The common explanation for the progression of the rainy season over Africa is the seasonal excursion of the ITCZ. The ITCZ paradigm stems from a time when tropical rainfall was assumed to be associated mainly with localized convection. Its development was also linked to the emergence of midlatitude frontal concepts. The paradigm has numerous shortcomings, including the diversity of definitions and the large number of parameters used to identify the ITCZ. A historical look at the concept shows that its use over Africa has long been controversial, with many eminent tropical meteorologists harshly criticizing its applicability over this continent. However, the seasonal excursion of the ITCZ remains the classical explanation for African rainy seasons, especially in the equatorial region. This article underscores the shortcomings of the paradigm in equatorial Africa by examining various aspects of the circulation associated with the spatial and temporal patterns of rainfall during the equatorial rainy seasons. The overall conclusion is that a deeper understanding of the seasonal cycle in the equatorial regions of Africa still needs to be developed.
This article challenges the assumption that the seasonal cycle of rainfall in equatorial Africa is controlled by the seasonal excursion of the ITCZ and calls for additional research on the seasonal cycle.
The intertropical convergence zone (ITCZ) has long been assumed to be a major control on tropical rainfall over both oceans and land. Over Africa, the cycle of the rainfall seasons is generally associated with its north–south displacement as this zone “follows” the sun. The assumed link to the tropical rainy seasons stems back to a time when tropical rainfall was thought to be produced primarily by local thunderstorms, with the ascent in the ITCZ promoting their development when thermodynamic conditions were favorable. This assumption of purely localized convection has long been overturned, with numerous studies showing the importance of mesoscale convective systems (e.g., Nesbitt et al. 2000; Nesbitt and Zipser 2003) and globally propagating convection-triggering waves (e.g., Mekonnen et al. 2008; Janiga and Thorncroft 2016) and the role of inertial instability in generating convection (e.g., Tomas and Webster 1997). In view of this evolution of our understanding of tropical rainfall, it seems meaningful to also reexamine the ITCZ concept. Here, this question is raised specifically concerning equatorial Africa, where rainy seasons occur twice annually—in the boreal spring and autumn.
This article commences with an overview of various published definitions of the ITCZ. This is followed by a historical look at the development of the image of the ITCZ over Africa, as well as controversies surrounding it. The development of the current scenario for West Africa is then described. The availability of reanalysis datasets permits the evaluation of this scenario. Here, the issue is examined for the two equatorial rainy seasons, focusing on April and November.
DEFINITIONS OF THE ITCZ.
The ambiguity of the ITCZ concept is made clear with a brief survey of definitions in the literature. The usages differ with respect to the variable chosen to define this zone. In the Encyclopedia of World Climatology, Yan (2005) states that the ITCZ is “an east-west oriented low-pressure region near the equator where surface northeasterly and southeasterly trade winds meets”; that is, the emphasis is on pressure. According to Miller (1996) in the Encyclopedia of Weather and Climate, it is “a region near the equator where the trade winds converge.” The American Meteorological Society’s (AMS) Glossary of Meteorology (Glickman 2000) provides a similar definition: “The axis, or a portion thereof, of the broad trade-wind current of the tropics” and the “dividing line between the southeast trades and the northeast trades.” A second definition there equates it with the meteorological equator. Holton et al. (1971) define the ITCZ as the “loci of cloud clusters associated with westward-propagating tropical wave disturbances,” a definition shared by Lockwood (1974).
