African easterly jet. J. Atmos. Sci., 66, 23032316.

    • Search Google Scholar
    • Export Citation
  • Leroux, S., N. M. J. Hall, and G. N. Kiladis, 2010: A climatological study of transient–mean–flow interactions over West Africa. Quart. J. Roy. Meteor. Soc., 136, 397410.

    • Search Google Scholar
    • Export Citation
  • Leroux, S., N. M. J. Hall, and G. N. Kiladis, 2011: Intermittent African easterly wave activity in a dry atmospheric model: Influence of the extratropics. J. Climate, 24, 53785396.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., and D. L. Hartmann, 1998: Frictional moisture convergence in a composite life cycle of the Madden–Julian Oscillation. J. Climate, 11, 23872403.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., and D. L. Hartmann, 2000a: Modulation of eastern North Pacific hurricanes by the Madden–Julian Oscillation. J. Climate, 13, 14511460.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., and D. L. Hartmann, 2000b: Modulation of hurricane activity in the Gulf of Mexico by the Madden–Julian oscillation. Science, 287, 20022004.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., and A. H. Sobel, 2007: Idealized hot spot experiments with a general circulation model. J. Climate, 20, 908925.

  • Maloney, E. D., and J. Shaman, 2008: Intraseasonal variability of the West African monsoon and the Atlantic ITCZ. J. Climate, 21, 28982918.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., A. H. Sobel, and W. M. Hannah, 2010: Intraseasonal variability in an aquaplanet general circulation model. J. Adv. Model Earth Syst., 2, Article 5, 24 pp., doi:10.3894/JAMES.2010.2.5.

    • Search Google Scholar
    • Export Citation
  • Matthews, A. J., 2004: Intraseasonal variability over tropical Africa during northern summer. J. Climate, 17, 24272440.

  • Mekonnen, A., and W. B. Rossow, 2011: The interaction between deep convection and easterly waves over tropical North Africa: A weather state perspective. J. Climate, 24, 42754293.

    • Search Google Scholar
    • Export Citation
  • Mekonnen, A., C. D. Thorncroft, and A. R. Aiyyer, 2006: Analysis of convection and its association with African easterly waves. J. Climate, 19, 54055421.

    • Search Google Scholar
    • Export Citation
  • Mekonnen, A., C. D. Thorncroft, A. R. Aiyyer, and G. N. Kiladis, 2008: Convectively coupled Kelvin waves over tropical Africa during the boreal summer: Structure and variability. J. Climate, 21, 66496667.

    • Search Google Scholar
    • Export Citation
  • Mo, K. C., 2000: The association between intraseasonal oscillations and tropical storms in the Atlantic basin. Mon. Wea. Rev., 128, 40974107.

    • Search Google Scholar
    • Export Citation
  • Pohl, B., S. Janicot, B. Fontaine, and R. Marteau, 2009: Implication of the Madden–Julian oscillation in the 40-day variability of the West African monsoon. J. Climate, 22, 37693785.

    • Search Google Scholar
    • Export Citation
  • Raymond, D. J., and H. Jiang, 1990: A theory for long-lived mesoscale convective systems. J. Atmos. Sci., 47, 30673077.

  • Robinson, G. J., and K. I. Hodges, 2005: Cloud archive user service user guide. NERC Environmental Systems Science Centre, 18 pp. [Available online at http://badc.nerc.ac.uk/data/claus/userguide.pdf.]

  • Simmons, A. J., R. Mureau, and T. Petroliagis, 1995: Error growth and estimates of predictability from the ECMWF forecasting system. Quart. J. Roy. Meteor. Soc., 121, 17391771.

    • Search Google Scholar
    • Export Citation
  • Spiegel, M. R., 1992: Theory and Problems of Statistics. McGraw-Hill, 504 pp

  • Straub, K. H., and G. N. Kiladis, 2003: Extratropical forcing of convectively coupled Kelvin waves during austral winter. J. Atmos. Sci., 60, 526543.

    • Search Google Scholar
    • Export Citation
  • Sultan, B., and S. Janicot, 2003: West African monsoon dynamics. Part II: The “preonset” and the “onset” of the summer monsoon. J. Climate, 16, 34073427.

    • Search Google Scholar
    • Export Citation
  • Thorncroft, C. D., and B. J. Hoskins, 1994: An idealised study of African easterly waves. 1. A linear view. Quart. J. Roy. Meteor. Soc., 120, 953982.

    • Search Google Scholar
    • Export Citation
  • Thorncroft, C. D., N. M. J. Hall, and G. N. Kiladis, 2008: Three dimensional structure and dynamics of African easterly waves. Part III: Genesis. J. Atmos. Sci., 65, 35963607.

    • Search Google Scholar
    • Export Citation
  • Tromeur, E., and W. B. Rossow, 2010: Interaction of tropical deep convection with the large-scale circulation in the MJO. J. Climate, 23, 18371853.

    • Search Google Scholar
    • Export Citation
  • Vitart, F., and F. Molteni, 2010: Simulation of the Madden– Julian Oscillation and its teleconnections in the ECMWF forecast system. Quart. J. Roy. Meteor. Soc., 136, 842855.

    • Search Google Scholar
    • Export Citation
  • Wheeler, M. C., and K. M. Weickmann, 2001: Real-time monitoring and prediction of modes of coherent synoptic to intraseasonal tropical variability. Mon. Wea. Rev., 129, 26772694.

    • Search Google Scholar
    • Export Citation
  • 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, 19171932.

    • Search Google Scholar
    • Export Citation
  • Yanai, M., S. K. Esbensen, and J. H. Chu, 1973: Determination of bulk properties of tropical cloud clusters from large-scale heat and moisture budgets. J. Atmos. Sci., 30, 611627.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    EOF1 and EOF2 of the equatorial-averaged (15°S–15°N), 30–90-day bandpass-filtered CLAUS TB (black), ERA-I 850-hPa zonal wind (red), and ERA-I 200-hPa zonal wind (blue). The values in the vertical axis are normalized.

  • View in gallery

    MJO phase composites of 30–90-day variability over tropical North Africa from June to September. The 30–90-day CLAUS TB anomalies (shading) and 30–90-day ERA-I 850-hPa vector wind (vectors). The shading interval for TB anomalies is 0.4 K, and the reference vector is 0.75 m s−1. The minimum and maximum vector lengths are 0.1 and 0.75 m s−1, respectively. The TB anomalies with 90% significance are marked by stippling while vector wind anomalies with 90% significance are black. The trigger region is marked by a box.

  • View in gallery

    MJO phase composites of 30–90-day ERA-I 700-hPa EKE anomalies for June–September. The shading interval is 0.08 m2 s−2. Solid contours, with an interval of 0.08 m2 s−2, identify values outside of the shading range. Stippling represents data that are significant with 90% confidence. Only even phases are included. The trigger region is marked by a box.

  • View in gallery

    As in Fig. 3, but for total precipitable water. The shading interval is 0.2 mm. Dots represent data that are significant with 90% confidence.

  • View in gallery

    (a) The mean, unfiltered apparent heat source (K day−1) averaged in the trigger region from June to September (1989–2005). (b) Phase vs pressure plot of 30–90-day ERA-I apparent heat source anomalies for June–September in the trigger region. The shading interval is 0.02 K day−1. In general, the heating anomalies above and below ∼650 hPa (AEJ level) are significant at the 90% confidence level.

  • View in gallery

    MJO phase composites for 30–90-day ERA-I 200-hPa vector wind anomalies filtered for eastward zonal wavenumbers 0–10 (vectors) and 30–90-day CLAUS TB anomalies (shading). The reference vector at the top left of the top panel is 5 m s−1. Black vectors are significant at the 90% confidence level. The shading interval is 1 K. The trigger region is marked by a box.

  • View in gallery

    As in Fig. 6, but filtered for westward zonal wavenumbers −1 to −10.

  • View in gallery

    Phase vs longitude for anomalous 30–90-day ERA-I 400-hPa air temperature for eastward zonal wavenumbers 0 to 10 from 0° to 5°N. The plot shows the propagation of temperature anomalies as a function of phase.

  • View in gallery

    MJO phase composites of 30–90-day ERA-I 400-hPa air temperature anomalies for June–September. (a) All zonal wavenumbers. (b) Filtered for eastward zonal wavenumbers 0–10. The shading interval is 0.06 K. Dots represent data that are significant with 90% confidence.

  • View in gallery

    The 30–90-day variability of ERA-I 650-hPa zonal wind shown for June–September. (a) MJO phase-1 composite, (b) MJO phase-5 composite, and (c) phase cycle of 650-hPa zonal wind in the trigger region for total (blue solid), eastward (green dashed) and westward (red dotted) zonal wavenumbers. For (a) and (b), the shading represents zonal wind anomalies with an interval of 0.16 m s−1, while the contours represent zonal wind values ≤−7 m s−1 with a contour interval of 1 m s−1. Dots represent data that are significant with 90% confidence. In (a) and (b), the trigger region is marked by a box.

