• Ambenje, P. G., 1990: Mean tropospheric motion field over Kenya during the 1984 long rains season (March–May). Extended Abstracts, Third WMO Symp. on Meteorological Aspects of Tropical Droughts with Emphasis on Long-Range Forecasting, Niamey, Niger, WMO, 83.

    • Search Google Scholar
    • Export Citation
  • Balas, N., , S. E. Nicholson, , and D. Klotter, 2007: The relationship of rainfall variability in west central Africa to sea-surface temperature fluctuations. Int. J. Climatol., 27, 13351349.

    • Search Google Scholar
    • Export Citation
  • Behera, S. K., and Coauthors, 2005: Paramount impact of the Indian Ocean dipole on East African short rains: A CGCM study. J. Climate, 18, 45144530.

    • Search Google Scholar
    • Export Citation
  • Beltrando, G., , and D. L. Cadet, 1990: Interannual variability of the short rain season in East Africa: Relationships with general atmospheric circulation. Veille Climatique Satell., 33, 1936.

    • Search Google Scholar
    • Export Citation
  • Black, E., , J. Slingo, , and K. R. Sperber, 2003: An observational study of the relationship between excessively strong short rains in coastal East Africa and Indian Ocean SST. Mon. Wea. Rev., 131, 7494.

    • Search Google Scholar
    • Export Citation
  • Camberlin, P., 1995: June–September rainfall in north-eastern Africa and atmospheric signals over the tropics: A zonal perspective. Int. J. Climatol., 15, 773783.

    • Search Google Scholar
    • Export Citation
  • Camberlin, P., , and J. Wairoto, 1997: Intraseasonal wind anomalies related to wet and dry spells during the “long” and “short” rainy seasons in Kenya. Theor. Appl. Climatol., 58, 5769.

    • Search Google Scholar
    • Export Citation
  • Camberlin, P., , and N. Philippon, 2002: The East African March–May rainy season: Associated atmospheric dynamics and predictability over the 1968–97 period. J. Climate, 15, 10021019.

    • Search Google Scholar
    • Export Citation
  • Carton, J. A., , and B. Huang, 1994: Warm events in the tropical Atlantic. J. Phys. Oceanogr., 24, 888903.

  • Christian, H. J., and Coauthors, 2003: Global frequency and distribution of lightning as observed from space by the Optical Transient Detector. J. Geophys. Res.,108, 4005, doi:10.1029/2002JD002347.

  • Clark, C. O., , P. J. Webster, , and J. E. Cole, 2003: Interdecadal variability of the relationship between the Indian Ocean zonal mode and East African coastal rainfall anomalies. J. Climate, 16, 548554.

    • Search Google Scholar
    • Export Citation
  • Dezfuli, A. K., 2011a: Spatio-temporal variability of seasonal rainfall in western equatorial Africa. Theor. Appl. Climatol., 104, 5769.

    • Search Google Scholar
    • Export Citation
  • Dezfuli, A. K., 2011b: The relationship of interannual variability in western equatorial Africa to the tropical oceans and atmospheric circulation. Ph.D. dissertation, The Florida State University.

  • Dezfuli, A. K., , and S. E. Nicholson, 2011: A note on the long-term variations of the African easterly jet. Int. J. Climatol., 31, 20492054.

    • Search Google Scholar
    • Export Citation
  • Dezfuli, A. K., , and S. E. Nicholson, 2013: The relationship of rainfall variability in western equatorial Africa to the tropical oceans and atmospheric circulation. Part II: The boreal autumn. J. Climate,26, 66–84.

  • Gong, X., , and M. B. Richman, 1995: On the application of cluster analysis to growing season precipitation data in North America east of the Rockies. J. Climate, 8, 897931.

    • Search Google Scholar
    • Export Citation
  • Grist, J. P., , and S. E. Nicholson, 2001: A study of the dynamic factors influencing the interannual variability of rainfall in the West African Sahel. J. Climate, 14, 13371359.

    • Search Google Scholar
    • Export Citation
  • Grodsky, S. A., , and J. A. Carton, 2003: The intertropical convergence zone in the South Atlantic and the equatorial cold tongue. J. Climate, 16, 723733.

    • Search Google Scholar
    • Export Citation
  • Hastenrath, S., 2001: In search of zonal circulations in the equatorial Atlantic sector from the NCEP–NCAR reanalysis. Int. J. Climatol., 21, 3747.

    • Search Google Scholar
    • Export Citation
  • Hastenrath, S., , D. Polzin, , and C. Mutai, 2011: Circulation mechanisms of Kenya rainfall anomalies. J. Climate, 24, 404412.

  • Hirst, A. C., , and S. Hastenrath, 1983a: Atmosphere–ocean mechanisms of climate anomalies in the Angola–tropical Atlantic sector. J. Phys. Oceanogr., 13, 11461157.

    • Search Google Scholar
    • Export Citation
  • Hirst, A. C., , and S. Hastenrath, 1983b: Diagnostics of hydrometeorological anomalies in the Zaire (Congo) basin. Quart. J. Roy. Meteor. Soc., 109, 881892.

    • Search Google Scholar
    • Export Citation
  • Indeje, M., , F. H. M. Semazzi, , and L. J. Ogallo, 2000: ENSO signals in East African rainfall seasons. Int. J. Climatol.,20, 19–46.

  • Jackson, B., , S. E. Nicholson, , and D. Klotter, 2009: Mesoscale convective systems over western equatorial Africa and their relationship to large-scale circulation. Mon. Wea. Rev., 137, 12721294.

    • Search Google Scholar
    • Export Citation
  • Joyce, T. M., , C. Frankignoul, , J. Yang, , and H. E. Phillips, 2004: Ocean response and feedback to the SST dipole in the tropical Atlantic. J. Phys. Oceanogr., 34, 25252540.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471.

  • Kijazi, A. L., , and C. J. C. Reason, 2005: Relationships between intraseasonal rainfall variability of coastal Tanzania and ENSO. Theor. Appl. Climatol., 82, 153176.

    • Search Google Scholar
    • Export Citation
  • Kistler, R., and Coauthors, 2001: The NCEP–NCAR 50-Year Reanalysis: Monthly means CD-ROM and documentation. Bull. Amer. Meteor. Soc., 82, 247267.

    • Search Google Scholar
    • Export Citation
  • Laing, A., , and J. M. Fritsch, 1993: Mesoscale convective complexes in Africa. Mon. Wea. Rev., 121, 22542263.

  • Laing, A., , R. Carbone, , V. Levizzani, , and J. Tuttle, 2008: The propagation and diurnal cycles of deep convection in northern tropical Africa. Quart. J. Roy. Meteor. Soc., 134, 93109.

    • Search Google Scholar
    • Export Citation
  • Lamb, P. J., , and R. Peppler, 1992: Further case studies of tropical Atlantic surface atmospheric and oceanic patterns associated with sub-Saharan drought. J. Climate, 5, 476488.

    • Search Google Scholar
    • Export Citation
  • Lienou, G., and Coauthors, 2008: Evolution of hydrological regimes in the equatorial area of Cameroon: An impact of climate variability in equatorial Africa? Hydrol. Sci. J., 53, 789801.

    • Search Google Scholar
    • Export Citation
  • Macodras, M. W., , P. N. Nthusi, , and J. Mwikya, 1989: Synoptic features associated with the failure of 1984 long rains in Kenya. Coping with Drought in Kenya: National and Local Strategies, T. E. Downing, K. W. Gitu, and C. M. Kamau, Eds., Lynne Rienner, 69–81.

  • McCollum, J. R., , A. Gruber, , and M. B. Ba, 2000: Discrepancy between gauges and satellite estimates of rainfall in equatorial Africa. J. Appl. Meteor., 39, 666679.

    • Search Google Scholar
    • Export Citation
  • Mohr, K. L., , and E. J. Zipser, 1996: Mesoscale convective systems defined by their 85-GHz ice scattering signature: Size and intensity comparison over tropical oceans and continents. Mon. Wea. Rev., 124, 24172437.

    • Search Google Scholar
    • Export Citation
  • Mutai, C. C., , and M. N. Ward, 2000: East African rainfall and the tropical circulation/convection on intraseasonal to interannual timescales. J. Climate, 13, 39153939.

    • Search Google Scholar
    • Export Citation
  • Nguyen, H., , and J. P. Duvel, 2008: Synoptic wave perturbations and convective systems over equatorial Africa. J. Climate, 21, 63726388.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., 1986: The spatial coherence of African rainfall anomalies: Interhemispheric teleconnections. J. Climate Appl. Meteor., 25, 13651381.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., 1996: A review of climate dynamics and climate variability in Eastern Africa. The Limnology, Climatology, and Paleoclimatology of the East African Lakes, T.C. Johnson and E. O. Odada, Eds., Gordon and Breach Publications, 25–56.

  • Nicholson, S. E., 2009a: On the factors modulating the intensity of the tropical rainbelt over West Africa. Int. J. Climatol., 29, 673689.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., 2009b: A revised picture of the structure of the “monsoon” and land ITCZ over West Africa. Climate Dyn., 32, 11551171.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., , and D. Entekhabi, 1987: Rainfall variability in equatorial and southern Africa: Relationships with sea surface temperatures along the southwestern coast of Africa. J. Climate Appl. Meteor., 26, 561578.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., , and B. S. Nyenzi, 1990: Temporal and spatial variability of SSTs in the tropical Atlantic and Indian Oceans. Meteor. Atmos. Phys., 42A, 117.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., , and I. M. Palao, 1993: A re-evaluation of rainfall variability in the Sahel. Part I. Characteristics of rainfall fluctuations. Int. J. Climatol., 13, 371389.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., , and J. Y. Kim, 1997: The relationship of the El Niño–Southern Oscillation to African rainfall. Int. J. Climatol., 17, 117135.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., , and J. P. Grist, 2003: The seasonal evolution of the atmospheric circulation over West Africa and equatorial Africa. J. Climate, 16, 10131030.

    • Search Google Scholar
    • Export Citation
  • Okoola, R. E., 1998: Spatial evolutions of the active convective patterns across the equatorial eastern Africa region during Northern Hemisphere spring season using outgoing longwave radiation records. Meteor. Atmos. Phys., 66, 5163.

    • Search Google Scholar
    • Export Citation
  • Okoola, R. E., 1999a: A diagnostic study of the eastern Africa monsoon circulation during the Northern Hemisphere spring season. Int. J. Climatol., 19, 143168.