Perhaps as a result of this ambiguity in definition, the tracking of the ITCZ has variously been based on the pressure minimum, surface wind convergence (Grodsky et al. 2003), the rainfall maximum (Sultan and Janicot 2000; Philander et al. 1996), the vorticity maximum (Magnusdottir and Wang 2008), the minimum in outgoing longwave radiation (Gu and Zhang 2002), or the cloudiness maximum (Waliser and Gautier 1993). The availability of satellite photos has led to the last two parameters being frequently utilized as a matter of convenience (Waliser and Gautier 1993). The use of so many different parameters has been justified by the long-held assumptions that 1) the pressure minimum and rainfall maximum are collocated with each other and with the wind convergence, 2) maximum cloudiness is roughly collocated with maximum rainfall, and 3) longwave radiation is at a minimum at that location. Unfortunately, these assumptions, especially the first one, do not stand up to close scrutiny. Even over the oceans the zone of minimum pressure does not generally coincide with that of the wind convergence or the rainfall maximum (Tomas and Webster 1997; Toma and Webster 2010). Over West Africa even the maximum in surface convergence is some 350 km south of the surface wind discontinuity between easterly and westerly winds (Hastenrath 1988).
Notably, most definitions emphasize the convergence of the trade winds. This may be appropriate over some oceans sectors. However, for the most part, the trade winds do not exist over the tropical landmasses. Thus, over Africa depictions of the surface ITCZ instead represent the meeting of the northeasterlies and southwesterly monsoon flow. However, the term is all too frequently applied to the rainfall maximum. Recognizing that the surface convergence zone and the rainfall maximum are not closely coupled, many authors now avoid the use of the term ITCZ in discussing rainfall over Africa. For example, in describing the rainfall maximum, Ross and Krishnamurti (2007) prefer the term equatorial rain belt. Zhang et al. (2006) use the term rain band and Nicholson (2009) substitutes the term tropical rain belt. The latter term is used in this article. Clearly, over most of the tropical landmasses, and over Africa in particular, the use of the term ITCZ should be avoided except in coastal areas where the trade winds are influential.
HISTORICAL DEVELOPMENT OF THE ITCZ CONCEPT.
A search of the early meteorological literature uncovers just as much confusion and ambiguity about the ITCZ and its historical origin. Nieuwolt (1977), in his book on tropical meteorology, suggests that the concept goes back to Hadley’s (1735) model. Hadley, however, did not mention the convergence, but it is implicit in his model of the vertical cells. The concept had definitely come into vogue by the 1920s and 1930s, when meteorologists attempted to apply midlatitude frontal concepts of the Bergen school to the tropics (Barry and Chorley 1992; Palmer 1951). The first mention of the convergence of the trade winds between the two hemispheres may have been in a paper by Brooks and Braby (1921) entitled “The Clash of the Trades in the Pacific.” This feature, identified by streamline confluence and not horizontal wind convergence, became known as the intertropical front (ITF). When the importance of wind convergence in tropical weather was realized in the 1940s and 1950s (Barry and Chorley 1992), the trade wind convergence was designated the intertropical convergence zone, initially abbreviated as ITC (Fletcher 1945) (Fig. 1).
Most of the work that led to the development of the ITCZ concept was based on conditions over the Pacific. Later, Bjerknes et al. (1933) described the ITF over all three tropical oceans and also attempted to trace its course over land in Africa and Southeast Asia. Petterssen (1941, 1958) further promoted the global picture of an ITCZ but defined it as the meeting of the trade winds. He used it to explain the basic rainfall pattern that exists over much of the tropics: bimodal rainfall seasonality in the equatorial latitudes with the twice-equatorial transit of the ITCZ, unimodal in the outer tropics when the ITCZ reaches its extreme latitudinal positions. Palmén and Newton (1969) subsequently published a global picture of surface streamlines with the land portions essentially representing a subjective compromise between the locations shown by the analyses of Mintz and Dean (1952) and Riehl (1954). Variants of their maps are repeated in nearly every climatological textbook, as well as in a multitude of articles and textbooks from a broad range of disciplines. The portion over Africa was generally adopted as the January and July locations of the ITCZ (Fig. 2). Eventually, the picture emerged of a global zone in which the pressure minimum, cloud and rainfall maxima, and low-level wind convergence coincided.