  • View in gallery

    MJO phase composites of 30–90-day ERA-I 400-hPa omega anomalies for June–September. Only phases 4 and 8 are shown. The shading interval is 1 × 10−3 Pa s−1. Dots represent data that are significant with 90% confidence. The trigger region is marked by a box.

  • View in gallery

    MJO phase composites for vertically integrated 30–90-day ERA-I moisture tendency anomalies. The shading interval is 4 × 10−7 kg m−2 s−1. Dots represent data that are significant with 90% confidence. The trigger region is marked by a box.

  • View in gallery

    (a) Phase vs pressure plot of vertically integrated 30–90-day ERA-I moisture budget term anomalies for June–September in the trigger region. Terms include moisture tendency (black solid), zonal moisture advection (green dashed), meridional moisture advection (red dotted), moisture convergence minus precipitation (blue double dot–dashed), and evaporation (gray dot–dashed). (b) Phase vs pressure plot of vertically integrated 30–90-day meridional moisture advection perturbation budget term anomalies for June–September in the trigger region. Terms include total meridional moisture advection (black solid), anomalous meridional wind advecting mean moisture (red dashed), mean meridional wind advecting anomalous moisture (blue dotted), and (green dot–dashed). Units are kg m−2 s−1.

  • View in gallery

    MJO phase composites for vertically integrated ERA-I for June–September. (a) Phases 1–3 for . (b) Phases 1–3 for . The shading interval is 1 × 10−6 kg m−2 s−1.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 283 124 17
PDF Downloads 193 101 11

The Influence of the MJO on Upstream Precursors to African Easterly Waves

View More View Less
  • 1 Department of Atmospheric Science, Colorado State University, Fort Collins, Colorado
Full access

Abstract

The Madden–Julian oscillation (MJO) produces alternating periods of increased and reduced precipitation and African easterly wave (AEW) activity in West Africa. This study documents the influence of the MJO on the West African monsoon system during boreal summer using reanalysis and brightness temperature fields. MJO-related West African convective anomalies are likely induced by equatorial Kelvin and Rossby waves generated in the Indian Ocean and West Pacific by the MJO, which is consistent with previous studies. The initial modulation of tropical African convection occurs upstream of West Africa, near the entrance of the African easterly jet (AEJ). Previous studies have hypothesized that an area to the east of Lake Chad is an initiation region for AEWs. Called the “trigger region” in this study, this area exhibits significant intraseasonal convection and wave activity anomalies prior to the wet and dry MJO phases in the West African monsoon region.

In the trigger region, cold tropospheric temperature anomalies and high precipitable water, as well as an eastward extension of the African easterly jet, appear to precede and contribute to the wet MJO phase in West Africa. An anomalous stratiform heating profile is observed in advance of the wet MJO phase with anomalous PV generation maximized at the jet level. The opposite behavior occurs in advance of the dry MJO phase. The moisture budget is examined to provide further insight as to how the MJO modulates and initiates precipitation and AEW variability in this region. In particular, meridional moisture advection anomalies foster moistening in the trigger region in advance of the wet MJO phase across West Africa.

Corresponding author address: Ghassan Alaka, Department of Atmospheric Science, Colorado State University, 1371 Campus Delivery, Fort Collins, CO 80523-1371. E-mail: gusalaka@atmos.colostate.edu

Abstract

The Madden–Julian oscillation (MJO) produces alternating periods of increased and reduced precipitation and African easterly wave (AEW) activity in West Africa. This study documents the influence of the MJO on the West African monsoon system during boreal summer using reanalysis and brightness temperature fields. MJO-related West African convective anomalies are likely induced by equatorial Kelvin and Rossby waves generated in the Indian Ocean and West Pacific by the MJO, which is consistent with previous studies. The initial modulation of tropical African convection occurs upstream of West Africa, near the entrance of the African easterly jet (AEJ). Previous studies have hypothesized that an area to the east of Lake Chad is an initiation region for AEWs. Called the “trigger region” in this study, this area exhibits significant intraseasonal convection and wave activity anomalies prior to the wet and dry MJO phases in the West African monsoon region.

In the trigger region, cold tropospheric temperature anomalies and high precipitable water, as well as an eastward extension of the African easterly jet, appear to precede and contribute to the wet MJO phase in West Africa. An anomalous stratiform heating profile is observed in advance of the wet MJO phase with anomalous PV generation maximized at the jet level. The opposite behavior occurs in advance of the dry MJO phase. The moisture budget is examined to provide further insight as to how the MJO modulates and initiates precipitation and AEW variability in this region. In particular, meridional moisture advection anomalies foster moistening in the trigger region in advance of the wet MJO phase across West Africa.

Corresponding author address: Ghassan Alaka, Department of Atmospheric Science, Colorado State University, 1371 Campus Delivery, Fort Collins, CO 80523-1371. E-mail: gusalaka@atmos.colostate.edu

1. Introduction

A complete understanding of intraseasonal variability over West Africa during boreal summer requires analysis of how the Madden–Julian oscillation (MJO) modulates the West African monsoon and, consequently, African easterly waves (AEWs). Previous studies have provided evidence of 30–90-day spectral peaks in various atmospheric variables over tropical North Africa (Sultan and Janicot 2003; Janicot et al. 2011; de Coëtlogon et al. 2010). The 30–90-day variability of rainfall and winds in tropical West Africa and the Atlantic intertropical convergence zone (ITCZ) is at least partly governed by the MJO (Matthews 2004; Maloney and Shaman 2008; Pohl et al. 2009; Janicot et al. 2009). While some studies have found a relatively weak relationship between the MJO and tropical North Africa (Maloney and Hartmann 2000a; Knutson and Weickmann 1987; Annamalai and Slingo 2001; Wheeler and Weickmann 2001; Knutson et al. 1986), other recent work suggests that the MJO may have a significant remote influence on the region during boreal summer, including a modulation of the West African monsoon (e.g., Matthews 2004; Janicot et al. 2009, 2011; Leroux and Hall 2009) and Gulf of Guinea SSTs (de Coëtlogon et al. 2010). These studies have shown that increased convection in the West African monsoon coincides with MJO convective initiation over the Indian Ocean. It is hypothesized that Kelvin and Rossby waves, while traveling in opposite zonal directions, force variability in the West African monsoon region at approximately the same time (Matthews 2004; Janicot et al. 2009). This modulation by the MJO extends to Atlantic tropical cyclones (Maloney and Hartmann 2000b; Mo 2000; Matthews 2004; Maloney and Shaman 2008; Camargo et al. 2009). Because the MJO may be predictable a few weeks in advance, understanding how and why the MJO impacts the West African monsoon may have a profound influence on Atlantic tropical cyclone prediction (Vitart and Molteni 2010).

This study aims to identify how the MJO induces convection and AEW variability in tropical North Africa. In particular, we concentrate on the role of the MJO in inducing convective anomalies to the east of Lake Chad. Some early AEW studies cited the instability of the AEJ as the mechanism most responsible for AEW initiation, and did not place importance on East African convection (Burpee 1972, 1974; Albignat and Reed 1980). However, Carlson (1969a,b) proposed the importance of convection associated with topographical features east of ∼10°E to AEWs. More recent studies hypothesized that upstream convection, in an East African region including the Darfur Mountains (15°N, 23°E) and the Ethiopian Highlands (10°N, 35°E), initiates AEWs near the entrance of the AEJ (Hall et al. 2006; Kiladis et al. 2006; Mekonnen et al. 2006; Thorncroft et al. 2008, hereafter THK08; Leroux and Hall 2009). Recently, based on climatological analysis of observational [e.g., the Cloud Archive User Service (CLAUS)] and reanalysis datasets, Mekonnen et al. (2006) showed that AEWs are forced in association with convective activity over Darfur and, occasionally, the Ethiopian Highlands. Kiladis et al. (2006) also noted the importance of weak convection over western and central Sudan that precedes AEW activity in West Africa.

In a modeling study, THK08 used a primitive equation model to analyze how effective different heating profiles (shallow, deep, stratiform) and their horizontal locations are at initiating AEW variability when given a common basic state AEJ. When heating profiles were applied at (15°N, 20°E), which is in the Darfur region near the African easterly jet (AEJ) entrance, strong divergence anomalies near the AEJ level produced high-amplitude wave disturbances downstream, particularly when using shallow convection and stratiform heating profiles. Therefore, the entrance region of the AEJ, and the influence of the MJO on this location, will be a significant focus throughout our study. By varying the convective trigger location and profile, THK08 showed that the AEJ entrance was the most sensitive region for AEW initiation. A modeling study by Hall et al. (2006) also emphasized the importance of upstream convection based on the same dry model that THK08 utilized. THK08 and Hall et al. (2006) assert that the weak instability and the short length of the AEJ indicate that the AEJ alone is insufficient in supporting observed wave growth. As further evidence, Hsieh and Cook (2005) promoted the idea that moist convection is a necessary condition for AEWs, while the AEJ is not.