    • Search Google Scholar
    • Export Citation
  • Okoola, R. E., 1999b: Midtropospheric circulation patterns associated with extreme dry and wet episodes over equatorial eastern Africa during the northern hemisphere spring. J. Appl. Meteor., 38, 11611169.

    • Search Google Scholar
    • Export Citation
  • Okumura, Y., , and S.-P. Xie, 2004: Interaction of the Atlantic equatorial cold tongue and the African monsoon. J. Climate, 17, 35893602.

    • Search Google Scholar
    • Export Citation
  • Pohl, B., , and P. Camberlin, 2006: Influence of the Madden–Julian oscillation on East Africa rainfall. II: March–May season extremes and interannual variability. Quart. J. Roy. Meteor. Soc., 132, 25412558.

    • Search Google Scholar
    • Export Citation
  • Rao, A. R., , and V. V. Srinivas, 2006: Regionalization of watersheds by hybrid-cluster analysis. J. Hydrol., 318, 3756.

  • Samba, G., , D. Nganga, , and M. Mpounza, 2007: Rainfall and temperature variations over Congo-Brazzaville between 1950 and 1998. Theor. Appl. Climatol., 91, 8597.

    • Search Google Scholar
    • Export Citation
  • Segele, Z. T., , P. J. Lamb, , and L. A. Leslie, 2009: Large-scale atmospheric circulation and global sea surface temperature associations with Horn of Africa June–September rainfall. Int. J. Climatol., 29, 10751100.

    • Search Google Scholar
    • Export Citation
  • Smith, T. M., , R. W. Reynolds, , T. C. Peterson, , and J. Lawrimore, 2008: Improvements to NOAA’s historical merged land–ocean surface temperature analysis (1880– 2006). J. Climate, 21, 22832296.

    • Search Google Scholar
    • Export Citation
  • Stickler, A., , and S. Brönnimann, 2011: Significant bias of the NCEP/NCAR and twentieth-century reanalyses relative to pilot balloon observations over the West African monsoon region (1940–1957). Quart. J. Roy. Meteor. Soc., 137, 14001416.

    • Search Google Scholar
    • Export Citation
  • Terray, P., , P. Delecluse, , S. Labattu, , and L. Terray, 2003: Sea surface temperature associations with the late Indian summer monsoon. Climate Dyn., 21, 593618.

    • Search Google Scholar
    • Export Citation
  • Todd, M. C., , and R. Washington, 2004: Climate variability in central equatorial Africa: Influence from the Atlantic sector. Geophys. Res. Lett., 31, L23202, doi:10.1029/2004GL020975.

    • Search Google Scholar
    • Export Citation
  • Ummenhofer, C. C., , A. Sen Gupta, , M. E. England, , and C. J. C. Reason, 2009: Contribution of Indian Ocean sea surface temperatures to enhanced East African rainfall. J. Climate, 22, 9931013.

    • Search Google Scholar
    • Export Citation
  • Unal, Y., , T. Kindap, , and M. Karaca, 2003: Redefining the climate zones of Turkey using cluster analysis. Int. J. Climatol., 23, 10451055.

    • Search Google Scholar
    • Export Citation
  • Vigaud, N., , Y. Richard, , M. Rouault, , and N. Fauchereau, 2007: Water vapour transport from the tropical Atlantic and summer rainfall in tropical southern Africa. Climate Dyn., 28, 113123.

    • Search Google Scholar
    • Export Citation
  • Xie, S.-P., , and J. A. Carton, 2004: Tropical Atlantic variability: Patterns, mechanisms, and impacts. Earth’s Climate: The Ocean–Atmosphere Interaction, Geophys. Monogr., Vol. 147, Amer. Geophys. Union, 121–142.

  • Yin, X. G., , and A. Gruber, 2010: Validation of the abrupt change in GPCP precipitation in the Congo River basin. Int. J. Climatol., 30, 110119.

    • Search Google Scholar
    • Export Citation
  • Zipser, E. J., , D. J. Cecil, , C. T. Liu, , S. W. Nesbitt, , and D. P. Yorty, 2006: Where are the most intense thunderstorms on Earth? Bull. Amer. Meteor. Soc., 87, 10571071.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    March and April rainfall along the coast in warm water years and cold water years (from Nicholson and Entekhabi 1987).

  • View in gallery

    (left) Homogeneous rainfall regions for the April–June (AMJ) season, superimposed upon a map of topographic relief. Asterisks indicate stations utilized. The typical seasonal cycle for each region is also indicated. (right) Mean rainfall (mm) during the AMJ season for the period 1948 to 1988.

  • View in gallery

    Regionally averaged rainfall for the AMJ season, 1948 to 1988, for regions 1 to 5. The data are presented as standardized anomalies from the mean for the period (anomaly divided by the standard deviation).

  • View in gallery

    Mean (top) sea surface temperatures (°C) and (bottom) sea level pressure (SLP; hPa) during AMJ for 1948 to 1988.

  • View in gallery

    Mean wind vectors at four levels during AMJ for 1948 to 1988.

  • View in gallery

    Correlation between regional AMJ rainfall and concurrent sea surface temperatures for 1948–88. The 5% significance level is indicated by the dash-dotted line. Rainfall is also correlated with SST averages for the indicated boxes, with values given in Table 2.

  • View in gallery

    Standardized SST anomalies for the wet composite and the dry composite of regions 2, 3, and 5. Anomalies are based on the period 1948–88.

  • View in gallery

    Correlation between regional AMJ rainfall of region 2 and concurrent sea level pressure for the period 1948–88. The 5% significance level is indicated by the dash-dotted line. Rainfall is also correlated with SLP averages for the indicated box, with value given in Table 2.

  • View in gallery

    Latitude–height cross sections of specific humidity (g kg−1) for regions (top) 2, (middle) 3, and (bottom) 5 during AMJ. Specific humidity is zonally averaged over 10°–15°E, 10°–20°E, and 15°–30°E for regions 2, 3, and 5, respectively. The panel on the right shows the difference between the wet and dry composites; shading indicates areas where the difference is significant at 5% level.

  • View in gallery

    Latitude–height cross sections of omega (10−2 Pa s−1) for regions (top) 2, (middle) 3, and (bottom) 5 during AMJ. Omega is zonally averaged over 10°–15°E, 10°–20°E, and 15°–30°E, for regions 2, 3, and 5, respectively. Areas of rising motion are shaded. The panel on the right shows the difference between the wet and dry composites; shading indicates areas where the difference is significant at 5% level.

  • View in gallery

    Latitude–height cross sections of zonal winds (m s−1) for regions 2 (top), 3 (middle) and 5 (bottom) during AMJ. Zonal wind is averaged over 10°–15°E, 10°–20°E, and 15°–30°E, for regions 2, 3, and 5, respectively. Areas of easterly winds are shaded. The panel on the right shows the difference between the wet and dry composites; shading indicates areas where the difference is significant at 5% level.

  • View in gallery

    (top) Wet-minus-dry composites of divergence (10−6 s−1) at 200 hPa, and zonal wind (m s−1) at (middle) 200 hPa and (bottom) 850 hPa. The three-panel diagrams clockwise from top left represent regions 2, 3, and 5, respectively. The dashed line shows areas in which the difference between the wet and dry composites is significant at the 5% level.

  • View in gallery

    Correlation between AMJ rainfall and zonal wind at 200, 850, and 1000 hPa for regions 2 (top), 3 (middle), and 5 (bottom). Correlations are based on the period 1948–88 and only those exceeding the 5% significance level are shown. Rainfall is also correlated with zonal wind averages for the indicated boxes, with values given in Table 2.

  • View in gallery

    Omega (10−2 Pa s−1) for wet, dry, and wet-minus-dry composites. The three-panel diagrams clockwise from top left represent regions 2, 3, and 5, respectively. Omega of each region is meridionally averaged over its corresponding latitudes. The dashed lines in the bottom panel of each set of diagrams show areas in which the difference between the wet and dry composites is significant at the 5% level.

  • View in gallery

    Zonal wind (m s−1, top) and divergence (10−6 s−1, bottom) at 600 hPa for the wet, dry, and wet-minus-dry composites of region 5.

  • View in gallery

    Daily rainfall (mm) during AMJ for region 5 for the two wettest years (2002, 2006) and the two driest years (1998, 2004) of the period 1998–2010. The data are from TRMM 3B42.

  • View in gallery

    Omega (10−2 Pa s−1) for the wettest and driest 6-day periods in April of region 5 during the four years shown in Fig. 16. Wet composite is for 12–17 Apr 2006, and dry composite is for 25–30 Apr 1998.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 0 0 0
PDF Downloads 0 0 0

The Relationship of Rainfall Variability in Western Equatorial Africa to the Tropical Oceans and Atmospheric Circulation. Part I: The Boreal Spring

View More View Less
  • 1 Department of Earth, Ocean and Atmospheric Science, The Florida State University, Tallahassee, Florida
  • 2 Department of Earth and Planetary Sciences, Johns Hopkins University, Baltimore, Maryland
© Get Permissions
Full access

Abstract

This paper examines the factors governing rainfall variability in western equatorial Africa (WEA) during the April–June rainy season. In three of the five regions examined some degree of large-scale forcing is indicated, particularly in the region along the Atlantic coast. Interannual variability in this coastal sector also demonstrates a strong link to changes in local sea surface temperatures (SSTs) and the South Atlantic subtropical high.

To examine potential causal mechanisms, various atmospheric parameters are evaluated for wet and dry composites. The results suggest that the intensity of the zonal circulation in the global tropics is a crucial control on rainfall variability over WEA. A La Niña (El Niño)–like signal in both SSTs and zonal circulation over the Pacific is apparent in association with the wet (dry) conditions in the western sector. However, remote forcing from the Pacific modulates the circulation over Africa indirectly by way of synchronous changes in the entire Indian or Atlantic Ocean.

Anomalies in the local zonal winds are similar in all three regions: the wet (dry) composite is associated with an intensification (weakening) of the upper-tropospheric easterlies and low-level westerlies, but a weakening (intensification) of the midlevel easterlies. This work also suggests that, in most cases, the relationship between local SSTs and rainfall reflects a common remote forcing by the large-scale atmosphere–ocean system. This forcing is manifested via changes in the zonal circulation. Thus, the statistical associations between rainfall and SSTs do not indicate direct forcing by local SSTs. One point of evidence for this conclusion is the stronger association with atmospheric parameters than with SSTs.