These ideas were not without controversy. Trewartha (1943) pointed out the ambiguity of the ITCZ and the rather capricious approach to delineating it. He states that there is “considerable vagueness in exactly defining this front, some writers making it synonymous with the doldrums, others locating it at the equatorial margins of the trades.” Riehl (1979), in his classical book on tropical meteorology, focused attention on the pressure field (i.e., the equatorial trough). He argued against the use of the term ITCZ, stating that convergence is intermittent and that the term relates back to “old times when it was thought that the meeting of northern and southern air masses occurred right at the equator.” One fact that Riehl cited as contrary to this picture was the independent seasonal displacement of the rainfall maximum, compared to the displacement of the pressure minimum and surface convergence. Palmén and Newton (1969) pointed out that, in what they termed the trade confluence zone, there were generally relatively clear skies and little precipitation. Ramage (1971) was particularly harsh in his criticism of the ITCZ concept. He states that the original concept of the intertropical front or ITF confused tropical meteorologists, who noted that “the worst weather” and the ITF do not coincide. He concluded that the confusion was compounded when the term ITCZ came into use and that “confusion became chaos” when the tropical meteorologists started to use the terms ITF and ITCZ “indiscriminately and interchangeably.” Notably, Ramage refused to use either term in his book.
Hastenrath (1988) openly called for an abandonment of the “outmoded notions” of the global coincidence of the various parameters. In their widely used climatology textbook, Barry and Chorley (1992) distinguished between an ITCZ over the ocean and an ITF over land. They mentioned that the formation is discontinuous in time and space and not well developed in the doldrums. Notably, the term ITCZ does not appear in two other widely used textbooks, Wallace and Hobbs’s (2006) ,Atmospheric Science and Krishnamurti’s (1979) ,Compendium of Tropical Meteorology. Over Africa, the latter text refers instead to a “wind separation line.” Moreover, detailed studies of the global ITCZ, such as that of Schneider et al. (2014), consider only ocean regions.
In summary, the origins of the ITCZ paradigm for Africa are murky, but it clearly emerged from attempts to parallel midlatitude concepts. However, its development was haphazard. Most of those who argued for and implemented the concept were midlatitude meteorologists. Tropical meteorologists, and especially those who actually worked in Africa, harshly criticized the ITCZ paradigm and its applications, with many suggesting that it was completely wrong. Unfortunately, the use of this paradigm has persisted.
THE ITCZ AND THE CYCLE OF SEASONS OVER AFRICA.
Hubert (1926) may have been the first to identify an entity equivalent to the ITCZ over Africa. He used the term ITF, defining it as the meeting of two air masses, the dry northeast (NE) harmattan and the moist southwest (SW) monsoon. At some point the diagram in Fig. 3 depicting weather zones with respect to the surface ITCZ over West Africa was published, but the origin of the initial model has been attributed to various sources (Trewartha 1961)—Solot (1943), Walker (1957, 1958), and Hamilton et al. (1945)—all of whom served as meteorologists in Africa, as well as A Pilot’s Primer of West African Weather (Knight and Smith 1944). Over West Africa, where the aforementioned meteorologists worked, the diagram more or less correctly depicts the spatial relationship between the ITCZ and the rainfall zones. However, as described later, the ITCZ and the rainfall zones are decoupled. Griffiths (1972) appears to have been the first to associate a latitudinal displacement of these zones with the seasonal cycle over West Africa.
Henderson et al. (1949) may have been the first to apply the ITCZ concept specifically to East Africa, attributing the seasonal cycle of rainfall there to the movement of the ITCZ toward and away from the equator. The association between the ITCZ and rainfall seasonality, as presented by Petterssen (1941), was also extrapolated to Africa in the form of a diagram similar to that in Fig. 4 (Miller 1971; Flohn 1964). This paradigm is used extensively not only in the meteorological literature but also appears in sources related to disciplines as disparate as geology, ecology, paleoclimate, and history.