Later, Leroux and Hall (2009) varied the basic-state structure of the jet while maintaining a deep convective heating profile, which confirmed the findings of THK08. By comparing model runs with many different basic states, Leroux and Hall (2009) conclude that in a simple modeling framework, the convective trigger is necessary, but not sufficient, and that intraseasonal variations in the AEJ determine whether or not a wave response will develop from the upstream convective trigger. Strong shear and PV reversals over an extended region near the AEJ are two conditions that produce stronger AEW responses (Leroux and Hall 2009).

The MJO modulation of convection and related quantities in the initiation region near the AEJ entrance will be one focus of this study to determine the extent to which this region may trigger intraseasonal AEW and convection variability in the West African monsoon during MJO events. More generally, Leroux et al. (2010) found that intraseasonal variability of AEWs tends to be preceded by anomalous convection in the Darfur region. We hypothesize that modulation of convection upstream of the AEJ by the MJO may be a key mechanism for producing intraseasonal AEW and convective variability in the West African monsoon. Of particular interest is how moisture and eddy kinetic energy (EKE) vary on MJO time scales in the initiation region and over the greater West African monsoon region.

Section 2 describes the data used and methodology, including details of the MJO index, filtering, and compositing technique. In section 3, we analyze the intraseasonal variability of tropical North Africa, and diagnose MJO precursor signals upstream of the African easterly jet. Section 4 examines the relationship between MJO heating in the Indo-Pacific warm pool and tropical North Africa via remote influences propagated by equatorial waves. Section 5 investigates, from a process standpoint, how these remote influences modulate large-scale dynamical and thermodynamical variables over tropical North Africa, with particular emphasis on processes in a trigger region upstream of the AEJ. Section 6 presents a discussion of results and conclusions.

2. Data and methodology

Most of the fields analyzed in this study are from the European Centre for Medium-Range Weather Forecasts (ECMWF) Interim Re-Analysis (ERA-I; e.g., Simmons et al. 1995) dataset. ERA-I features a T255 spherical-harmonic representation of basic dynamical fields that is reduced to a horizontal resolution of 1.5° for the fields we use here. We use model output with four times daily temporal resolution from 1989 to 2005, concentrating on the boreal summer months of June–September. We note that observational data included in the reanalysis may be sparse for the African region, although ERA-I should be satisfactory on synoptic and larger scales.

The CLAUS brightness temperature (TB) dataset (Robinson and Hodges 2005) is utilized as a proxy for convective activity. The CLAUS archive is produced from International Satellite Cloud Climatology Project (ISCCP) level B3 products in a 10–12-μm thermal infrared window (channel 2). Hodges et al. (2000) resampled the TB data to a temporal resolution of 3 h and a spatial resolution of 0.5°. The CLAUS dataset allows for an excellent representation of convective variability in tropical North Africa, given that important topographically forced convective features may not be well represented by coarser-resolution datasets. Some caution should be used in interpreting brightness temperature data, as TB values for cirrus may be similar to those for deep convection.

To isolate how certain fields are modulated by the MJO, data are filtered to intraseasonal time scales via a linear nonrecursive digital bandpass filter with half-power points at 30 and 90 days (comprising 59-point high-pass and low-pass filters). The seasonal cycle is also removed by averaging each day of the year across the long-term record and subtracting this climatological seasonal cycle from the data.

Some fields are also spatially filtered by zonal wavenumber as a means to analyze the modulation of tropical North African fields via eastward- and westward-propagating teleconnected signals from the Indo-Pacific warm pool. Since the teleconnected wave responses are concentrated at low zonal wavenumbers, data are filtered to include wavenumbers 0–10 for eastward-propagating disturbances, including Kelvin waves, and wavenumbers −1 to −10 for westward-propagating disturbances, including Rossby waves. Zonal wavenumber zero represents the zonal mean and is included with eastward-propagating disturbances to capture the potentially fast eastward propagation of dry intraseasonal Kelvin waves forced by MJO heating and their modulation of the tropical zonal mean (e.g., Bantzer and Wallace 1996). Dry Kelvin wave fronts forced by Indo-Pacific heating anomalies can circumnavigate the globe in about 8 days, and hence may produce a strong zonal mean signature in intraseasonal filtered data. No westward-propagating counterpart exists.

Boreal summer (1 June–30 September) composites that describe the evolution of an MJO life cycle are constructed based on an index derived from empirical orthogonal functions (EOFs). Multivariate EOF analysis is performed on ERA-I 850- and 200-hPa zonal winds and CLAUS TB in a closely related method to that employed by Wheeler and Hendon (2004, hereafter WH04). All data are 30–90-day bandpass filtered and meridionally averaged from 15°S to 15°N in the computation of combined EOFs. The first two EOFs explain 30.39% and 24.77% of the total combined variance, respectively, with the first EOF in quadrature with the second EOF (Fig. 1). The amplitude for the first EOF peaks 11 days before (calculated by cross correlating the first two principal components) and 90° to the west of the second EOF, representing eastward propagation.

Fig. 1.
Fig. 1.

EOF1 and EOF2 of the equatorial-averaged (15°S–15°N), 30–90-day bandpass-filtered CLAUS TB (black), ERA-I 850-hPa zonal wind (red), and ERA-I 200-hPa zonal wind (blue). The values in the vertical axis are normalized.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

From the principal components (PCs) of the two leading EOFs, an “MJO index” is constructed that entrains daily information about MJO amplitude and phase. The amplitude and phase are calculated by (1) and (2), using the same method as WH04:
e1
e2
Sign information of individual PCs is retained in the computation of phase. The behavior of the PC time series derived in this study is similar to Fig. 5 in WH04.

The MJO cycle is split into eight phases of equal angular extent, with each phase representing enhanced MJO activity at a particular part of the world. With one MJO cycle taking approximately 40 days, each MJO phase represents ∼5 days. In accordance with WH04, only days with normalized amplitude over 1.0 are included in composites below in order to eliminate weak MJO events. Importantly, the data record includes at least 151 significant MJO days during boreal summer for each phase, which provides each phase composite with enough days to produce anomaly fields that are significantly different from zero.

A Student’s t test at the 90% confidence interval (e.g., Spiegel 1992) is performed in many of the analyses described below to discern whether or not the low-amplitude anomalies over tropical North Africa are significant. Degrees of freedom are calculated by dividing the average number of days in an MJO phase (i.e., 159) by the characteristic number of days that the MJO exists in a given phase (i.e., 5). Thus, degrees of freedom (i.e., 32) represent the fewest possible individual MJO events for a single phase.

3. The influence of the MJO on North Africa

General aspects of how the MJO impacts the West African monsoon during the boreal summer months of June through September are investigated, with precursor signals upstream of the AEJ in East Africa being emphasized, as they are the focus of this study. We only briefly describe general aspects of the influence of the MJO on West Africa because the general intraseasonal variability of the West African monsoon system has been summarized in previous studies (Leroux et al. 2010; Lebel et al. 2010; de Coëtlogon et al. 2010; Janicot et al. 2011). Using ERA-I and CLAUS fields, we will illustrate the relationship between EKE and convection in West Africa, convection in East Africa, and the MJO.

First, with the MJO index described in section 2, global MJO phase composites of CLAUS TB data were generated to verify the global structure of the MJO during boreal summer (see WH04, their Fig. 9). Overall, the composites feature an eastward-propagating region of positive convection anomalies, which initiates in the western Indian Ocean around 60°E (not shown). As the center of increased convection propagates east of 120°E about 20 days after initiation, a region of reduced convection initiates around 60°E and propagates to the east in the same manner as the envelope of increased convection. In both cases, northward propagation into the Indian subcontinent is attributed to the interaction of the MJO with the Indian monsoon.

Of initial focus is the boreal summer intraseasonal variability of the TB and low-level winds over tropical North Africa (Fig. 2). Since phase 1 represents the onset of MJO convection in the Indian Ocean, Fig. 2 reveals that tropical North Africa is convectively enhanced at the same time that the MJO initiates in the Indian Ocean (cf. with Fig. 6 below). In phase 1, significant negative 30–90-day TB anomalies are widespread over much of tropical North Africa, which are accompanied by significant low-level 30–90-day EKE anomalies in phase 2 (see Fig. 3 below).

Fig. 2.
Fig. 2.

MJO phase composites of 30–90-day variability over tropical North Africa from June to September. The 30–90-day CLAUS TB anomalies (shading) and 30–90-day ERA-I 850-hPa vector wind (vectors). The shading interval for TB anomalies is 0.4 K, and the reference vector is 0.75 m s−1. The minimum and maximum vector lengths are 0.1 and 0.75 m s−1, respectively. The TB anomalies with 90% significance are marked by stippling while vector wind anomalies with 90% significance are black. The trigger region is marked by a box.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

Fig. 3.
Fig. 3.