Corresponding author address: Amin K. Dezfuli, Department of Earth and Planetary Sciences, Johns Hopkins University, 3400 N. Charles Street, 301 Olin Hall, Baltimore, MD 21218. E-mail: dez@jhu.edu

Abstract

This paper examines the factors governing rainfall variability in western equatorial Africa (WEA) during the April–June rainy season. In three of the five regions examined some degree of large-scale forcing is indicated, particularly in the region along the Atlantic coast. Interannual variability in this coastal sector also demonstrates a strong link to changes in local sea surface temperatures (SSTs) and the South Atlantic subtropical high.

To examine potential causal mechanisms, various atmospheric parameters are evaluated for wet and dry composites. The results suggest that the intensity of the zonal circulation in the global tropics is a crucial control on rainfall variability over WEA. A La Niña (El Niño)–like signal in both SSTs and zonal circulation over the Pacific is apparent in association with the wet (dry) conditions in the western sector. However, remote forcing from the Pacific modulates the circulation over Africa indirectly by way of synchronous changes in the entire Indian or Atlantic Ocean.

Anomalies in the local zonal winds are similar in all three regions: the wet (dry) composite is associated with an intensification (weakening) of the upper-tropospheric easterlies and low-level westerlies, but a weakening (intensification) of the midlevel easterlies. This work also suggests that, in most cases, the relationship between local SSTs and rainfall reflects a common remote forcing by the large-scale atmosphere–ocean system. This forcing is manifested via changes in the zonal circulation. Thus, the statistical associations between rainfall and SSTs do not indicate direct forcing by local SSTs. One point of evidence for this conclusion is the stronger association with atmospheric parameters than with SSTs.

Corresponding author address: Amin K. Dezfuli, Department of Earth and Planetary Sciences, Johns Hopkins University, 3400 N. Charles Street, 301 Olin Hall, Baltimore, MD 21218. E-mail: dez@jhu.edu

1. Introduction

Equatorial Africa comprises roughly half of the equatorial landmass, yet relatively little is known about the region’s meteorology. Most of the meteorological research on the region has focused on East Africa. The few studies of the more western sectors encompass rainfall climatology and interannual variability (Hirst and Hastenrath 1983a,b; McCollum et al. 2000; Todd and Washington 2004; Balas et al. 2007; Samba et al. 2007; Lienou et al. 2008; Yin and Gruber 2010; Dezfuli 2011), mesoscale convective systems (Laing and Fritsch 1993; Laing et al. 2008; Zipser et al. 2006; Jackson et al. 2009), atmospheric circulation (Nicholson and Grist 2003), and wave activity (Nguyen and Duvel 2008). The goal of the current study is to expand our knowledge of the region’s rainfall regime and, in particular, to identify the factors governing its interannual variability.

An outstanding characteristic of this region is the extreme spatial heterogeneity of interannual variability. This stands in sharp contrast to most other regions of Africa (Nicholson and Palao 1993; Nicholson 1996; Hastenrath et al. 2011; Pohl and Camberlin 2006), including the eastern equatorial region. It might be assumed that the heterogeneity is a result of the complex topography of the region. However, topography is equally complex in eastern equatorial parts of Africa, yet that region exhibits a coherent interannual signal throughout a sector that extends from roughly 5°N to 10°S and 28° to 43°E.

This heterogeneity was noted by Nicholson (1986), who speculated that it might reflect the quality of the gauge data, rather than true spatial heterogeneity. Balas et al. (2007) performed more extensive quality control and reevaluated the regionalization used by Nicholson (1986). That study confirmed the spatial heterogeneity and showed also that the factors governing interannual variability varied markedly within the region and from season to season. Dezfuli (2011a,b) extended that work by performing a regionalization at the seasonal scale. The current study is based on those results.

The one commonality within equatorial Africa is a bimodal rainy season, the typical “equatorial” regime with peaks occurring during the transition seasons and minima or dry seasons during the extreme seasons. The abovementioned studies and numerous works on East Africa have clearly demonstrated strong contrasts between the two rainy seasons in terms of their character and links to large-scale processes. Because of these contrasts, our work will be presented in two parts, with the two rainy seasons considered separately. Part I focuses on April–June (AMJ); Part II (Dezfuli and Nicholson 2013, hereafter Part II) focuses on October–December (OND) and on a comparison of interannual variability in the two seasons.

Section 2 of this article presents an overview of relevant past studies, including several dealing with eastern equatorial Africa. Section 3 describes data and methodology. It includes a brief summary of Dezfuli (2011a,b) and Part II, wherein the precipitation regions used in this study are delineated. Section 4 presents the mean climatology of the region, including precipitation, the wind regime and atmospheric circulation during April–June. Section 5 presents the results, examining three regional case studies in detail. These focus on the sea surface temperatures and atmospheric circulation patterns associated with AMJ seasons that are anomalously wet or dry. Section 6 examines several specific factors. Section 7 summarizes our main conclusions about the rainfall regime of this season and its contrasts with that in eastern equatorial Africa.

2. The equatorial rainy season of the boreal spring

Most regions of equatorial Africa experience two rainy seasons during the course of the year. These occur during the transition seasons, but with somewhat variable timing. The boreal spring is the main rainy season in eastern equatorial Africa, but it is the weaker of the two seasons in western equatorial Africa. It is traditionally considered to be associated with the northward equatorial transit of the intertropical convergence zone (ITCZ) and the mean ascent associated with it. However, over central Africa this zone remains well into the Northern Hemisphere throughout the year Dezfuli (2011b). During the boreal spring, the ITCZ’s position averages 15°N, while peak rainfall occurs at about 3° to 5°N. Thus the factors producing equatorial rainfall in the boreal spring are considerably more complex. A separate zone of ascent lies to the south of the ITCZ and is linked with the rainfall maximum. We term this the tropical rain belt.

Few studies have examined in detail the rainy season of the boreal spring in western equatorial Africa. The first to do so was that of Balas et al. (2007). Defining the season as March to May, they found that the association between the interannual variability of rainfall and tropical SSTs was both regionally and seasonally specific. The most pervasive associations are those of the Pacific ENSO and SSTs along the Benguela coast. During abnormally dry March–May seasons the El Niño pattern of Pacific SSTs is strongly developed. The study also concluded that an opposition between the Atlantic and Indian Oceans creates an east–west displacement of the convection. Changes in SSTs along the Benguela coast also appear to be associated with such an east–west displacement.

The link to SSTs along the Benguela coast is particularly strong in March and April at stations right along the coast from 5°S to about 12°S (Fig. 1). Areas that receive some 50 mm per month during cold-water years receive 100 to 300 mm in warm water years. The time scale of these variations is 5 to 6 years (Nicholson and Entekhabi 1987), the peak time scale of both ENSO and SST variation in the Atlantic and Indian Oceans (Nicholson and Nyenzi 1990; Nicholson 1996). This time scale also accounts for the largest proportion of annual variance throughout most of western equatorial Africa.

Fig. 1.
Fig. 1.

March and April rainfall along the coast in warm water years and cold water years (from Nicholson and Entekhabi 1987).

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

Jackson et al. (2009) examined convective activity and found it to be much stronger during the boreal spring than the boreal autumn in most, but not all of the region. They attributed this to the influence of the midtropospheric African easterly jet (AEJ) of the Southern Hemisphere (here called AEJ-S), which is strongly developed during the boreal autumn but weak or absent during the boreal spring. Convective activity, as indicated by the contribution of mesoscale convective systems (MCSs) and the frequency/intensity of lightning, is also much stronger in western equatorial Africa than in eastern equatorial Africa (Jackson et al. 2009; Mohr and Zipser 1996; Christian et al. 2003).

Convection and hence intraseasonal variability during the boreal spring appears to be linked to wave perturbations with a peak time scale of 5 to 6 days. These appear to be convectively coupled Kelvin waves, some of which originate in the equatorial Atlantic and propagate eastward (Nguyen and Duvel 2008). The waves do not trigger convective systems but enhance their development into larger, organized convection, especially over the Congo basin.

Much more research on the boreal spring rainy season of equatorial Africa has been carried out in eastern equatorial regions. The results are reviewed here because they may have some significance in interpreting our results for the more western regions. Compared to the boreal autumn, the spatial coherence of rainfall anomalies is weak (Nicholson 1996; Camberlin and Philippon 2002). The temporal coherence within the season is likewise weak and the links to atmospheric factors are markedly different in each month of the season (Beltrando 1990; Ambenje 1990; Macodras et al. 1989). May, in particular, stands out from the other months (Camberlin and Philippon 2002).

The causes of interannual variability of the boreal rainy season in eastern equatorial Africa are still largely unknown (Pohl and Camberlin 2006). The total amount of rainfall is dependent on a combination of several unrelated factors, such as number of rainfall events and their intensity and the onset and cessation of the season. Strong relationships to large-scale atmospheric or ocean anomalies are not apparent (Camberlin and Philippon 2002). As an example, the influence of ENSO is weak, absent, or confined to local areas such as coastal Tanzania (Nicholson and Kim 1997; Indeje et al. 2000; Mutai and Ward 2000; Kijazi and Reason 2005). The lack of large-scale control suggests that interannual variability is linked to internal “chaotic” factors in the atmosphere (Camberlin and Philippon 2002). Two such factors include the Madden–Julian oscillation (MJO) and equatorial westerly winds over the continent. Changes in MJO amplitude can account for 44% of the March-to-May seasonal rainfall variance (Pohl and Camberlin 2006). The equatorial winds appear to be the most robust link to rainfall during this season. Westerly wind anomalies are linked to rain events during this season (Camberlin and Wairoto 1997; Okoola 1998). Easterly anomalies at 700 mb characterize both dry years and dry spells within wet years (Okoola 1999a,b).

3. Data and methodology

Figure 2 shows the western equatorial sector considered in this study. It includes the Zaire basin and the highlands surrounding it on all sides. The western extension stretches along the Atlantic coast; the eastern extreme includes the western part of the Rift Valley highlands.

Fig. 2.
Fig. 2.