Eventually, the view for the extreme seasons published by Bjerknes et al. (1933) was expanded to the often-reproduced diagram in Fig. 5, showing the ITCZ in four seasons. The original source appears to be Thompson’s (1965) atlas, The Climate of Africa, but this depiction further appeared in the French doctoral dissertation of Dhonneur (1974), which is also often cited as the source. Over West Africa, where the concept is best developed, various authors have added further details of the link to the precipitation regime. Note that the image of the ITCZ over Africa in Fig. 5 differs markedly from the image based on the Palmén and Newton (1969) streamline maps (Fig. 2). Meanwhile, many other significantly different images of the African ITCZ have appeared in the literature.
As early as the 1950s, the application of the concept to Africa was highly criticized. Crowe (1951) showed how difficult it is to utilize the ITCZ to explain rainfall seasonality at coastal stations of Africa. Tschirhart (1959), a meteorologist working in equatorial Africa, criticized the ambiguity concerning the nature of the ITF and further claimed that the ITCZ and ITF should be considered distinct entities, with the term ITCZ limited to the ocean. Obasi (1976), a Nigerian working in Kenya, stated that the ITCZ plays no role in forecasting in East Africa. Leroux (2001) provides a scathing review of the concept over Africa. He points out the lack of an accepted definition, the subjective analyses used to formulate charts of the ITCZ, the highly divergent images of it that have been published, the quite varied parameters used to define it, and the plethora of labels attached to it. In discussing the surface zone where the NE harmattan encounters the SW monsoon, he patently states that “to call it an ITCZ is completely wrong,” and laments that terms “such as the ITF and ITCZ…are inappropriate but hallowed by use.”
Our picture of the meteorology over West Africa has fortunately changed dramatically, as a result of such milestones as the Global Atmospheric Research Program (GARP) Atlantic Tropical Experiment (GATE) in 1974, the launch of the Tropical Rainfall Measuring Mission (TRMM) satellite in 1997, and the African Monsoon Multidisciplinary Analysis (AMMA) research project and field campaign (Janicot et al. 2008; Redelsperger et al. 2006). The emphasis is now on the West African monsoon, the phases of which are described by Thorncroft et al. (2011). Zhang et al. (2006) and Nicholson (2009) independently described the monsoon in the boreal summer and presented markedly similar structures. The latter is shown in Fig. 6. The point of both representations is that the wind discontinuity, usually termed the ITCZ, represents a secondary cell of vertical motion that is completely independent of the main region of vertical ascent (although they merge in some years). The main ascent and the rainfall maximum lie between the axes of the midtropospheric African easterly jet and the upper-tropospheric tropical easterly jet. Although both the surface wind discontinuity and the zone of maximum rainfall shift latitudinally with the seasons, as well as from year to year, they fluctuate independently, with the latter exhibiting a much greater longitudinal displacement (Grist and Nicholson 2001).
Unfortunately, no such comprehensive paradigm has been published for equatorial Africa. However, several things are noteworthy about the situation in equatorial Africa. First, even the classical picture of Africa’s circulation (Fig. 5) shows no ITCZ over western equatorial Africa. Second, precipitation there is associated primarily with mesoscale convective complexes and they are so intense and frequent that Zipser et al. (2006) designate this region as having the most intense thunderstorms on Earth. Moreover, the dominant factor appears to be orographic forcing over the highlands surrounding the Congo basin (Jackson et al. 2009). Finally, in eastern equatorial Africa a previously neglected factor is the low-level Turkana jet. This appears to be a major factor in the region’s aridity (Nicholson 2016) and appears to have a major influence on rainfall seasonality as well. Nevertheless, the bulk of the literature on the region’s climate and interannual variability refers to seasonality as linked to a twice-annual passage of the ITCZ. The remainder of this article challenges that concept.
ANALYSIS OF THE ITCZ PARADIGM FOR EQUATORIAL AFRICA.