MJO phase composites of 30–90-day ERA-I 700-hPa EKE anomalies for June–September. The shading interval is 0.08 m2 s−2. Solid contours, with an interval of 0.08 m2 s−2, identify values outside of the shading range. Stippling represents data that are significant with 90% confidence. Only even phases are included. The trigger region is marked by a box.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

Due to this pattern, phase 1 will be referred to as the “wet” MJO phase over tropical North Africa. Importantly, a region in the vicinity of the Darfur Mountains (15°N, 23°E) exhibits significant negative intraseasonal TB anomalies during phase 1. Significant negative intraseasonal TB anomalies first appear in this region, which is upstream of the AEJ entrance, in phase 7. Consistent with observational (Berry and Thorncroft 2005; Mekonnen et al. 2006) and modeling studies (Leroux et al. 2010), anomalous upstream latent heating precedes, and may help initiate, periods of enhanced AEW activity. The presence of low-level southerly flow anomalies near the right entrance of the AEJ (see Fig. 10 for general location) in phases 6 and 7 hints that meridional moisture advection may play an important role in priming the area for increased convection since the mean moisture gradient is directed southward in this region.

Examination of intraseasonal 700-hPa EKE (Fig. 3) and total precipitable water (TPW; Fig. 4) anomalies in phase 8 indicate significant positive column moisture anomalies and enhanced wave activity extending from the Darfur region to northern Ethiopia prior to the wet MJO phase, which would both tend to support and be indicative of anomalous convection in the vicinity of the AEJ entrance. The strongest moisture anomalies occur in the vicinity of the right AEJ entrance. Here, EKE is computed by removing the running 5-day mean from the zonal and meridional wind fields. Figure 3 suggests that upstream EKE anomalies lead widespread West African EKE anomalies, again hinting that upstream precursors associated with the MJO help initiate downstream AEWs. The 700-hPa EKE anomalies spread westward out of the trigger region at about 10 m s−1 (approximately equal to the AEW propagation speed, e.g., Mekonnen et al. 2008), which is supported by examination of the odd phases (not shown). While the intraseasonal TPW does not vary much over West Africa, upstream column moisture anomalies exceed 1 mm when upstream convection is increased (phase 8; Fig. 4). Since tropical water vapor and precipitation are strongly correlated (Bretherton et al. 2004), the colocation of column moisture and convection anomalies is an expected result.

Fig. 4.
Fig. 4.

As in Fig. 3, but for total precipitable water. The shading interval is 0.2 mm. Dots represent data that are significant with 90% confidence.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

Phase 5 (Fig. 2) is dominated by significant positive intraseasonal TB anomalies over tropical North Africa and coincides with a reduction of MJO convection in the Indian Ocean (see Fig. 6). For that reason, phase 5 will be referred to as the “dry” MJO phase over tropical North Africa. In the opposite sense of phase 1, reduced convection in tropical North Africa is preceded by reduced upstream convection over the Darfur region (15°N, 20°E). Anomalously reduced convection in the Darfur region commences during phase 4, or before the dry MJO phase. Given the weighting of convective activity toward eastern North Africa in phase 4 before anomalies shift downstream at later phases, this supports the notion of convective activity initiating upstream and spreading downstream via the AEJ. The relationship between upstream and downstream regions is more tenuous in advance of the dry MJO phase than the wet MJO phase; however, since no significant reduced convection anomalies appear in the trigger box during Phase 3, upstream anomalies may be isolated from those to the west. Negative low-level EKE anomalies (Fig. 3) and negative TPW anomalies (Fig. 4) are widespread near the Darfur region in phase 4, in advance of the dry MJO phase.

The pattern of intraseasonal variability located to the east of Lake Chad in Figs. 24, particularly the presence of precursor moisture, convection, and EKE anomalies in the trigger region in advance of the wet MJO phase, suggests that convective anomalies upstream of the AEJ may foster anomalous convection and AEW activity downstream over West Africa. This is consistent with observational studies (Mekonnen et al. 2006; Berry and Thorncroft 2005), as well as dry model-based studies (Hall et al. 2006; THK08). This region may be important for detecting upstream convective heating anomalies that seed the AEJ, and, as a result we refer to it as the “trigger region.” The trigger region (10.5°–24°N, 16.5°–37.5°E) is prescribed to capture intraseasonal variability upstream of the AEJ. Two important topographical features are located within the trigger region: the Darfur Mountains and the Ethiopian Highlands.

Recent studies suggest that convective activity in the trigger region may be initiated in the Indian Ocean. Tromeur and Rossow (2010) asserted that convective activity in East Africa is linked to various “Weather States” in the Indian Ocean and West Pacific Ocean, while Mekonnen and Rossow (2011) suggested that East African convection originates in the Arabian Sea and initiates AEWs west of the Ethiopian Highlands. However, this convection is small scale and disorganized until it is west of the Ethiopian Highlands (Mekonnen and Rossow 2011), giving credence to focusing on the trigger region.

A key question is how much intraseasonal convective variability the MJO explains in the trigger region and related EKE anomalies downstream. To examine this, a “TB index” was created by averaging 30–90-day bandpass-filtered TB within the trigger region. Fifty-three significant events (25 negative, 28 positive) were identified, having TB index extrema greater than one standard deviation. For negative TB events the composite minimum values within the trigger region for the MJO and TB indices are −2.16 and −5.16 K, respectively. For positive TB events, the maximum values within the trigger region for the MJO and TB indices are 2.56 and 5.12 K, respectively. Thus, the MJO TB anomalies are about 40%–50% of the amplitude of intraseasonal TB anomalies derived from a local TB index.

Intraseasonal 700-hPa EKE anomalies within a box over the West African coast (6°–19.5°N, 22.5°–1.5°W) are also compared for composite MJO and TB index events. The area-averaged EKE is assessed for a lag of 5 days after the TB index minimum when EKE is highest (corresponding to phase 2 for the MJO index). The area-averaged EKE anomalies for the MJO and TB indices are 1.52 and 1.21 m2 s−2, respectively, with similar amplitude but opposite-signed anomalies associated with reduced convection periods. Composites from the MJO index may contain a stronger amplitude EKE signal than the TB index, even though triggering region convection anomalies are lower, because convection in the trigger region does not explain all downstream intraseasonal variability. Variations in the jet and influences from midlatitudes (e.g., Leroux et al. 2010, 2011) may also help modulate the amplitude of AEW activity on intraseasonal time scales.

To gain a better understanding of intraseasonal variability in convective heating profiles in the trigger region, the apparent heat source (Q1; Yanai et al. 1973) is presented in Fig. 5. The apparent heat source is given by
e3
where s = cpT + gz is the dry static energy. Terms on the right-hand side of Eq. (3) are computed from ERA-I output fields. THK08 showed that the amplitude of AEW activity downstream in the AEJ was sensitive to the exact structure of diabatic heating. The PV generation at the maximum level of the AEJ (∼650 mb) would be maximized for a higher vertical mode heating structure due to stronger heating gradients (Raymond and Jiang 1990).
Fig. 5.
Fig. 5.

(a) The mean, unfiltered apparent heat source (K day−1) averaged in the trigger region from June to September (1989–2005). (b) Phase vs pressure plot of 30–90-day ERA-I apparent heat source anomalies for June–September in the trigger region. The shading interval is 0.02 K day−1. In general, the heating anomalies above and below ∼650 hPa (AEJ level) are significant at the 90% confidence level.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

Above 800 hPa, the mean heating profile is weakly stratiform (Fig. 5a), with slight cooling occurring below the AEJ (−0.1 K day−1) and heating located above the AEJ (0.5 K day−1). Such a profile induces heating gradients across the AEJ, which THK08 found to force a strong wave response. However, THK08 utilized a maximum heating rate an order of magnitude greater than the values in Fig. 5a. This might be explained by the fact that THK08 used a dry, idealized model, and so the forcing had to be stronger to observe a reasonable response.

The weak stratiform heating profile, which straddles the AEJ entrance, may be significantly altered by intraseasonal heating anomalies in the trigger region (Fig. 5b). For example, phase 6 displays a shallow heating profile with heating anomalies occurring above and below the level of the AEJ that tend to add destructively to the mean heating profile (Fig. 5a). In fact, the intraseasonal anomalies at 700 hPa erode the slight cooling observed in the mean heating profile. By phase 7 or 8 (prior to the wet MJO phase), the trigger region features an anomalous stratiform heating profile that adds constructively to the mean heating profile. A stratiform heating profile was shown by THK08 to engender a strong AEW response downstream; and the significant enhancement of that profile prior to the wet MJO phase could contribute to a stronger AEW initialization. THK08 also concluded that shallow heating–cooling contributes to the strongest wave response, so the variations in heating just below the AEJ may be more important. Variations in intraseasonal Q1 in Fig. 5b represent about 50% of the mean near 700 hPa and 25% of the mean (Fig. 5a) near 500 hPa.