(left) Homogeneous rainfall regions for the April–June (AMJ) season, superimposed upon a map of topographic relief. Asterisks indicate stations utilized. The typical seasonal cycle for each region is also indicated. (right) Mean rainfall (mm) during the AMJ season for the period 1948 to 1988.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

We examine select aspects of rainfall variability for the April–June season in five regions within this geographical area. These regions were delineated (Dezfuli 2011a,b) using a combination of rotated principal component analysis and Ward’s clustering technique (e.g., Gong and Richman 1995; Unal et al. 2003; Rao and Srinivas 2006). Each is homogeneous with respect to the interannual variability of rainfall.

The regionalization was carried out using Tropical Rainfall Measuring Mission (TRMM) 3B42 precipitation estimates, which have a 0.25° × 0.25° spatial resolution. These are available online from the National Aeronautics and Space Administration (NASA; http://mirador.gsfc.nasa.gov). TRMM was used instead of gauge data to delineate regions because of the irregular spatial distribution of gauges in the region and the sparse network in some sectors. A comparison with regionalizations based on gauge data is presented in Dezfuli (2011b).

The regions bear some relationship to topography. Region 1, the easternmost, lies over the western highlands of the Rift Valley. Region 2, the westernmost region, lies roughly between the Atlantic coast and the highlands of Cameroon. Region 3 includes the highlands of Cameroon and the northern portion of the Zaire basin. Region 4 includes mainly the highlands of Central African Republic. Region 5 includes mainly the remaining areas of the Zaire basin and the northern slopes of the central African plateau.

Gauge data were utilized to produce time series for the regions shown in Fig. 2. The archive assembled by the first author includes 141 stations in the study area but those with more than 10% missing data were eliminated from the analysis. Most analyses in the study are limited to the period 1948 to 1988 because most stations are available throughout this period, allowing for a relatively homogeneous station network.

For each region, a standardized rainfall anomaly series for the three-month season is calculated, following Nicholson (1986). These time series (Fig. 3) are utilized for linear correlation and for identifying years that are anomalously wet or dry during the AMJ season. In each case, the four wettest and four driest AMJ seasons are identified (see Table 1) and used to construct composites of surface and atmospheric parameters. The period 1948–88 is chosen for the interannual analysis because it represents the most reliable continuous period of observations. There are individual years outside this period with adequate data for some of the regions. If the rainfall anomalies of those years are stronger than the extremes of the period 1948–88, we will include them in the composites. Thus, for two regions the year 1989 was considered as one of the extreme dry cases. A two-tailed t test was used to test the significance of correlation coefficients and a bootstrapping procedure was used to test the significance of the composite differences (Terray et al. 2003). The latter was chosen because the normality assumption of the composites may not be satisfied.

Fig. 3.
Fig. 3.

Regionally averaged rainfall for the AMJ season, 1948 to 1988, for regions 1 to 5. The data are presented as standardized anomalies from the mean for the period (anomaly divided by the standard deviation).

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

Table 1.

The four wettest and driest years for each region.

Table 1.

Notably, in all but region 1 there is a modest reduction in rainfall starting in the mid to late 1960s. This shift is roughly commensurate with the onset of drought conditions in the West African Sahel. The change is most strongly apparent in regions 2 and 3. In contrast, region 1 remained relatively dry until 1962, after which time positive anomalies were much more pronounced and frequent than negative anomalies. Except for region 1, there is some tendency for the wettest seasons to occur in the 1950s and 1960s and the driest to occur in the 1970s and 1980s (Table 1). For example, at least one of the years 1955 or 1966 is among the wettest years in regions 2 to 5.

The time series in Fig. 3 are correlated with sea surface temperatures and sea level pressure in the global tropics. The NOAA Extended Reconstructed SST (ERSST) V3 product, which is available at a 2° × 2° resolution, is utilized (Smith et al. 2008). Three regions that show particularly strong links to SSTs are then examined in detail, using composites of the four wettest and the four driest seasons. This includes composites of SST anomalies and various atmospheric parameters, such as winds, divergence, and vertical motion. A comparison is made with the mean climatology, described in section 4.

Surface pressure and other atmospheric variables are obtained from the National Centers for Environmental Prediction (NCEP)–National Center for Atmospheric Research (NCAR) reanalysis (Kalnay et al. 1996; Kistler et al. 2001). These are monthly data available on a 2.5° × 2.5° grid. For both datasets, anomalies are calculated with respect to the 1948–88 mean. Various shortcomings of the NCEP–NCAR reanalysis product have been discussed by several authors (e.g., Stickler and Brönnimann 2011). However, these have been used by the current authors in numerous published works and also validated over West Africa in the analysis by Grist and Nicholson (2001) via comparison with pilot balloon data. The results with NCEP were consistent with those from pibals, although they appeared to exaggerate some of the wet–dry contrasts. Moreover, the circulation features noted in NCEP data over Africa are extremely consistent with independent estimates of rainfall (e.g., vertical motion fields are strongly correlated in magnitude and space with rainfall). Furthermore, the results are internally consistent and consistent among several diverse analyses.

4. Mean climatology

Figure 2 shows the mean seasonal cycle for the five regions. Regions 1 and 2 exhibit the typical bimodal equatorial rainfall regime, with a minimum in the boreal summer. Region 1, the easternmost region, has more rainfall in the boreal spring, similar to eastern equatorial Africa. In region 2, the westernmost region, the second rainy season is the more intense. Regions 3 and 4 show a single maximum in the boreal summer and autumn. The seasonal cycle of region 5 represents a transition between the equatorial and subtropical rainfall regimes: a single dry season in the austral winter, but slight maxima in the two transition seasons.

Mean rainfall for the AMJ season is shown in Fig. 2. Within the study area of western equatorial Africa rainfall exceeds 25 mm month−1 in a zone that extends from roughly 13° to 14°N to 14°S. It exceeds 100 mm month−1 from roughly 10° to 5°N. Maximum rainfall occurs along the coast of eastern Nigeria and Cameroon, just north of the equator. The pattern for May (not shown) is much like that of the seasonal mean. For April (June) the latitudinal stretch receiving 25 mm or more is larger (smaller) and the pattern is shifted southward (northward) several degrees of latitude.

The region is strongly influenced by its proximity to the Atlantic Ocean. Aridity near the coast is enhanced by the presence of the St. Helena subtropical high (Fig. 4), which is most intense in the austral winter and has its core at roughly 30°S during the AMJ season. The high creates winds parallel to the coast, with upwelling in the latitudes 15° to 30°S (Fig. 4). The upwelling is weak during the AMJ season. Over the Atlantic the temperature maximum extends from roughly 5°N to the Guinea coast, and a weak “cold tongue” (e.g., Okumura and Xie 2004; Grodsky and Carton 2003) appears right at the equator.

Fig. 4.
Fig. 4.

Mean (top) sea surface temperatures (°C) and (bottom) sea level pressure (SLP; hPa) during AMJ for 1948 to 1988.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

Figure 5 shows the mean vector winds at four levels. In the mid and upper troposphere easterly winds are clearly dominant. The well-known African easterly jet that dominates the meteorology of the Sahel region is well developed at 600 hPa. Its core lies over the equatorial Atlantic and its mean speeds over the season exceed 12 m s−1. A weakly developed tropical easterly jet (TEJ) is evident at 200 hPa, but its speeds over Africa are lower than those of the AEJ. The weak TEJ may suggest an incomplete zonal circulation cell along the equatorial Atlantic (Hastenrath 2001). The mean speed of the TEJ is about 8 to 10 m s−1 over equatorial Africa and 10 to 12 m s−1 over the western Indian Ocean. In the lower troposphere at 925 and 850 hPa, southerly flow prevails in the western equatorial region. At 925 hPa the southerly flow is weak throughout most of the region. These winds start out as southeasterlies in the Southern Hemisphere and take on a southwesterly direction after they cross the equator. The westerly component becomes well developed near the surface.

Fig. 5.
Fig. 5.

Mean wind vectors at four levels during AMJ for 1948 to 1988.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

5. Results

Figure 6 shows the simultaneous correlation between AMJ rainfall and SSTs. Only three of the five regions appear to show significant correlations; these are discussed in detail in the following sections. However, large-scale patterns in the correlations suggest that in four of the five regions, high rainfall is associated with a preponderance of lower than normal SSTs over the tropical and subtropical oceans and negative correlations with sea level pressure over most of the tropics (not shown). Only region 1, the easternmost region that includes the western Rift Valley highlands, does not fit this pattern; a possible explanation is discussed in section 6.

Fig. 6.
Fig. 6.

Correlation between regional AMJ rainfall and concurrent sea surface temperatures for 1948–88. The 5% significance level is indicated by the dash-dotted line. Rainfall is also correlated with SST averages for the indicated boxes, with values given in Table 2.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

a. Region 2: The Atlantic coastal region

In the region stretching along the Atlantic coast high rainfall is associated with above normal SSTs in the equatorial and South Atlantic and in extratropical regions of the western Pacific. The correlations (Table 2) reach the 1% significance level in the Gulf of Guinea (r = 0.39) and along the Benguela coast of the eastern Atlantic (r = 0.48). Areas of significant negative correlation include the western tropical Indian Ocean (r = −0.39) and the eastern equatorial Pacific and western coast of South America (r = −0.39).

Table 2.

The correlation between rainfall and selected variables. The values of r corresponding to the 0.05 and 0.01 significance levels are 0.30 and 0.39, respectively. Variables are averaged over the indicated regions, which are also depicted in Figs. 6, 8, and 13.

Table 2.

SST anomalies for wet and dry composites (Fig. 7) show a general reversal of the anomaly sign between the wet and dry cases, with anomalies in the eastern Pacific respectively resembling typical La Niña/El Niño patterns. The SST anomalies are much greater in the dry case than in the wet case. Dry conditions along the Atlantic coast are also associated with generally positive anomalies in the Indian Ocean, but with negative anomalies in the equatorial Atlantic and along the Benguela coast of the Atlantic. The well-known Atlantic SST dipole—positive (negative) anomalies north (south) of the equator—associated with abnormally high rainfall in the Sahel (e.g., Lamb and Peppler 1992; Joyce et al. 2004) is also well developed in the dry composite.

Fig. 7.
Fig. 7.

Standardized SST anomalies for the wet composite and the dry composite of regions 2, 3, and 5. Anomalies are based on the period 1948–88.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

The association with sea level pressure helps to explain the SST anomalies over the Atlantic. Rainfall in the Atlantic coastal sector and sea level pressure are negatively correlated throughout the central and eastern sectors of the equatorial Atlantic (Fig. 8), indicating a weaker (stronger) South Atlantic subtropical high commensurate with wetter (drier) conditions.