The core of the ITCZ paradigm over Africa is an annual north–south migration, with a twice-annual equatorial passage corresponding to the two equatorial rainy seasons. Accordingly, there is a corresponding north–south migration of the rain belt associated with low-level convergence. This section examines circulation and rainfall, in order to see whether or not this paradigm can explain the seasonal cycle in equatorial latitudes. Here, for the sake of argument, the commonly accepted definition of the ITCZ over Africa is adopted (i.e., the location where the dry northeasterly harmattan meets the moist southerly flow of the monsoon). Some of the analyses presented will consider the entire equatorial sector, but the focus is on the central and western equatorial regions. The meteorology of these regions is more poorly understood than that of the eastern equatorial regions, about which considerable research has been carried out.
Here, the European Centre for Medium-Range Weather Forecasts (ECMWF) interim reanalysis (ERA-Interim) dataset is utilized to examine vector winds, divergence, and vertical motion. The dataset commences in 1979 and runs to the present, providing a spatial resolution of roughly 80 km (0.75° latitude–longitude) and a time step of 6 h (Dee et al. 2011). Each of the variables was also examined using the 40-yr version of the ECMWF Re-Analysis (ERA-40; Uppala et al. 2005) and the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis dataset (Kalnay et al. 1996; Kistler et al. 2001). As the results were completely consistent among the three datasets, only the results based on ERA-Interim are presented. Most analyses presented here examine the four months of January, April, August, and October, with January and August representing the annual extremes and April and October being used to represent the equatorial rainy seasons. A more difficult decision is what level to utilize to represent surface conditions. In Northern Hemisphere sectors the terrain is relatively low and 1,000 hPa is appropriate, but in much of the Southern Hemisphere the surface elevation exceeds 1,000 m. As a compromise, the 925-hPa level is utilized. However, similar results are achieved if 1,000 hPa is considered.
Rainfall is evaluated from an independent gauge dataset compiled by the author. The original dataset is described in Nicholson (1986) and Nicholson et al. (2000). Numerous stations have been added and the records updated (Nicholson et al. 2017). The dataset includes a total of 2,091 stations within the analysis sector (Fig. 7). Rainfall data have been converted to a ½° grid, using a natural-neighbor technique (Watson 1999). Mean rainfall results for April and November, based on the years 1979–2014, are shown in Fig. 8.
Low-level winds and divergence.
As indicated, the common explanation for the seasonal cycle in equatorial Africa is the north–south migration of the ITCZ. Accordingly, it moves between the two hemispheres and transits the equator twice a year, producing the bimodal seasonal cycle in equatorial regions. An examination of low-level winds and divergence indicates that this is not the case.
Figure 9 shows the mean vector winds and divergence at 925 hPa during four months of the year. During August the southwesterly “monsoon” prevails over West Africa up to roughly 20°N. The ITCZ (i.e., the interaction of the monsoon with the northeast harmattan) is clearly evident in the east, but much less so west of ∼10°E, where the low-level circulation around the Saharan heat low (not shown) interrupts the pattern. At 925 hPa a continuous zone of convergence is roughly coincident with the ITCZ. In April, the core of the first equatorial rainy season, the flow over northern Africa shows some similarity to that of August, but with the pattern displaced equatorward by roughly 6°–10° of latitude. The ITCZ is situated at roughly 10°–12°N and is marked by strong convergence. In November, the core of the second equatorial rainy season, the southwesterly monsoon is extremely weak, but southerly flow gives way to the northeasterly harmattan at roughly 8°–10°N. However, the main area of convergence is somewhat farther north, well within the northerly flow, and convergence is weaker than in April. In January, when the ITCZ is generally represented as having traversed equatorial Africa and residing at roughly 20°S (Fig. 2), the wind shift and zone of convergence clearly lie at roughly 6°–10°N. Notably, these locations are strongly in accord with the ITCZ locations indicated in Fig. 5.