4. Equatorial wave teleconnections: Large-scale context

With the evidence of significant 30–90-day variability in the trigger region presented in section 3, we investigate if equatorial waves attributed to MJO heating impact this upstream region. Intraseasonal Kelvin wave fronts have been shown to propagate rapidly into the African monsoon region after emission by the MJO in the Indo-Pacific warm pool, and Rossby waves impinge from the east (e.g., Matthews 2004). Hendon and Salby (1994) identify two equatorial wave responses to an MJO heating anomaly: a coupled response that resembles a coupled Rossby–Kelvin wave (e.g., Gill 1980), and a radiating response that appears as a Kelvin wave propagating into the Western Hemisphere away from MJO heating. The Kelvin wave response can be diagnosed in the upper troposphere, with flow anomalies at lower levels being of opposite sign (Hendon and Salby 1994). Once the MJO heating anomaly has been established, upper-level anticyclonic Rossby gyres intensify to the west of MJO heating (Knutson and Weickmann 1987; Hendon and Salby 1994). Therefore, anomalous upper-tropospheric winds are used to verify the role of MJO-induced equatorial waves for affecting the flow over tropical North Africa during boreal summer, with emphasis on how they may impact the trigger region. This initial analysis will provide evidence that the MJO modulates tropical North Africa via equatorial waves, which provides the context for the physical mechanisms proposed to affect convection in the triggering region in section 5. Links between equatorial waves and thermodynamic–dynamic modulation of large-scale conditions in the trigger region will be used to diagnose how the MJO actually triggers convective variability in the trigger region.

a. Kelvin wave response

The propagation of Kelvin waves from MJO heating in the Indo-Pacific warm pool to Africa is analyzed by filtering ERA-I 30–90-day bandpass-filtered 200-hPa vector wind anomalies for eastward-propagating disturbances (Fig. 6). In phase 1, the initiation of MJO heating spawns anomalous westerlies in the upper troposphere. These upper-level westerlies are located to the east of MJO heating, and propagate rapidly eastward to the east of the date line after phase 3, consistent with the notion of a Kelvin wave response to heating that expands to the east of the heating center as documented by previous studies (e.g., Hendon and Salby 1994; Maloney and Hartmann 1998). By phase 5, the 30–90-day upper-level westerly vector wind anomalies arrive at the African coast. The westerly anomalies propagate through equatorial Africa in phases 5–8, and then reappear in the Indian Ocean. The propagation of intraseasonal Kelvin waves forced by the MJO into Africa is consistent with other studies (Hendon and Salby 1994; Maloney and Hartmann 1998; Straub and Kiladis 2003; Matthews 2004; Mekonnen et al. 2008; Janicot et al. 2011). As we will show, these westerlies aloft are accompanied by a warm equatorial troposphere consistent with a Kelvin wave response, and are associated with a cold off-equatorial troposphere in the triggering region of east Africa.

Fig. 6.
Fig. 6.

MJO phase composites for 30–90-day ERA-I 200-hPa vector wind anomalies filtered for eastward zonal wavenumbers 0–10 (vectors) and 30–90-day CLAUS TB anomalies (shading). The reference vector at the top left of the top panel is 5 m s−1. Black vectors are significant at the 90% confidence level. The shading interval is 1 K. The trigger region is marked by a box.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

Anomalous negative MJO convective activity in the Indian Ocean is associated with upper-tropospheric easterly anomalies in phase 5. These anomalies spread to the east, arriving at the African coast by phase 1, which will be shown to be consistent with the propagation of cold equatorial midtropospheric temperature anomalies (see Fig. 8). Cold Kelvin waves generated by negative MJO heating anomalies in phase 5 propagate eastward arrive in tropical Africa in phase 1, and may help maintain positive convective anomalies over the region.

b. Rossby wave response

Several studies have cited the potential importance of Rossby waves in communicating MJO anomalies into tropical Africa (Janicot et al. 2009; Matthews 2004). Janicot et al. (2009) assert that Gill-type Rossby circulations originate from Indian heating and propagate into North Africa. Anomalous meridional flow and anomalous moisture advection are observable consequences of Rossby wave propagation into tropical North Africa. This work is different from that of Janicot et al. (2009) in that we focus on the MJO influence on the trigger region upstream of the AEJ.

The impact of Rossby wave–type disturbances on tropical North Africa is examined by filtering ERA-I 30–90-day 200-hPa vector wind anomalies for westward-propagating disturbances (Fig. 7). With the onset of MJO heating in the Indian Ocean, an upper-level anticyclone appears over India in phase 1. In phase 2, this gyre is located just to the west of its previous location, or over the northern Indian Ocean. In phase 3, two gyres are evident, with one located over Tibet and the other over Sudan. By phase 4, these centers have shifted west and resemble one elongated anticyclone, spanning from the Greenwich Meridian to 80°E, consistent with Matthews (2004) and Janicot et al. (2009). This elongation is representative of westward Rossby wave propagation into northern Africa. The westward propagation is more apparent by following significant easterly anomalies just north of the equator. Importantly, Rossby wave signatures arrive in the trigger region two phases, or about 10 days, in advance of Kelvin wave anomalies, which is consistent with previous studies. The disparate arrival times in the trigger region further suggests that Kelvin and Rossby waves modulate the trigger region in different ways.

Fig. 7.
Fig. 7.

As in Fig. 6, but filtered for westward zonal wavenumbers −1 to −10.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

The behavior associated with Rossby wave propagation spawned by negative Indian monsoon heating is similar, but with opposite sign. In phase 5, the cyclonic gyre is located over Tibet–India. By phase 6, the cyclonic center is located over India and the northern Indian Ocean. By phase 7, two gyres are apparent (Libya and Afghanistan). Phase 8 shows these gyres propagating west. The westward propagation is apparent by following significant westerly anomalies just north of the equator. In the next section, the modulation of specific fields in tropical North Africa in the context of equatorial waves is analyzed to better understand the relationship between MJO heating in the Indian Ocean and precursor disturbances within the trigger region.

5. Modulation of fields in the trigger region

In this section, we examine variations in large-scale environmental conditions in the trigger region that support generation of convective anomalies. These include precursor temperature signals, extension of the AEJ, and moisture advection.

a. Upper-level temperature

Negative upper-level air temperature anomalies decrease the static stability, which supports more vigorous deep convection. Upper-level air temperature anomalies in the West African monsoon region may be associated with equatorial waves (Matthews 2004; Maloney and Shaman 2008). Thus, air temperature anomalies are filtered by zonal wavenumber and frequency to examine the contributions of eastward- and westward-propagating disturbances, suggesting their respective contributions in modulating the static stability in the trigger region. Although the static stability is not specifically analyzed, if all else remain unchanged, upper-tropospheric temperature anomalies should modulate the lapse rate.

Figure 8 displays the 0°–5°N average of anomalous 30–90-day 400-hPa air temperature for eastward zonal wavenumbers (0–10). When viewed in conjunction with eastward-propagating vector wind anomalies (Fig. 6), the temperature anomalies in the Western Hemisphere are consistent with intraseasonal Kelvin waves emanating from the Indo-Pacific warm pool into North Africa. Specifically, positive (negative) temperature anomalies are associated with upper-level westerly (easterly) wind anomalies. Positive MJO convective anomalies initiate near the equator at approximately 75°–90°E near phase 1, with positive upper-tropospheric temperature anomalies growing to values of at least 0.32 K. The opposite is the case for the negative phase of the MJO, where negative temperature anomalies initiate in the Indian Ocean in phase 5. In both cases, these anomalies propagate slowly to the east at approximately 5 m s−1 until they reach the date line. Anomalies accelerate eastward to the east of the date line, with average phase speeds greater than 15 m s−1. Upon the phase speed increase, Kelvin wave disturbances propagate rapidly across the equatorial Western Hemisphere (including Africa). To clarify, the MJO forces upper-level temperature anomalies in association with deep convection in the Indo-Pacific warm pool, which are then propagated via Kelvin waves along the equator. Upon reaching Africa, these upper-level temperature anomalies modulate the static stability, and, as a result, support anomalous convection in Africa. Figure 8 suggests that 0°–5°N cold temperature anomalies initiated by the negative phase of the MJO (phase 5) arrive in northern Africa by phase 1 and maximize in phase 2, while warm temperature anomalies arrive in northern Africa by phases 5 and 6.

Fig. 8.
Fig. 8.

Phase vs longitude for anomalous 30–90-day ERA-I 400-hPa air temperature for eastward zonal wavenumbers 0 to 10 from 0° to 5°N. The plot shows the propagation of temperature anomalies as a function of phase.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

The response of temperature anomalies in the trigger region is analyzed in Fig. 9. While Fig. 8 provided evidence of negative (positive) equatorial temperature anomalies arriving in tropical North Africa during the wet (dry) MJO phase at the earliest, the off-equatorial response proves more nuanced. Figure 9a shows the time-filtered (but not space filtered) temperature anomalies at phases 3, 4, 7, and 8 of an MJO life cycle, while Fig. 9b shows the anomalies from Fig. 9a filtered for eastward-propagating disturbances. In phases 4 and 8 (Fig. 9a) of ERA-I 400-hPa T30–90, significant, off-equatorial anomalies dominate and extend into the trigger region. In particular, the anomalous positive 400-hPa T30–90 in phase 4 (Fig. 9a) support an increase in static stability in advance of the dry MJO phase, while the anomalous cold anomalies in phase 8 (Fig. 9a) correspond to a decrease in static stability just prior to the wet MJO phase. The reduction in static stability observed near the trigger region in phase 8 (Fig. 9a) would support more vigorous convection, which favors strong upstream precursor disturbances that may seed the AEJ.