Fig. 8.
Fig. 8.

Correlation between regional AMJ rainfall of region 2 and concurrent sea level pressure for the period 1948–88. The 5% significance level is indicated by the dash-dotted line. Rainfall is also correlated with SLP averages for the indicated box, with value given in Table 2.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

The weakening of the high influences Atlantic SSTs in three ways. The reduced southerly and southeasterly flow on its eastern flank reduces the advection of cold subpolar water and reduces upwelling. The relaxation of the equatorial trades brings warm water farther eastward along the equator. Such a condition is well known to occur in the boreal summer (Xie and Carton 2004; Carton and Huang 1994). Termed the “Atlantic Niño,” this phenomenon is analogous to the Pacific El Niño.

The strong contrast in coastal SSTs in the wet and dry composites suggests that local SSTs play a role in modulating rainfall in this coastal sector of Africa. However, an analysis of vertical motion and specific humidity suggests that the causal link to interannual variability is not the local SSTs but the changes in pressure that produce them. The potential influence of SSTs includes static stability changes, which influence vertical motion, and changes in atmospheric moisture. However, the wet and dry composites show little contrast in specific humidity (Fig. 9) and subsidence is actually greater near the surface in the wet composite (Fig. 10), when SSTs are anomalously high along the coast.

Fig. 9.
Fig. 9.

Latitude–height cross sections of specific humidity (g kg−1) for regions (top) 2, (middle) 3, and (bottom) 5 during AMJ. Specific humidity is zonally averaged over 10°–15°E, 10°–20°E, and 15°–30°E for regions 2, 3, and 5, respectively. The panel on the right shows the difference between the wet and dry composites; shading indicates areas where the difference is significant at 5% level.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

Fig. 10.
Fig. 10.

Latitude–height cross sections of omega (10−2 Pa s−1) for regions (top) 2, (middle) 3, and (bottom) 5 during AMJ. Omega is zonally averaged over 10°–15°E, 10°–20°E, and 15°–30°E, for regions 2, 3, and 5, respectively. Areas of rising motion are shaded. The panel on the right shows the difference between the wet and dry composites; shading indicates areas where the difference is significant at 5% level.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

The contrasts in vertical motion between the wet and dry composites can probably be attributed to the reduced (enhanced) influence of the subtropical high in the wet (dry) years rather than local SST anomalies. For the troposphere as a whole, the zone of subsidence associated with the subtropical high, which is directly over the Atlantic coastal region, is broader and more intense in the dry composite (Fig. 10). An even more striking contrast is apparent in the equatorial zone of ascent, just to the north. It is much broader and stronger in the wet composite.

The zonal wind field corresponding to wet and dry composites is shown locally in the cross sections in Fig. 11 and globally via anomalies at 850 mb and 200 hPa in Fig. 12. These figures indicate that in the wet composite, compared to the dry, the westerlies over the equatorial Atlantic and the West African continent are more strongly developed in the lower troposphere, while the easterlies are stronger at 200 hPa, particularly over the Atlantic (Fig. 12). The correlation between rainfall and wind is particularly high near the surface in the equatorial Atlantic, where it reaches 0.59 (Table 2). The correlation with 200-hPa winds reaches −0.53 over much of the Atlantic, but it is also strong over much of the continent (Fig. 13).

Fig. 11.
Fig. 11.

Latitude–height cross sections of zonal winds (m s−1) for regions 2 (top), 3 (middle) and 5 (bottom) during AMJ. Zonal wind is averaged over 10°–15°E, 10°–20°E, and 15°–30°E, for regions 2, 3, and 5, respectively. Areas of easterly winds are shaded. The panel on the right shows the difference between the wet and dry composites; shading indicates areas where the difference is significant at 5% level.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

Fig. 12.
Fig. 12.

(top) Wet-minus-dry composites of divergence (10−6 s−1) at 200 hPa, and zonal wind (m s−1) at (middle) 200 hPa and (bottom) 850 hPa. The three-panel diagrams clockwise from top left represent regions 2, 3, and 5, respectively. The dashed line shows areas in which the difference between the wet and dry composites is significant at the 5% level.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

Fig. 13.
Fig. 13.

Correlation between AMJ rainfall and zonal wind at 200, 850, and 1000 hPa for regions 2 (top), 3 (middle), and 5 (bottom). Correlations are based on the period 1948–88 and only those exceeding the 5% significance level are shown. Rainfall is also correlated with zonal wind averages for the indicated boxes, with values given in Table 2.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

A very striking feature is the strong and well-developed zone of low-level westerlies from the surface to nearly 700 hPa in the wet composite (Fig. 11). The stronger westerlies would promote ascent by 1) enhancing the wind component perpendicular to the coast, thereby 2) enhancing the orographic impact of the high terrain, and 3) enhancing surface convergence over the coastal region. The enhanced westerlies might also contribute to the ascent in the mid and upper troposphere. At the surface the flow is southwesterly. The intensification of this flow serves both to displace the ITCZ northward and intensify the convergence associated with it. Consequently, the ascending motion associated with the ITCZ links into the tropical rain belt (Fig. 10), contributing to its intensification. In the dry composite, these two zones of ascent are nearly independent.

The vertical structure of the zonal winds probably also contributes to the strong and wider column of ascent near the equator in the wet composite (Fig. 11). The vertical shear in the mid and upper troposphere (between 150 and 600 hPa) is much higher in the wet case: −8 m s−1 compared to 5 m s−1 in the dry case. Higher shear also characterizes years in which the tropical rain belt is anomalously intense over West Africa (Grist and Nicholson 2001; Nicholson 2009a).

These local changes in atmospheric circulation appear to be adequate explanations for the interannual variability of rainfall in the coastal sector. However, the significant correlations with SSTs in the Pacific and Indian Oceans (Fig. 6) suggest that some large-scale effects are also involved. The zonal wind at 200 hPa shows markedly stronger (weaker) easterlies [or weaker (stronger) westerlies] over the Atlantic (eastern Pacific) in the wet composite (Fig. 12). These changes in turn lead to reduced upper-level divergence over the eastern Pacific and Indian Oceans and enhanced upper-level divergence over most of Africa and the eastern Atlantic. This combination enhances vertical motion over Africa.

The omega fields for the wet and dry composites further indicate large changes in the intensity of the east–west zonal circulation over the entire tropics. In the wet (dry) composite the Pacific cell is stronger (weaker) and the Atlantic and Indian Ocean cells are weaker (stronger) (Fig. 14). Such intensification (weakening) of the Pacific cell characterizes La Niña (El Niño) episodes, consistent with the SST anomalies in the wet and dry composites. Thus, the global scale probably produces the local anomalies that enhance or reduce rainfall in coastal region 2.

Fig. 14.
Fig. 14.

Omega (10−2 Pa s−1) for wet, dry, and wet-minus-dry composites. The three-panel diagrams clockwise from top left represent regions 2, 3, and 5, respectively. Omega of each region is meridionally averaged over its corresponding latitudes. The dashed lines in the bottom panel of each set of diagrams show areas in which the difference between the wet and dry composites is significant at the 5% level.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

The changes that occur over the Atlantic in the wet and dry composites appear to result from a complex feedback between the SSTs and atmospheric circulation. The SST patterns in the equatorial Atlantic are such that the east–west SST gradient is reduced in the wet years and enhanced in the dry years. The result is a weakened (enhanced) east–west zonal circulation cell over the Atlantic in the wet (dry) years. Consequently, subsidence is reduced over the Atlantic coastal sector in the wet case, thus favoring increased rainfall. The reduced subsidence, in turn, weakens the subtropical high in the eastern Atlantic, which would further increase SSTs in that region.

b. Region 3: Cameroon highlands and northern Zaire basin

Region 3, the Cameroon highlands and northern Zaire basin, lies adjacent to region 2, the coastal Atlantic sector, and rainfall time series for the two regions are well correlated (r = 0.49). The two regions make an interesting comparison because the patterns of SSTs and winds associated with the wet and dry composites are very similar, but with a striking exception: the equatorial and South Atlantic SSTs. This contrast, apparent in every analysis, suggests that region 3 has a weaker association with the large scale.

Anomalously high (low) AMJ rainfall in both regions is associated with negative (positive) SST anomalies over the Indian Ocean and eastern tropical Pacific. The correlations are significant in the eastern equatorial Pacific and the equatorial Indian Ocean (Fig. 6; Table 2). Notable anomalously dry conditions in both region 2 and region 3 are associated with strong positive SST anomalies characteristic of El Niño and in the wet composites show negative La Niña–like anomalies in the eastern Pacific. In the equatorial and South Atlantic, the association with SSTs is very different for the two regions: strongly positive for the Atlantic coastal sector (region 2) and weakly negative for region 3, located to the north.

The links to zonal wind anomalies likewise indicates similar global-scale patterns for the two regions (but weaker associations for region 3) and marked contrast in the equatorial and South Atlantic. In the wet (dry) composites (Figs. 11 and 12) the upper-tropospheric easterlies and low-level westerlies are anomalously strong (weak) and midtropospheric easterlies are anomalously weak (strong). However, the anomalies over the Atlantic associated with region 3 are weak and do not reach the 5% significance level.

Over the Atlantic and western equatorial Africa, the wet and dry composites of region 3 show little contrast in either upper-level divergence or vertical motion (Figs. 12 and 14). These contrasts are strong for region 2: stronger (weaker) divergence at 200 hPa in the wet (dry) composites and stronger (weaker) subsidence over the continent.

Surprisingly, the vertical motion over most of the region is marginally stronger in the dry composite, although the differences are not highly significant. This is apparent in the global vertical motion fields (Fig. 14) and in the zonal cross sections of vertical motion (Fig. 10). The latter figure represents an average for the longitude span 10° to 20°E. Because of the northwest to southeast orientation of region 3 and the counterintuitive result of weaker vertical motion in the wet composite, zonal cross sections of vertical motion were also constructed for 10° to 15°E and 15° to 20°E (not shown). These confirm that the contrast in vertical motion between wet and dry composites is minimal for region 3, with marginally more intense vertical motion in the dry composite.