Thus, in all cases the ITCZ remains well north of the equator. In each of the four months a contiguous zone of convergence, located close to the ITCZ, extends across the continent. This zone does migrate with the seasons, in accordance with the shifts in the wind regime seen in Fig. 9. However, its migration bears little relationship to that of the rain belt. This is clearly seen from a comparison of Figs. 8 and 9. In both April and November, the ITCZ and associated convergence lie well to the north of the equatorial locations where this feature is assumed to bring rainfall. Moreover, the latitudinal span of the rain belt is roughly 3–4 times greater than the latitudinal span of the low-level convergence.
More interesting is the pattern of divergence in the equatorial locations during April and November. Instead of low-level convergence, divergence prevails over much of the region, especially across the Congo basin (Fig. 9) (see also Jackson et al. 2009). Over eastern Africa, the divergence patterns in both months bear a resemblance to that of January, the heart of the dry season. Yang et al. (2015) similarly noted prevailing low-level divergence over parts of East Africa during the rainy seasons. Much of the divergence is associated with the low-level Turkana jet (Nicholson 2016), seen in Fig. 9 at roughly 0°–5°N and 35°–40°E (Nicholson 2016).
In summary, during the equatorial rainy seasons the low-level convergence associated with the ITCZ lies well to the north of the regions experiencing rainfall at these times. Further, the low-level winds are divergent, on average, over much of the region during these seasons. Thus, the observed patterns of wind further contradict the ITCZ paradigm as an explanation for the seasonal cycle in Africa’s equatorial latitudes.
Rainfall and vertical motion.
A further tenet of the ITCZ paradigm is a progressive north–south shift of the rain belt during the course of the year, following the path of the overhead sun. Figure 10 shows the latitudinal profile of rainfall during the two equatorial rainy seasons along the two transects shown in Fig. 7. Profiles are shown for three months of each season: March–May for the first season and October–December for the second season. These show some degree of north–south displacement, but only one of the four cases truly fits the classic ITCZ paradigm.
Over western equatorial Africa (16°E) there is little latitudinal displacement of the rain belt between March and April. Only between April and May is there a northward shift apparent. The width of the rain belt is relatively constant throughout the season. In the second rainy season, a latitudinal shift is evident between October and November, but the rain belt shows little change in latitude between November and December. The situation is very different in central equatorial Africa, over the Congo basin (25°E). During the first rainy season only a small northward shift is apparent, but the latitudinal extent of the rain belt steadily decreases between March and May. During the second rainy season at both longitudes a latitudinal shift is evident in each month. Only here and in this season is there a strong semblance of the classic scenario of the ITCZ.
The ITCZ paradigm further relates rainfall to ascent produced by the low-level convergence. The patterns of vertical velocity during the two equatorial rainy seasons, as represented by April and November, belie this scenario. Figure 11 depicts vertical velocity via omega (vertical velocity in pressure coordinates) along the two transects at 16° and 25°E, as well as with latitudinal profiles of topography and rainfall. Although shallow areas of ascent around 10°N are associated with the surface positions of the ITCZ, the ascent extends only to the midtroposphere and prevails over less than 10° of latitude. Shallow areas of ascent are also evident in association with topographic peaks. These appear at ∼21°N and ∼13°S along the transect at 16°E. At 25°E, they appear around 14°N in both months and also at ∼12°S in November only.
Within the latitudinal span of the rain belt (very roughly 5°N–10°S in April and 5°N–15°S in November) the situation is quite different. Low-level subsidence prevails in both months and along both transects. These areas of low-level subsidence largely correspond to areas of low-level divergence or at most very weak convergence (Fig. 9). Aloft, above 850 hPa, strong ascent that reaches into the upper troposphere spans the latitudes of the rain belt. It should be noted that the correspondence between vertical motion and rainfall is not a model-generated result, since the rainfall dataset is independent of the reanalysis dataset. The deep column of ascent appears to be decoupled from the shallow areas of ascent associated with the ITCZ around 10°N. The decoupling is particularly clear in November. This further indicates that the ITCZ should not be considered to be the bearer of seasonal rainfall over Africa’s equatorial region.