Fig. 9.
Fig. 9.

MJO phase composites of 30–90-day ERA-I 400-hPa air temperature anomalies for June–September. (a) All zonal wavenumbers. (b) Filtered for eastward zonal wavenumbers 0–10. The shading interval is 0.06 K. Dots represent data that are significant with 90% confidence.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

Wavenumber decomposition between eastward and westward wavenumbers indicates that the observed temperature signal is most likely attributed to large-scale, eastward propagation (Fig. 9b), although the contribution from westward-propagating disturbances is not negligible. The explanation behind off-equatorial, eastward-propagating T30–90 is unclear. Maloney and Sobel (2007) showed a similar response in idealized hot spot experiments using a general circulation model. They showed that a forced Kelvin wave in response to a switch-on of an SST hotspot produces remote positive temperature anomalies along the equator and negative off-equatorial anomalies (their Fig. 4, days 40 and 50). The equatorial temperature anomalies decay more rapidly than their off-equatorial counterparts, resulting in a remnant off-equatorial response much like that shown in Fig. 9b (cf. top two panels). Chauvin et al. (2010) observed that the summer temperature field over North Africa was modulated by the southward penetration of upper-level Rossby wave disturbances on intraseasonal time scales. Given that significant temperature anomalies in Fig. 9 extend north of 35°N, an interaction with the midlatitudes is also entirely plausible.

b. African easterly jet extension

The AEJ is important for the structure, propagation, and maintenance of AEWs west of 10°E (Thorncroft and Hoskins 1994). While several studies have found that the AEJ is not necessary for AEW initiation over eastern Africa (Hsieh and Cook 2005; Hall et al. 2006; Mekonnen et al. 2006), the AEJ is vital for propagating AEW precursors downstream. The influence of the MJO on the AEJ is examined through the 30–90-day ERA-I 650-hPa zonal wind (Fig. 10). The entrance region is of particular interest since an extension of the AEJ to the east will not only increase the number of precursor disturbances that enter the AEJ, but would also tend to enhance baroclinic and barotropic energy conversions in the trigger region and allow for more wave growth since disturbances have more residence time in the jet. Furthermore, the strength and areal coverage of the PV gradient reversal may be important for producing a stronger AEW response to a convective precursor (Leroux and Hall 2009).

Fig. 10.
Fig. 10.

The 30–90-day variability of ERA-I 650-hPa zonal wind shown for June–September. (a) MJO phase-1 composite, (b) MJO phase-5 composite, and (c) phase cycle of 650-hPa zonal wind in the trigger region for total (blue solid), eastward (green dashed) and westward (red dotted) zonal wavenumbers. For (a) and (b), the shading represents zonal wind anomalies with an interval of 0.16 m s−1, while the contours represent zonal wind values ≤−7 m s−1 with a contour interval of 1 m s−1. Dots represent data that are significant with 90% confidence. In (a) and (b), the trigger region is marked by a box.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

The maximum extension of the jet (Fig. 10a) coincides with the wet MJO phase and convective maximum over tropical North Africa (Fig. 2). In phase 1, the −10 m s−1 contour extends to the east of Lake Chad. In addition, westerly anomalies to the south of the AEJ in West Africa induce a cyclonic vorticity anomaly at the jet level, which would enhance AEW development through barotropic conversion and precede maximized EKE anomalies in phase 2 (Fig. 3). The maximum contraction of the AEJ occurs during the dry MJO phase and convective minimum over tropical North Africa (Fig. 10b). In this case, the −8 m s−1 contour is the largest observed to the east of Lake Chad. A 2 m s−1 difference in the easterly winds exists between the wet and dry MJO phases in the vicinity of the trigger region. Although the remaining phase composites are not included, it follows that the jet expands eastward between phases 5 and 1, while contracting between phases 1 and 5, as shown in Fig. 10c. In fact, we argue that the extension of the jet before phase 1 into the trigger region supports stronger convective disturbances there and allows for more precursor disturbances to enter the jet and contribute the convective maximum observed in phase 1. The aforementioned results are consistent with the intraseasonal variability of the AEJ examined in Leroux et al. (2010), specifically the extension of the AEJ prior to increased AEW activity. We also note that opposite-signed zonal wind anomalies flank the jet to the west of the Greenwich Meridian, resembling the first EOF of zonal wind at 650 hPa over West Africa in Leroux et al. (2010).

The pattern of pressure velocity in the trigger region suggests influence from an expanding and contracting AEJ. Figure 11 displays 30–90-day ERA-I 400-hPa omega anomalies in phases 4 and 8, or just before the wet and dry MJO phases, respectively. Phase 8 shows that significant anomalous ascent in the trigger region at 400 hPa exceeding 0.05 Pa s−1 coincides with the AEJ extension prior to the wet MJO phase and would support increased upstream convection in the trigger region seeding the AEJ with stronger precursor disturbances (THK08; Leroux and Hall 2009). During AEJ contraction, the trigger region is characterized by anomalous descent at 400 hPa, as shown in phase 4 of Fig. 11. The zonal gradient of zonal wind is enhanced in the trigger region at this time. While not shown here, Q-vector analysis (e.g., Kiladis et al. 2006) suggests that these vertical velocity anomalies may be at least partially driven by dry dynamical variations of the jet, although we cannot rule out that convection itself helps to drive these vertical velocity anomalies.

Fig. 11.
Fig. 11.

MJO phase composites of 30–90-day ERA-I 400-hPa omega anomalies for June–September. Only phases 4 and 8 are shown. The shading interval is 1 × 10−3 Pa s−1. Dots represent data that are significant with 90% confidence. The trigger region is marked by a box.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

c. Moisture budget

The vertically integrated moisture budget is examined to assess processes responsible for moistening (drying) in the trigger region in advance of the wet (dry) MJO phases. Figure 4 showed the appearance of a localized positive moisture anomaly in the trigger region in advance of the wet MJO phase that would support convection. The vertically integrated moisture budget equation was derived by Maloney et al. (2010) and is given by
e4
where q is the specific humidity, E represents the surface evaporation, P represents precipitation, and R is the residual. Equation (4) was computed using ERA-I fields and vertical integration was conducted from the surface to 200 hPa.

In Fig. 12, reveals a significant moistening signal in the trigger region from phase 5 until phase 7, which suggests that the atmosphere in the trigger region becomes more favorable for vigorous convection prior to the wet MJO phase (e.g., Bretherton et al. 2004). Intraseasonal moisture anomalies maximize when transitions from positive to negative. Thus, Fig. 12 suggests that moisture maximizes within the trigger region in phase 8, consistent with the moisture fields above (Fig. 4). The signal is first observed in phase 5 at 20°N, 30°E. By phase 7, the region of significant has spread southwest toward Lake Chad. Similarly, drying from phase 1 to phase 3 would tend to create negative anomalies that suppress convective activity upstream of the AEJ. Interestingly, during the moistening and drying periods in the trigger region, moisture tendency anomalies elsewhere in West Africa are weak, suggesting that the precursor signal does indeed occur near the region suggested by THK08 for seeding AEW development.

Fig. 12.
Fig. 12.

MJO phase composites for vertically integrated 30–90-day ERA-I moisture tendency anomalies. The shading interval is 4 × 10−7 kg m−2 s−1. Dots represent data that are significant with 90% confidence. The trigger region is marked by a box.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

The vertically integrated anomalous moisture budget terms for June–September are plotted as a function of phase within the trigger region (Fig. 13a). Vertically integrated meridional moisture advection is the dominant term on intraseasonal time scales. The quantity suggests moistening from phases 5 to 7, or before the wet MJO phase, and drying from phases 1 to 3, or before the dry MJO phase. The quantity is more than sufficient to explain the amplitude of the tendency anomalies. Figure 13a suggests that a considerable residual exists, which is not unexpected given that reanalysis products often contain a substantial analysis increment that must be added that constrains the model to be close to observations, and we calculated advective terms based on model data interpolated to pressure levels, and not the native model grid.

Fig. 13.
Fig. 13.