Overall, the analyses described above do not show any clear-cut reason for the occurrence of wet or dry extremes in region 3. It is likely that more local factors play a role. We hypothesize that the low-level westerlies, enhanced by a strong cross-equatorial pressure gradient, may be the key. The westerlies would enhance the orographic effects of the highlands over Cameroon. Rainfall is exceedingly high in parts of this region, reaching 10 000 to 14 000 mm yr−1 at Debundscha and other locations near Mt. Cameroon. The eastern extent of region 3 lies downstream from the highlands and might be influenced by rain systems developed over them. Consistent with this hypothesis, omega fields averaged for the latitudes of 0° to 7°N (not shown) indicate a local intensification of ascending motion over the highlands and over Mt. Cameroon in particular. Some evidence of that is also seen in Fig. 14, where in the wet composite there is a narrow but significant area of enhanced vertical motion in the midtroposphere around 0° to 10°E.

c. Region 5: Southern Zaire basin

The interannual variability of rainfall in region 5, which is predominantly Southern Hemisphere sectors of the Zaire basin and the slopes of the highlands to the south, does not appear to be closely linked to large-scale factors. The only large area with significant correlation between SSTs and rainfall (Fig. 6) is the central tropical Indian Ocean (r = −0.50). Notably, the overall spatial pattern of correlation with SSTs is similar to that of region 3, the northern Zaire basin and Cameroon highlands: wet (dry) conditions tend to be associated with negative (positive) SST anomalies in the global tropics (Fig. 7). However, the association is stronger with dry conditions in region 3 and wet conditions in region 5.

The link to the zonal wind is relatively weak compared to regions 2 and 3 (Fig. 13). However, the opposition of zonal anomalies at 200 and 850 hPa in the wet-minus-dry composites of Fig. 12 suggests that there is some association between rainfall in the southern Zaire basin and global-scale processes. The most direct association is with zonal winds over the eastern Atlantic around 10°S (i.e., in proximity to region 5). The negative (positive) correlations with 200 hPa (850 hPa) zonal wind suggest a weakening (intensification) of the zonal circulation over the south equatorial Atlantic in the wet (dry) composite (Fig. 13). The strongest correlation, +0.63, is with 850-hPa wind around 10°S. The global omega fields (Fig. 14) confirm the changes in the zonal circulation over the south equatorial Atlantic. Weaker changes in the zonal circulation over the Pacific and Indian Oceans are also apparent.

Over the continent, the wet/dry contrasts in zonal wind at 200 and 850 hPa are weaker (Fig. 11). There the strongest contrast is apparent at 600 hPa (Figs. 11 and 15), the level of the African easterly jet of the Northern Hemisphere (AEJ-N). In the wet (dry) composite the AEJ-N is exceedingly weak (strong) over equatorial Africa. The correlation of 600-hPa zonal wind with rainfall in region 5 reaches 0.5.

Fig. 15.
Fig. 15.

Zonal wind (m s−1, top) and divergence (10−6 s−1, bottom) at 600 hPa for the wet, dry, and wet-minus-dry composites of region 5.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

Another notable difference between the wet and dry composites for region 5 is the intensity and latitudinal extent of prevailing ascent (Fig. 10). In the wet composite, ascending motion extends above most of region 5, across a latitude span of roughly 20 degrees. In the dry composite the zone of ascent is displaced northward and is marginally weaker. The global omega fields of Fig. 14 also show weaker ascent in the dry composite between 10° and 40°E. The key to understanding the interannual variability of April–June rainfall in the Zaire basin is an explanation for this broad zone of ascent, but an explanation is not readily apparent from the foregoing analyses.

Traditionally the boreal spring rainfall in this region and the zone of rising motion associated with it are assumed to be linked to the northward advance of the ITCZ. Nicholson (2009b) showed the shortcomings of this paradigm over West Africa. A major one is the decoupling of the low-level convergence/rising motion of the ITCZ and the broad region of ascent throughout the troposphere that is associated with rainfall. The latter she termed the “tropical rain belt.” The two features are not decoupled over region 5. However, the “rain belt” is much broader than the region of ascent at ~13°N that is associated with the low-level ITCZ.

We speculate that local topography contributes to this zone of ascent and that the combination of topography and the aforementioned wind anomalies play an important role in interannual variability. The rainfall regime of the entire Zaire basin is dominated by an orographic system with ascent over the surrounding highlands during the day and descent at night. The descending flow off the highlands converges into the basin, resulting in abnormally strong mesoscale convective systems (Jackson et al. 2009; Zipser et al. 2006). Figure 10 shows that the mean zone of ascending motion coincides with the expanse of the Zaire basin, where this flow converges. Subsidence prevails over most of region 5 (the southern basin), which roughly spans the latitudes of 2° to 10°S. Region 5 lies on the slopes of the highlands to the south and receives rainfall from MCSs that form over the highlands and move downslope at night.

Jackson et al. (2009) showed that midlevel convergence over the Zaire basin appears to promote the development of MCSs. The much weaker AEJ-N in the wet composite for region 5 results in enhanced convergence at 600 hPa, suggesting enhanced MCS activity in the wet case. A similar mechanism was shown for the variability of summer rainfall over the central Sahel (Dezfuli and Nicholson 2011). The anomalously strong westerlies over the south equatorial Atlantic (Figs. 12 and 13) enhance the trigger for uplift over the highlands, a prerequisite for the MCS formation.

If a change in the number and/or intensity of MCS activity plays a major role in the interannual variability of rainfall in region 5, the southern Zaire basin, this should be evident in the frequency distribution of individual rain events. Daily rainfall in wet and dry years, examined from TRMM data, shows contrasts in rainfall intensity that are consistent with enhanced convective activity (Fig. 16). In two years with anomalously high AMJ rainfall in region 5, 2002 and 2006, rainfall exceeds 10 mm on 20 days (on 6 of which it exceeds 20 mm). It is greater than 5 mm on a total of 29 days. In the two years with anomalously low AMJ rainfall, 1998 and 2004, it exceeds 10 mm on only 2 days and exceeds 5 mm on only 11 days. Figure 17 shows the mean vertical motion in the wettest and driest 6-day periods of April during these four years. Maximum contrast occurs around 600 hPa, where the mean vertical motion for the wet period is 3 to 4 times greater than during the dry period. This is consistent with the occurrence of strong MSCs in this region.

Fig. 16.
Fig. 16.

Daily rainfall (mm) during AMJ for region 5 for the two wettest years (2002, 2006) and the two driest years (1998, 2004) of the period 1998–2010. The data are from TRMM 3B42.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

Fig. 17.
Fig. 17.

Omega (10−2 Pa s−1) for the wettest and driest 6-day periods in April of region 5 during the four years shown in Fig. 16. Wet composite is for 12–17 Apr 2006, and dry composite is for 25–30 Apr 1998.

Citation: Journal of Climate 26, 1; 10.1175/JCLI-D-11-00653.1

It is also worth noting that May was almost completely dry in the dry years but received substantial rainfall during the wet years. Hence a longer rainy season characterized the two wet years of Fig. 17. In the traditional paradigm of the ITCZ bringing rains to this region, this might be interpreted as an earlier northward progression of the ITCZ in the dry years. Two facts contradict that interpretation. One is that the surface ITCZ (the zone of low-level vertical motion at ~13°N in Fig. 10) is well to the north of region 5 in both composites, nor has it shifted northward in the dry composite.

Another factor that may play a role in this region is equatorially propagating Kelvin waves. During March and April these waves move eastward off the Atlantic and interact with convective systems that are typically triggered over the Rift Valley highlands (Nguyen and Duvel 2008). The presence of low-level westerlies enhances the propagation. Although the waves do not trigger convection, they modify their development into larger mesoscale convective systems. Pohl and Camberlin (2006) also note that the amplitude of the Madden–Julian oscillation influences interannual variability during this season.

6. Discussion

The analyses carried out in this study identify associations between rainfall in western equatorial Africa during the April–June season and the tropical oceans and atmosphere. In only one case, the Atlantic coastal region 2, can a definitive causal explanation for rainfall variability be put forth. This is the only region in which remote forcing appears to be the major control. In the other regions, some remote impact is apparent but the control appears to be largely local and strongly modulated by orographic effects. For these two cases, hypotheses concerning the controls on interannual variability are presented.

For all three case studies, anomalies in the local zonal winds are similar. In each of these regions, the wet composite is associated with anomalously strong low-level westerlies and upper-tropospheric easterlies, but reduced easterly winds in the midtroposphere and a poleward displacement of the upper-tropospheric westerlies in both hemispheres. The opposite pattern prevails in the dry composites. These patterns are best developed in region 2, along the Atlantic coast, and most weakly developed for region 5, the central Zaire basin.

The wet and dry extremes in region 2 can be readily explained in terms of large-scale SST, pressure, and upper-level wind anomalies that alter the east–west zonal circulation. Changes in these variables alter both the low-level wind field over the equatorial Atlantic and the vertical motion field over western equatorial Africa. The impact is probably locally enhanced by changes in the intensity of the South Atlantic subtropical high and by the impact of the low-level westerlies as they meet the coastal terrain. Stronger westerlies enhance orographic effects, low-level convergence, and vertical and horizontal shear.

Region 3 shows a weaker relationship to the large-scale, despite an apparent relationship between El Niño–like SST anomalies and dry conditions. In both the wet and dry composites the SST anomalies in the Atlantic serve to enhance zonal SST gradients across the Atlantic. Consequently, the interannual variability is probably controlled by local factors that are modulated by the global tropical circulation. This is a region of high terrain and intense convective activity. Our hypothesis is that the low-level westerlies have a large impact by enhancing the orographic effects in this region, where annual rainfall can reach 14 000 mm in some areas.

Region 5, the southern Zaire basin, shows even weaker links to the large scale. A similar conclusion was drawn by Pohl and Camberlin (2006), who examined the March–May season in an area of East Africa that extends into region 5. They did not find a link to large-scale processes. They did show that interannual variability is clearly linked to a change in the intensity of individual rainfall events, consistent with our finding for region 5. Topography plays a large role in determining the intensity of convection in this region (Jackson et al. 2009). A change in the intensity of the Atlantic zonal circulation cell also contributes.

In all three regions examined contrasts between wet and dry composites are apparent in the global zonal circulation cells (Fig. 14). The strongest anomalies arise over equatorial Africa when the zonal circulation is anomalous over both the Pacific and Indian Oceans. This is the case for Atlantic coastal region 2, with strong wet/dry contrasts in vertical motion over Africa from 30°W to 30°E. More moderate contrasts in the zonal cell over Africa are apparent for regions 3 and 5. In the former case, the Cameroon highlands and northern Zaire basin, contrasts are strong in the Pacific cell but relatively weak in the Indian Ocean cell. The opposite is true for region 5, the southern Zaire basin.