SUMMARY AND CONCLUSIONS.
The ITCZ over Africa has long been a point of contention. Much of the picture of the ITCZ over equatorial Africa has been derived from vague ideas concerning the circulation over Africa and much of the original work was conjecture. Tropical meteorologists have long suggested that it is not an appropriate paradigm for the seasonal cycle over equatorial Africa, yet the use of this paradigm persists.
The structure of the motion field during the rainy seasons of the boreal spring and autumn was examined over central and western equatorial Africa. In both locations, the structure showed little similarity to the classic ITCZ paradigm, which entails surface convergence leading directly to ascent and hence rainfall. Low-level subsidence underlies much of the region of maximum rainfall. It results from the divergence of the mountain breezes over the surrounding highlands (Jackson et al. 2009). Ascent associated with the tropical rain belt over Africa commences higher up in the atmosphere. Clearly, the latitudinal progression of the equatorial rainy season does not follow that of a surface convergence zone.
The ITCZ paradigm further suggests a progressive northward shift of the rain belt during the course of the first rainy season and a progressive southward movement of the rain belt within the course of the second rainy season. This pattern is apparent only at 25°E and only in the second rainy season. Figure 10 suggests that the changes during the first rainy season might be better described as a progressive contraction of the rain belt, as the northern edge shows relatively little change from month to month.
If the ITCZ paradigm is incorrect, the obvious question is what does produce the seasonal cycle over equatorial Africa, a region of extraordinarily intense thunderstorms and convective activity. That question is beyond the scope of this article, but several potential factors have been identified in other studies. An important one is the storms that develop in the highlands surrounding the Congo basin; these move into the basin at night when katabatic flow prevails (Jackson et al. 2009). The East African highlands, on the eastern rim of the Congo basin, are particularly important in the initiation of mesoscale convective systems that propagate westward and traverse the equatorial latitudes (Hartman 2017, manuscript submitted to Mon. Wea. Rev.). The intense development of these storms over the Congo basin (i.e., central and western equatorial Africa) is favored by high values of equivalent potential temperature and convective available potential energy (CAPE) during the two rainy seasons (Hartman 2017, manuscript submitted to Mon. Wea. Rev.). The motion fields associated with midlevel jet streams also appear to promote ascent (Nicholson and Grist 2003; Jackson et al. 2009).
Yang et al. (2015), looking at the eastern equatorial region, have emphasized such factors as mean static energy, saturation mean static energy, and vertically integrated moisture from the Indian Ocean. Here, in particular, the role of topography in the development of rain-bearing systems must also be considered, as well as large-scale phenomena such as the Madden–Julian oscillation (Berhane and Zaitchik 2014) and the vertical cells over the Indian Ocean (Hastenrath et al. 2011; Nicholson 2015; Nicholson et al. 2017, manuscript submitted to Global Planet. Change).
The early meteorologists who suggested the ITCZ paradigm did not have the sophisticated tools at their disposal that modern researchers have. Satellite imagery, reanalysis datasets, field experiment results, and computer-generated models allow for a much more detailed picture of the atmosphere in the horizontal and the vertical. This led to the revised model of the monsoon over West Africa. These same tools need to be applied to further our understanding of the seasonal cycle over equatorial Africa. Clearly, our current paradigm over the ITCZ and its seasonal migration does not explain the seasonal cycle in this region.
The assistance of Douglas Klotter in analyses and graphics is gratefully acknowledged. Support for this work was provided by three grants from the National Science Foundation: AGS 1158984, 1445605, and 1535439. The author would like to thank the editor, Brian Mapes, and four anonymous reviewers whose suggestions led to a substantially improved manuscript.