(a) Phase vs pressure plot of vertically integrated 30–90-day ERA-I moisture budget term anomalies for June–September in the trigger region. Terms include moisture tendency (black solid), zonal moisture advection (green dashed), meridional moisture advection (red dotted), moisture convergence minus precipitation (blue double dot–dashed), and evaporation (gray dot–dashed). (b) Phase vs pressure plot of vertically integrated 30–90-day meridional moisture advection perturbation budget term anomalies for June–September in the trigger region. Terms include total meridional moisture advection (black solid), anomalous meridional wind advecting mean moisture (red dashed), mean meridional wind advecting anomalous moisture (blue dotted), and (green dot–dashed). Units are kg m−2 s−1.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

Meridional moisture advection is analyzed further by computing a linearized budget (Fig. 13b), which is given by
e5
where and . Here, overbars represent the 50-day running mean, and primes represent the perturbation from the 50-day running mean. The terms and appear to dominate the meridional advection anomalies. The low-level flow (Fig. 2) suggests anomalous southerly flow in the trigger region centered near 20°N prior to the wet MJO phase (e.g., phase 7) and anomalous northerly flow prior to the dry MJO phase (e.g., phase 3), which supports the importance of . The mean moisture gradient is directed toward the south in this region. Furthermore, appears to be associated with decay of existing moisture anomalies in the trigger region through advection by the mean meridional wind. The meridional wind perturbations are filtered by eastward and westward zonal wavenumbers to determine their contributions to the moisture tendency anomalies observed within the trigger region (Fig. 12). The meridional wind perturbation is defined as , where accounts for zonal wavenumbers 0–120, and accounts for zonal wavenumbers −1 to −120. The quantity is then partitioned using these eastward and westward wind components because of its importance in the moisture budget of the trigger region.

Figure 14, which shows the partitioning in advance of the dry MJO phase (the period in advance of the wet MJO phase is similar but opposite), indicates that both components of the meridional flow are important for producing the observed intraseasonal anomalies, although the contribution from westward-propagating zonal wavenumbers is larger. As the trigger region begins drying in phase 1, which is suggested by in Fig. 12, meridional moisture advection is mostly responsible. Meridional wind anomalies may be contributed by Rossby waves propagating through tropical North Africa from the Indian Ocean (Matthews 2004; Janicot et al. 2009).

Fig. 14.
Fig. 14.

MJO phase composites for vertically integrated ERA-I for June–September. (a) Phases 1–3 for . (b) Phases 1–3 for . The shading interval is 1 × 10−6 kg m−2 s−1.

Citation: Journal of Climate 25, 9; 10.1175/JCLI-D-11-00232.1

6. Conclusions

This study investigated how the Madden–Julian oscillation (MJO) produces variability in convective and easterly wave activity in the West African monsoon during boreal summer using CLAUS TB as a proxy for convection and ERA-I reanalysis fields. Based on the analysis presented here, we hypothesize that MJO-related convection and AEW variability in the West African monsoon region are first initiated by modulation of precursor convective disturbances in the Darfur Mountains and Ethiopian Highlands, which has been shown to be a sensitive region for producing variability in African easterly waves downstream in the African easterly jet (THK08; Leroux and Hall 2009).

A multivariate MJO index was created from upper- and lower-tropospheric ERA-I zonal wind and CLAUS TB, and composites were generated during June–September of 1989–2005. MJO-induced variability in tropical North Africa is diagnosed through an analysis of convection, moisture, apparent heat source, EKE and wind anomalies. Periods when precipitation and easterly wave activity are enhanced (wet MJO phase) or reduced (dry MJO phase) across West Africa and the Atlantic were determined based on composite analysis. Kelvin and Rossby wave signatures communicate MJO-related anomalies from the Indian Ocean to tropical North Africa. Rossby waves tend to arrive 5–10 days prior to Kelvin waves, although both wave types modulate tropical North Africa, consistent with previous studies.

Prior to the wet and dry MJO phases, a region near the AEJ entrance that we refer to as the “trigger region” exhibits significant intraseasonal variability, especially in TB, moisture, and low-level EKE fields. Convection, moisture, and wave activity anomalies appear in the trigger region about 10 days before maximizing over the remainder of tropical West Africa. The fact that initial anomalies occur in the trigger region during MJO events supports the hypothesis that increased upstream convection and wave activity prior to the wet MJO phase may seed the AEJ prior to AEW development and enhance convection across West Africa and the Atlantic at later lags (THK08; Leroux and Hall 2009). Specifically, these MJO-specific results support the previous findings of Leroux et al. (2010), who found that stronger convective activity in the trigger region tends to precede intraseasonal periods of enhanced AEW activity. Similarly, reduced convection in the trigger region in advance of the dry MJO phase may suppress downstream AEW growth at later lags. The MJO accounts for approximately 50% of the amplitude of intraseasonal convective anomalies within the trigger region, as diagnosed by comparing the MJO index with a trigger region TB index. In West Africa, the MJO index produced stronger composite EKE anomalies than the TB index, which suggests that variability in the trigger region is not the only factor for downstream wave growth.

MJO teleconnections from the Indo-Pacific warm pool appear to affect precursor activity in the trigger region through three mechanisms. One mechanism is the reduction of 400-hPa temperatures prior to the wet MJO phase, which reduces upstream static stability and supports more vigorous convection. Another mechanism is the eastward AEJ extension into the trigger region in advance of the wet MJO phase, and a shortening of the AEJ in advance of the dry MJO phase. Jet variations affect the ability of eddies to grow through strong barotropic and baroclinic conversions in the trigger region. Furthermore, anomalous ascent (descent) is found above the jet level in the trigger region due to expansion (contraction) of the jet in advance of the wet (dry) MJO phase.

A final mechanism of importance is meridional moisture advection, which appears to dominate growth–decay of moisture anomalies within the trigger region, and cause growth of positive (negative) moisture anomalies in advance of the wet (dry) MJO phase. Southerly wind anomalies advect moisture down a strong climatological precipitable water gradient in the trigger region prior to the wet MJO phase, while the opposite occurs before the dry MJO phase. Although eastward and westward zonal wavenumber flow components both contribute to meridional advection, westward-propagating flow components are slightly larger, which suggests the importance of Rossby waves.

Future work will continue focus on the relationship between the MJO and AEWs. Primitive equation modeling studies will be performed to determine whether jet variations or the strength of precursor disturbances are more important for producing EKE variations associated with the MJO in the West African monsoon system. In particular, the relationship between MJO phase and AEW characteristics and amplitude will be investigated by employing different basic states and precursor heating strengths, based on our composites shown here. More in-depth analysis of the contributions of eastward and westward components of the flow is necessary to better understand the independent influence of each on tropical Africa and the influence of the MJO on this region. General circulation modeling studies with selective filtering of westward- and eastward-propagating disturbances that impinge on West Africa may be productive avenues for further exploration. Additionally, in-depth analysis of the EKE and EAPE budgets is necessary to analyze independent contributions from diabatic heating and baroclinic and barotropic conversion to wave growth in the trigger region and downstream in the jet during MJO events.

Acknowledgments

Three reviewers provided comments that substantially improved the manuscript. The authors thank Wayne Schubert and Karan Venayagamoorthy for valuable criticism on an earlier version of the manuscript. The authors would also like to credit the European Centre for Medium-Range Weather Forecasts for the use of ERA-Interim datasets and the British Atmospheric Data Centre for the use of CLAUS data. This work was supported by the Climate and Large-Scale Dynamics Program of the National Science Foundation under Grants ATM-0828531 and AGS-0946911. This work has also been funded by Award NA08OAR4320893 from the National Oceanic and Atmospheric Administration, U.S. Department of Commerce. The statements, findings, conclusions, and recommendations do not necessarily reflect the views of NSF, NOAA, or the Department of Commerce.

REFERENCES

  • Albignat, J. P., and R. J. Reed, 1980: The origin of African wave disturbances during Phase III of GATE. Mon. Wea. Rev., 108, 18271839.

    • Search Google Scholar
    • Export Citation
  • Annamalai, H., and J. M. Slingo, 2001: Active/break cycles: Diagnosis of the intraseasonal variability of the Asian Summer Monsoon. Climate Dyn., 18, 85102.

    • Search Google Scholar
    • Export Citation
  • Bantzer, C. H., and J. M. Wallace, 1996: Intraseasonal variability in tropical mean temperature and precipitation and their relation to the tropical 40–50 day oscillation. J. Atmos. Sci., 53, 30323045.

    • Search Google Scholar
    • Export Citation
  • Berry, G., and C. D. Thorncroft, 2005: Case study of an intense African easterly wave. Mon. Wea. Rev., 133, 752766.

  • Bretherton, C. S., M. E. Peters, and L. E. Back, 2004: Relationships between water vapor path and precipitation over the tropical oceans. J. Climate, 17, 15171528.

    • Search Google Scholar
    • Export Citation
  • Burpee, R. W., 1972: The origin and structure of easterly waves in the lower troposphere of North Africa. J. Atmos. Sci., 29, 7790.

  • Burpee, R. W., 1974: Characteristics of North African easterly waves during the summers of 1968 and 1969. J. Atmos. Sci., 31, 15561570.

    • Search Google Scholar
    • Export Citation
  • Camargo, S. J., M. C. Wheeler, and A. H. Sobel, 2009: Diagnosis of the MJO modulation of tropical cyclogenesis using an empirical index. J. Atmos. Sci., 66, 30613074.

    • Search Google Scholar
    • Export Citation
  • Carlson, T. N., 1969a: Some remarks on African disturbances and their progress over the tropical Atlantic. Mon. Wea. Rev., 97, 716726.