These results collectively suggest that the Pacific Ocean has little direct influence on western equatorial Africa. Its influence is manifested via changes in the zonal circulation over the Atlantic and Indian Ocean sectors. This is consistent with the conclusions of numerous studies that have compared the influence of the Indian and Pacific Oceans on rainfall in East Africa (e.g., Clark et al. 2003; Black et al. 2003; Behera et al. 2005; Ummenhofer et al. 2009; Hastenrath et al. 2011). However, those studies have examined the East African “short rains” season of boreal autumn. Our results indicate a similar situation in the boreal spring rainy season (i.e., AMJ).

A synthesis of our results suggests that, while some regions show a strong association with local SSTs in the eastern Atlantic or western Indian Oceans, the association may not be a causal one. Large changes in the atmospheric circulation over the whole of the tropical Pacific and Indian Oceans are associated with the interannual variability of rainfall over western equatorial Africa. These represent coupled variability in the ocean and atmosphere over large sectors of the tropics. Variations in local cloudiness, local wind stress, or remote forcing of wind stress can all induce changes in SSTs (Hirst and Hastenrath 1983a).

Thus, some of the associations that have been demonstrated between SSTs and rainfall, especially the more local links to the Gulf of Guinea or Benguela coast, can reflect the common forcing by the large-scale atmosphere, which at the same time impacts equatorial rainfall via vertical motion fields and zonal winds. The modified SST fields work in tandem to enhance the effects. This is similar to the situation described for the link between East African rainfall during the boreal autumn and the Indian Ocean (e.g., Hastenrath et al. 2011).

The stronger association with atmospheric processes is confirmed by the correlation between rainfall, SSTs, SLP, and zonal winds. For region 2, the absolute values of the strongest correlations with SSTs range from 0.39 to 0.48. Absolute values of the correlation with zonal wind range from 0.52 to 0.59; correlation with sea level pressure over the equatorial Atlantic is −0.57. For region 3 the range with SSTs is −0.32 to −0.46, but 0.48 to 0.54 for wind. For region 5 the correlation with low-level zonal wind over the Atlantic reaches 0.63.

A direct effect of local SSTs would be seen as changes in specific humidity, moisture convergence, or static stability. The change in specific humidity between wet and dry composites for all three case studies is negligible (Fig. 9). Moisture convergence (not shown) does not show consistent contrast either. Vertical motion (e.g., Fig. 10) suggests that the changes in static stability are probably confined to the layers just above the surface. Prior studies similarly examined the links between rainfall and SSTs in this region (Hirst and Hastenrath 1983a,b). Impacts of coastal SSTs on atmospheric moisture and stability were demonstrated, but they appeared to have a clear association with rainfall variability only at stations right on the coast (Hirst and Hastenrath 1983a). In the interior (i.e., in the Zaire basin), the more important factor appeared to be changes in the zonal circulation over the Atlantic (Hirst and Hastenrath 1983b), consistent with our conclusions about the importance of the zonal circulation. It is worth noting that since there are some concerns about using NCEP–NCAR reanalysis data, we evaluated our hypotheses utilizing 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) data (not shown). The conclusions suggested by both datasets were overall consistent.

7. Summary and conclusions

This study has examined rainfall variability in five sectors of western equatorial Africa during the AMJ season. Large-scale forcing was apparent for three of the regions, with wet (dry) conditions being associated with positive (negative) SST anomalies in the tropical Pacific and Indian Oceans. The large-scale forcing is particularly strong for region 2, stretching along the Atlantic coast. In the remaining two, the southern Zaire basin and the Cameroon highlands/northern Zaire basin, the forcing appears to have a strong local component representing a combination of topographic and mesoscale effects.

A critical factor is the intensity of the east–west zonal circulations over the three tropical ocean basins. In some cases an ENSO-like signal in both SSTs and the zonal circulation over the Pacific is evident in association with the interannual variability of rainfall. However, changes in the Pacific sector do not appear to impact Africa directly but instead modulate the zonal circulation over Africa via influences on the Atlantic and Indian Oceans.

Consistent changes in the zonal winds over Africa are apparent in association with wet and dry conditions in all three regions evaluated in depth (i.e., the Atlantic coastal sector, Cameroon highlands/northern Zaire basin, and southern Zaire basin). Wet (dry) composites show an intensification (weakening) of the tropical easterly jet and low-level equatorial westerlies, but a weakening (intensification) of the midlevel easterly flow. Studies of variability in other sections of Africa have likewise identified these aspects of the zonal flow as important factors (e.g., Segele et al. 2009; Camberlin 1995; Vigaud et al. 2007; Todd and Washington 2004).

This study also demonstrated numerous significant correlations between rainfall in western equatorial Africa and SSTs. In most cases, it appears that the local SSTs and rainfall are forced by the same remote, large-scale atmospheric and oceanic patterns, including a coupled ENSO-like change in the Pacific and changes in the zonal circulation over the Indian Ocean. Indicative of this is the higher correlations with atmospheric parameters than with SSTs. The more local SSTs may enhance the impact of remote forcing. Our finding on indirect effect of local SSTs mainly relies on the NCEP–NCAR reanalysis data because of unavailability of soundings for the years examined. This mechanism, however, may need to be reconfirmed in future when such data are available. Using regional climate models would be an alternative tool to reevaluate this hypothesis. Our work also suggests that in the equatorial regions the zonal SST gradient may be more important than the sign of SST anomalies, consistent with studies of the boreal autumn rains of East Africa.

Acknowledgments

This work was supported by the National Science Foundation Grants 00047479, 016820, and 0813930 and also by the National Oceanic and Atmospheric Administration Grant NA09OAR4310731. We would like to acknowledge the help of Douglas Klotter for programming and production of figures. The authors are also thankful to three anonymous reviewers for their critical analysis of the manuscript.

REFERENCES

  • Ambenje, P. G., 1990: Mean tropospheric motion field over Kenya during the 1984 long rains season (March–May). Extended Abstracts, Third WMO Symp. on Meteorological Aspects of Tropical Droughts with Emphasis on Long-Range Forecasting, Niamey, Niger, WMO, 83.

    • Search Google Scholar
    • Export Citation
  • Balas, N., , S. E. Nicholson, , and D. Klotter, 2007: The relationship of rainfall variability in west central Africa to sea-surface temperature fluctuations. Int. J. Climatol., 27, 13351349.

    • Search Google Scholar
    • Export Citation
  • Behera, S. K., and Coauthors, 2005: Paramount impact of the Indian Ocean dipole on East African short rains: A CGCM study. J. Climate, 18, 45144530.

    • Search Google Scholar
    • Export Citation
  • Beltrando, G., , and D. L. Cadet, 1990: Interannual variability of the short rain season in East Africa: Relationships with general atmospheric circulation. Veille Climatique Satell., 33, 1936.

    • Search Google Scholar
    • Export Citation
  • Black, E., , J. Slingo, , and K. R. Sperber, 2003: An observational study of the relationship between excessively strong short rains in coastal East Africa and Indian Ocean SST. Mon. Wea. Rev., 131, 7494.

    • Search Google Scholar
    • Export Citation
  • Camberlin, P., 1995: June–September rainfall in north-eastern Africa and atmospheric signals over the tropics: A zonal perspective. Int. J. Climatol., 15, 773783.

    • Search Google Scholar
    • Export Citation
  • Camberlin, P., , and J. Wairoto, 1997: Intraseasonal wind anomalies related to wet and dry spells during the “long” and “short” rainy seasons in Kenya. Theor. Appl. Climatol., 58, 5769.

    • Search Google Scholar
    • Export Citation
  • Camberlin, P., , and N. Philippon, 2002: The East African March–May rainy season: Associated atmospheric dynamics and predictability over the 1968–97 period. J. Climate, 15, 10021019.

    • Search Google Scholar
    • Export Citation
  • Carton, J. A., , and B. Huang, 1994: Warm events in the tropical Atlantic. J. Phys. Oceanogr., 24, 888903.

  • Christian, H. J., and Coauthors, 2003: Global frequency and distribution of lightning as observed from space by the Optical Transient Detector. J. Geophys. Res.,108, 4005, doi:10.1029/2002JD002347.

  • Clark, C. O., , P. J. Webster, , and J. E. Cole, 2003: Interdecadal variability of the relationship between the Indian Ocean zonal mode and East African coastal rainfall anomalies. J. Climate, 16, 548554.

    • Search Google Scholar
    • Export Citation
  • Dezfuli, A. K., 2011a: Spatio-temporal variability of seasonal rainfall in western equatorial Africa. Theor. Appl. Climatol., 104, 5769.

    • Search Google Scholar
    • Export Citation
  • Dezfuli, A. K., 2011b: The relationship of interannual variability in western equatorial Africa to the tropical oceans and atmospheric circulation. Ph.D. dissertation, The Florida State University.

  • Dezfuli, A. K., , and S. E. Nicholson, 2011: A note on the long-term variations of the African easterly jet. Int. J. Climatol., 31, 20492054.

    • Search Google Scholar
    • Export Citation
  • Dezfuli, A. K., , and S. E. Nicholson, 2013: The relationship of rainfall variability in western equatorial Africa to the tropical oceans and atmospheric circulation. Part II: The boreal autumn. J. Climate,26, 66–84.

  • Gong, X., , and M. B. Richman, 1995: On the application of cluster analysis to growing season precipitation data in North America east of the Rockies. J. Climate, 8, 897931.

    • Search Google Scholar
    • Export Citation
  • Grist, J. P., , and S. E. Nicholson, 2001: A study of the dynamic factors influencing the interannual variability of rainfall in the West African Sahel. J. Climate, 14, 13371359.

    • Search Google Scholar
    • Export Citation
  • Grodsky, S. A., , and J. A. Carton, 2003: The intertropical convergence zone in the South Atlantic and the equatorial cold tongue. J. Climate, 16, 723733.

    • Search Google Scholar
    • Export Citation
  • Hastenrath, S., 2001: In search of zonal circulations in the equatorial Atlantic sector from the NCEP–NCAR reanalysis. Int. J. Climatol., 21, 3747.