    • Search Google Scholar
    • Export Citation
  • Carlson, T. N., 1969b: Synoptic histories of three African disturbances that developed into Atlantic hurricanes. Mon. Wea. Rev., 97, 256276.

    • Search Google Scholar
    • Export Citation
  • Chauvin, F., R. Roehrig, and J.-P. Lafore, 2010: Intraseasonal variability of the Saharan heat low and its link with midlatitudes. J. Climate, 23, 25442561.

    • Search Google Scholar
    • Export Citation
  • de Coëtlogon, G., S. Janicot, and A. Lazar, 2010: Intraseasonal variability of the ocean–atmosphere coupling in the Gulf of Guinea during boreal spring and summer. Quart. J. Roy. Meteor. Soc., 136, 426441.

    • Search Google Scholar
    • Export Citation
  • Gill, A. E., 1980: Some simple solutions for heat-induced tropical circulation. Quart. J. Roy. Meteor. Soc., 106, 447462.

  • Hall, N. M. J., G. N. Kiladis, and C. D. Thorncroft, 2006: Three-dimensional structure of African easterly waves. Part II: Dynamical modes. J. Atmos. Sci., 63, 22312245.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., and M. L. Salby, 1994: The life cycle of the Madden–Julian Oscillation. J. Atmos. Sci., 51, 22252237.

  • Hodges, K. I., D. W. Chappell, G. J. Robinson, and G. Yang, 2000: An improved algorithm for generating global window brightness temperatures from multiple satellite infrared imagery. J. Atmos. Oceanic Technol., 17, 12961312.

    • Search Google Scholar
    • Export Citation
  • Hsieh, J.-S., and K. H. Cook, 2005: Generation of African easterly wave disturbances: Relationship to the African easterly jet. Mon. Wea. Rev., 133, 13111327.

    • Search Google Scholar
    • Export Citation
  • Janicot, S., F. Mounier, N. M. J. Hall, S. Leroux, B. Sultan, and G. N. Kiladis, 2009: Dynamics of the West African monsoon. Part IV: Analysis of 25–90-day variability of convection and the role of the Indian monsoon. J. Climate, 22, 15411565.

    • Search Google Scholar
    • Export Citation
  • Janicot, S., and Coauthors, 2011: Intraseasonal variability of the West African monsoon. Atmos. Sci. Lett., 12, 5866.

  • Kiladis, G. N., C. D. Thorncroft, and N. M. J. Hall, 2006: Three-dimensional structure and dynamics of African easterly waves. Part I: Observations. J. Atmos. Sci., 63, 22122230.

    • Search Google Scholar
    • Export Citation
  • Knutson, D. R., and K. M. Weickmann, 1987: 30–60 day atmospheric oscillations: Composite life cycles of convection and circulation anomalies. Mon. Wea. Rev., 115, 14071436.

    • Search Google Scholar
    • Export Citation
  • Knutson, D. 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, 605623.

    • Search Google Scholar
    • Export Citation
  • Lebel, T., and Coauthors, 2010: The AMMA field campaigns: Multiscale and multidisciplinary observations in the West African region. Quart. J. Roy. Meteor. Soc., 136, 833.

    • Search Google Scholar
    • Export Citation
  • Leroux, S., and N. M. J. Hall, 2009: On the relationship between African Easterly waves and the African easterly jet. J. Atmos. Sci., 66, 23032316.

    • Search Google Scholar
    • Export Citation
  • Leroux, S., N. M. J. Hall, and G. N. Kiladis, 2010: A climatological study of transient–mean–flow interactions over West Africa. Quart. J. Roy. Meteor. Soc., 136, 397410.

    • Search Google Scholar
    • Export Citation
  • Leroux, S., N. M. J. Hall, and G. N. Kiladis, 2011: Intermittent African easterly wave activity in a dry atmospheric model: Influence of the extratropics. J. Climate, 24, 53785396.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., and D. L. Hartmann, 1998: Frictional moisture convergence in a composite life cycle of the Madden–Julian Oscillation. J. Climate, 11, 23872403.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., and D. L. Hartmann, 2000a: Modulation of eastern North Pacific hurricanes by the Madden–Julian Oscillation. J. Climate, 13, 14511460.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., and D. L. Hartmann, 2000b: Modulation of hurricane activity in the Gulf of Mexico by the Madden–Julian oscillation. Science, 287, 20022004.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., and A. H. Sobel, 2007: Idealized hot spot experiments with a general circulation model. J. Climate, 20, 908925.

  • Maloney, E. D., and J. Shaman, 2008: Intraseasonal variability of the West African monsoon and the Atlantic ITCZ. J. Climate, 21, 28982918.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., A. H. Sobel, and W. M. Hannah, 2010: Intraseasonal variability in an aquaplanet general circulation model. J. Adv. Model Earth Syst., 2, Article 5, 24 pp., doi:10.3894/JAMES.2010.2.5.

    • Search Google Scholar
    • Export Citation
  • Matthews, A. J., 2004: Intraseasonal variability over tropical Africa during northern summer. J. Climate, 17, 24272440.

  • Mekonnen, A., and W. B. Rossow, 2011: The interaction between deep convection and easterly waves over tropical North Africa: A weather state perspective. J. Climate, 24, 42754293.

    • Search Google Scholar
    • Export Citation
  • Mekonnen, A., C. D. Thorncroft, and A. R. Aiyyer, 2006: Analysis of convection and its association with African easterly waves. J. Climate, 19, 54055421.

    • Search Google Scholar
    • Export Citation
  • Mekonnen, A., C. D. Thorncroft, A. R. Aiyyer, and G. N. Kiladis, 2008: Convectively coupled Kelvin waves over tropical Africa during the boreal summer: Structure and variability. J. Climate, 21, 66496667.

    • Search Google Scholar
    • Export Citation
  • Mo, K. C., 2000: The association between intraseasonal oscillations and tropical storms in the Atlantic basin. Mon. Wea. Rev., 128, 40974107.

    • Search Google Scholar
    • Export Citation
  • Pohl, B., S. Janicot, B. Fontaine, and R. Marteau, 2009: Implication of the Madden–Julian oscillation in the 40-day variability of the West African monsoon. J. Climate, 22, 37693785.

    • Search Google Scholar
    • Export Citation
  • Raymond, D. J., and H. Jiang, 1990: A theory for long-lived mesoscale convective systems. J. Atmos. Sci., 47, 30673077.

  • Robinson, G. J., and K. I. Hodges, 2005: Cloud archive user service user guide. NERC Environmental Systems Science Centre, 18 pp. [Available online at http://badc.nerc.ac.uk/data/claus/userguide.pdf.]

  • Simmons, A. J., R. Mureau, and T. Petroliagis, 1995: Error growth and estimates of predictability from the ECMWF forecasting system. Quart. J. Roy. Meteor. Soc., 121, 17391771.

    • Search Google Scholar
    • Export Citation
  • Spiegel, M. R., 1992: Theory and Problems of Statistics. McGraw-Hill, 504 pp

  • Straub, K. H., and G. N. Kiladis, 2003: Extratropical forcing of convectively coupled Kelvin waves during austral winter. J. Atmos. Sci., 60, 526543.

    • Search Google Scholar
    • Export Citation
  • Sultan, B., and S. Janicot, 2003: West African monsoon dynamics. Part II: The “preonset” and the “onset” of the summer monsoon. J. Climate, 16, 34073427.

    • Search Google Scholar
    • Export Citation
  • Thorncroft, C. D., and B. J. Hoskins, 1994: An idealised study of African easterly waves. 1. A linear view. Quart. J. Roy. Meteor. Soc., 120, 953982.

    • Search Google Scholar
    • Export Citation
  • Thorncroft, C. D., N. M. J. Hall, and G. N. Kiladis, 2008: Three dimensional structure and dynamics of African easterly waves. Part III: Genesis. J. Atmos. Sci., 65, 35963607.

    • Search Google Scholar
    • Export Citation
  • Tromeur, E., and W. B. Rossow, 2010: Interaction of tropical deep convection with the large-scale circulation in the MJO. J. Climate, 23, 18371853.

    • Search Google Scholar
    • Export Citation
  • Vitart, F., and F. Molteni, 2010: Simulation of the Madden– Julian Oscillation and its teleconnections in the ECMWF forecast system. Quart. J. Roy. Meteor. Soc., 136, 842855.

    • Search Google Scholar
    • Export Citation
  • Wheeler, M. C., and K. M. Weickmann, 2001: Real-time monitoring and prediction of modes of coherent synoptic to intraseasonal tropical variability. Mon. Wea. Rev., 129, 26772694.

    • Search Google Scholar
    • Export Citation
  • 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, 19171932.

    • Search Google Scholar
    • Export Citation
  • Yanai, M., S. K. Esbensen, and J. H. Chu, 1973: Determination of bulk properties of tropical cloud clusters from large-scale heat and moisture budgets. J. Atmos. Sci., 30, 611627.

    • Search Google Scholar
    • Export Citation
Save