    • Search Google Scholar
    • Export Citation
  • Hastenrath, S., , D. Polzin, , and C. Mutai, 2011: Circulation mechanisms of Kenya rainfall anomalies. J. Climate, 24, 404412.

  • Hirst, A. C., , and S. Hastenrath, 1983a: Atmosphere–ocean mechanisms of climate anomalies in the Angola–tropical Atlantic sector. J. Phys. Oceanogr., 13, 11461157.

    • Search Google Scholar
    • Export Citation
  • Hirst, A. C., , and S. Hastenrath, 1983b: Diagnostics of hydrometeorological anomalies in the Zaire (Congo) basin. Quart. J. Roy. Meteor. Soc., 109, 881892.

    • Search Google Scholar
    • Export Citation
  • Indeje, M., , F. H. M. Semazzi, , and L. J. Ogallo, 2000: ENSO signals in East African rainfall seasons. Int. J. Climatol.,20, 19–46.

  • Jackson, B., , S. E. Nicholson, , and D. Klotter, 2009: Mesoscale convective systems over western equatorial Africa and their relationship to large-scale circulation. Mon. Wea. Rev., 137, 12721294.

    • Search Google Scholar
    • Export Citation
  • Joyce, T. M., , C. Frankignoul, , J. Yang, , and H. E. Phillips, 2004: Ocean response and feedback to the SST dipole in the tropical Atlantic. J. Phys. Oceanogr., 34, 25252540.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471.

  • Kijazi, A. L., , and C. J. C. Reason, 2005: Relationships between intraseasonal rainfall variability of coastal Tanzania and ENSO. Theor. Appl. Climatol., 82, 153176.

    • Search Google Scholar
    • Export Citation
  • Kistler, R., and Coauthors, 2001: The NCEP–NCAR 50-Year Reanalysis: Monthly means CD-ROM and documentation. Bull. Amer. Meteor. Soc., 82, 247267.

    • Search Google Scholar
    • Export Citation
  • Laing, A., , and J. M. Fritsch, 1993: Mesoscale convective complexes in Africa. Mon. Wea. Rev., 121, 22542263.

  • Laing, A., , R. Carbone, , V. Levizzani, , and J. Tuttle, 2008: The propagation and diurnal cycles of deep convection in northern tropical Africa. Quart. J. Roy. Meteor. Soc., 134, 93109.

    • Search Google Scholar
    • Export Citation
  • Lamb, P. J., , and R. Peppler, 1992: Further case studies of tropical Atlantic surface atmospheric and oceanic patterns associated with sub-Saharan drought. J. Climate, 5, 476488.

    • Search Google Scholar
    • Export Citation
  • Lienou, G., and Coauthors, 2008: Evolution of hydrological regimes in the equatorial area of Cameroon: An impact of climate variability in equatorial Africa? Hydrol. Sci. J., 53, 789801.

    • Search Google Scholar
    • Export Citation
  • Macodras, M. W., , P. N. Nthusi, , and J. Mwikya, 1989: Synoptic features associated with the failure of 1984 long rains in Kenya. Coping with Drought in Kenya: National and Local Strategies, T. E. Downing, K. W. Gitu, and C. M. Kamau, Eds., Lynne Rienner, 69–81.

  • McCollum, J. R., , A. Gruber, , and M. B. Ba, 2000: Discrepancy between gauges and satellite estimates of rainfall in equatorial Africa. J. Appl. Meteor., 39, 666679.

    • Search Google Scholar
    • Export Citation
  • Mohr, K. L., , and E. J. Zipser, 1996: Mesoscale convective systems defined by their 85-GHz ice scattering signature: Size and intensity comparison over tropical oceans and continents. Mon. Wea. Rev., 124, 24172437.

    • Search Google Scholar
    • Export Citation
  • Mutai, C. C., , and M. N. Ward, 2000: East African rainfall and the tropical circulation/convection on intraseasonal to interannual timescales. J. Climate, 13, 39153939.

    • Search Google Scholar
    • Export Citation
  • Nguyen, H., , and J. P. Duvel, 2008: Synoptic wave perturbations and convective systems over equatorial Africa. J. Climate, 21, 63726388.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., 1986: The spatial coherence of African rainfall anomalies: Interhemispheric teleconnections. J. Climate Appl. Meteor., 25, 13651381.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., 1996: A review of climate dynamics and climate variability in Eastern Africa. The Limnology, Climatology, and Paleoclimatology of the East African Lakes, T.C. Johnson and E. O. Odada, Eds., Gordon and Breach Publications, 25–56.

  • Nicholson, S. E., 2009a: On the factors modulating the intensity of the tropical rainbelt over West Africa. Int. J. Climatol., 29, 673689.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., 2009b: A revised picture of the structure of the “monsoon” and land ITCZ over West Africa. Climate Dyn., 32, 11551171.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., , and D. Entekhabi, 1987: Rainfall variability in equatorial and southern Africa: Relationships with sea surface temperatures along the southwestern coast of Africa. J. Climate Appl. Meteor., 26, 561578.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., , and B. S. Nyenzi, 1990: Temporal and spatial variability of SSTs in the tropical Atlantic and Indian Oceans. Meteor. Atmos. Phys., 42A, 117.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., , and I. M. Palao, 1993: A re-evaluation of rainfall variability in the Sahel. Part I. Characteristics of rainfall fluctuations. Int. J. Climatol., 13, 371389.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., , and J. Y. Kim, 1997: The relationship of the El Niño–Southern Oscillation to African rainfall. Int. J. Climatol., 17, 117135.

    • Search Google Scholar
    • Export Citation
  • Nicholson, S. E., , and J. P. Grist, 2003: The seasonal evolution of the atmospheric circulation over West Africa and equatorial Africa. J. Climate, 16, 10131030.

    • Search Google Scholar
    • Export Citation
  • Okoola, R. E., 1998: Spatial evolutions of the active convective patterns across the equatorial eastern Africa region during Northern Hemisphere spring season using outgoing longwave radiation records. Meteor. Atmos. Phys., 66, 5163.

    • Search Google Scholar
    • Export Citation
  • Okoola, R. E., 1999a: A diagnostic study of the eastern Africa monsoon circulation during the Northern Hemisphere spring season. Int. J. Climatol., 19, 143168.

    • Search Google Scholar
    • Export Citation
  • Okoola, R. E., 1999b: Midtropospheric circulation patterns associated with extreme dry and wet episodes over equatorial eastern Africa during the northern hemisphere spring. J. Appl. Meteor., 38, 11611169.

    • Search Google Scholar
    • Export Citation
  • Okumura, Y., , and S.-P. Xie, 2004: Interaction of the Atlantic equatorial cold tongue and the African monsoon. J. Climate, 17, 35893602.

    • Search Google Scholar
    • Export Citation
  • Pohl, B., , and P. Camberlin, 2006: Influence of the Madden–Julian oscillation on East Africa rainfall. II: March–May season extremes and interannual variability. Quart. J. Roy. Meteor. Soc., 132, 25412558.

    • Search Google Scholar
    • Export Citation
  • Rao, A. R., , and V. V. Srinivas, 2006: Regionalization of watersheds by hybrid-cluster analysis. J. Hydrol., 318, 3756.

  • Samba, G., , D. Nganga, , and M. Mpounza, 2007: Rainfall and temperature variations over Congo-Brazzaville between 1950 and 1998. Theor. Appl. Climatol., 91, 8597.

    • Search Google Scholar
    • Export Citation
  • Segele, Z. T., , P. J. Lamb, , and L. A. Leslie, 2009: Large-scale atmospheric circulation and global sea surface temperature associations with Horn of Africa June–September rainfall. Int. J. Climatol., 29, 10751100.

    • Search Google Scholar
    • Export Citation
  • Smith, T. M., , R. W. Reynolds, , T. C. Peterson, , and J. Lawrimore, 2008: Improvements to NOAA’s historical merged land–ocean surface temperature analysis (1880– 2006). J. Climate, 21, 22832296.

    • Search Google Scholar
    • Export Citation
  • Stickler, A., , and S. Brönnimann, 2011: Significant bias of the NCEP/NCAR and twentieth-century reanalyses relative to pilot balloon observations over the West African monsoon region (1940–1957). Quart. J. Roy. Meteor. Soc., 137, 14001416.

    • Search Google Scholar
    • Export Citation
  • Terray, P., , P. Delecluse, , S. Labattu, , and L. Terray, 2003: Sea surface temperature associations with the late Indian summer monsoon. Climate Dyn., 21, 593618.

    • Search Google Scholar
    • Export Citation
  • Todd, M. C., , and R. Washington, 2004: Climate variability in central equatorial Africa: Influence from the Atlantic sector. Geophys. Res. Lett., 31, L23202, doi:10.1029/2004GL020975.

    • Search Google Scholar
    • Export Citation
  • Ummenhofer, C. C., , A. Sen Gupta, , M. E. England, , and C. J. C. Reason, 2009: Contribution of Indian Ocean sea surface temperatures to enhanced East African rainfall. J. Climate, 22, 9931013.

    • Search Google Scholar
    • Export Citation
  • Unal, Y., , T. Kindap, , and M. Karaca, 2003: Redefining the climate zones of Turkey using cluster analysis. Int. J. Climatol., 23, 10451055.

    • Search Google Scholar
    • Export Citation
  • Vigaud, N., , Y. Richard, , M. Rouault, , and N. Fauchereau, 2007: Water vapour transport from the tropical Atlantic and summer rainfall in tropical southern Africa. Climate Dyn., 28, 113123.

    • Search Google Scholar
    • Export Citation
  • Xie, S.-P., , and J. A. Carton, 2004: Tropical Atlantic variability: Patterns, mechanisms, and impacts. Earth’s Climate: The Ocean–Atmosphere Interaction, Geophys. Monogr., Vol. 147, Amer. Geophys. Union, 121–142.

  • Yin, X. G., , and A. Gruber, 2010: Validation of the abrupt change in GPCP precipitation in the Congo River basin. Int. J. Climatol., 30, 110119.

    • Search Google Scholar
    • Export Citation
  • Zipser, E. J., , D. J. Cecil, , C. T. Liu, , S. W. Nesbitt, , and D. P. Yorty, 2006: Where are the most intense thunderstorms on Earth? Bull. Amer. Meteor. Soc., 87, 10571071.

    • Search Google Scholar
    • Export Citation
Save