Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

3. Precipitation and climatic change in central Sudan

A. H. Parry

Editor's Introduction

Of all the climatological features, precipitation is by far the most important single one in any assessment of the agricultural potential of this region. Figures 3.1 and 3.2 provide a shorthand summary of average climatic conditions at Khartoum and Ed Dueim, which mark respectively the northern and southern limits of the project area. Air temperatures of less than 5 C are unknown and mean monthly minima exceed 15 C, so low temperatures are never a hazard to plant growth. Wilting does however sometimes prove a minor hazard if strong dry northerly winds blow during the hot season, though in practice the cause of wilting is nearly always a lack of ground moisture to compensate for the strains exerted on the plant. Monthly mean daily maxima exceed 40 C for the months of May, June, and July at Khartoum and for April and May at Ed Dueim.

FIG. 3.1. Average climatic conditions at Khartoum

FIG. 3.2. Average climatic conditions at Ed Dueim

However, the problem arises once rainfall and temperature are considered together. The 1921-1950 rainfall normals (a generally "wetter" period) reveal average figures for Khartoum and Ed Dueim of 181 mm and 330 mm respectively with only two months at Khartoum (July and August) and four months at Ed Dueim (June to September inclusive) when average falls exceed 25 mm. Under these circumstances in no month of the Year does precipitation exceed (potential) evapotranspiration (Davies, 1973), and in Koppen's classification the whole region falls into his BWhw category, implying a hot dry climate throughout the year. In the rather "drier" period for 1951 to 1980 the aridity is emphasized.

In the light of comments in the chapters on geomorphology and soils, it is fair to say that if these average relationships between temperature and rainfall could be relied upon, certain areas could easily be identified as having good potential for agricultural development under rain-fed conditions. Unfortunately this is not the case. Although temperatures vary considerably from year to year, it is the variability in rainfall, in time and space and from one year to another, which is the real problem. Even when ten- and five- year running means are calculated, wide variations over time can still be discerned (fig. 3.3). This is the problem with this area and it forms the core of this chapter.

Characteristics of the Precipitation

The Precipitation Record

In the Sudan's 2.5 million sq. km, there are less than 1,000 rain gauges, more than three-quarters of which are found in central Sudan between latitudes 10 and 16N (fig. 3.4).

Even within this area the distribution is very uneven, as the Gezira/White Nile area has nearly half the rain gauges of the Sudan. In view of the high spatial variability of rainfall the present network is quite inadequate to fully assess rainfall variations even in the central Sudan. At many stations the length of record available is short, but Oliver (1965) has noted that by means of overlapping and consolidating recordings from a number of sites within the general area of the Three Towns a fairly long record can be compiled for Khartoum. The quality of the record varies considerably, though the weather stations at Khartoum and El Obeid are designated as first class and are equipped with autographic instruments.

One possible way of compensating for the poor station network and assessing the often complex and highly variable precipitation patterns that occur is by making use of satellite imagery, and some of the results of experiments using this source for precipitation assessment will be refferred to in a later section.

The Precipitation Regime

Because a large number of readily accessible accounts of Sudanese rainfall and its causes are available (notably El Tom, 1975; Osman and Hastenrath, 1969), it is proposed in this account to present only an outline of the precipitation regime in the project area, drawing attention to its major characteristics and reviewing what is currently known about the causes of rainfall.

It cannot be too highly stressed at the outset that rainfall, and the resulting run off and evaporation, are among the most important climatic features of the central Sudan.

FIG. 3.3. Rainfall variations in the White Nile

This is because under the prevailing arid, tropical climate, for many purposes and in most years, annual rainfall imposes a severe constraint on the way of life of the inhabitants of the area. Although care will be taken in subsequent sections of this report to point out the artificial and often misleading nature of mean values in the context of the Sudan, a brief look at the distribution of average annual rainfall (fig. 3.5) reveals that precipitation decreases from south to north and isohyets would run from west to east in a subparallel fashion. In the project area rainfall averages for 1921-1950 varied from 181 mm at Khartoum to 330 mm at Ed Dueim. During the period 1951-1980 the variation was from 161 mm to 262 mm. As figure 3.5 shows, rainfall increases further south, but the 1921-1950 figure continues to exceed the 1951-1980 figure to at least as far south as Kosti.

A distinctive precipitation regime is a feature of the climate of central Sudan. The rainy season is of variable length but usually occurs between June and September. It occurs as a result of the northward movement of the inter" tropical convergence zone (ITCZ). This marks the boundary between dry north-east winds and moist south-west winds at the surface, and normally by late August the ITCZ reaches its most northerly position between latitudes 18 and 20N. This movement of the ITCZ is one of the most significant weather controls over the Sudan, representing a major discontinuity in the surface moisture and wind fields. Maximum condensation and precipitation occurs in a zone to the south of the ITCZ in association with an upper tropospheric tropical easterly jet stream. Beneath this jet axis upward motion allows the moist surface air to be lifted to great heights and allows the development of thick cumuliform clouds. The retreat of the rainy season from September onwards is characterized by a southward shift of the ITCZ and the disappearance of the tropical easterly jet in the upper troposphere.

In addition to these seasonal movements of the ITCZ it is now known that periodic diurnal displacements (usually of 1 to 2 degrees of latitude) and forward surges occur with intervening periods of retreat. Thus the exact position of the ITCZ is variable from day to day. The withdrawal takes place more rapidly than the advance, although in some years occasional re-advances of the ITCZ can extend the rainy season into October.

FIG. 3.4. Rain gauges in central Sudan

The heaviest rainfall occurs in a zone to the south of the ITCZ where the depth of moist, conditionally unstable air is at its maximum. Highest maxima occur between 4 to 6 degrees south of the ITCZ. Osman and Hastenrath (1969) have given examples of year to year variations in the circulation which can be related to anomalous wet or dry seasons. Thus in the wet yet 1958 the ITCZ advanced as far as 1950'N, whilst in the abnormally dry July of 1966 it stayed south of 17N. More comprehensive synoptic studies of extreme seasons are necessary to increase knowledge of the possible ranges of behaviour of the circulation.

The term "rainy season" is a relative one, since in most years it will consist of a large number of predominantly dry days with only a small number of heavy but short rainstorms. At least three-fifths of the annual rainfall is associated with thunderstorms, and Barbour (1961) has pointed out that "the wide variation in seasonal rainfall totals is almost entirely due to variations in the number of storms occurring and not to any significant differences in the average amounts of rain yielded by individual storms." At Khartoum there are on average six days per year with 10 mm or more and 19 days with 1 mm or more. The maximum fall noted in one day is 100 mm. Because the total rainfall is received from a relatively small number of rain-days, it tends to be variations in the few falls of moderate or heavy amounts (2040 mm) that are most influential in determining the nature of the annual rainfall record. The Sudan Meteorological Service (1963) notes that rain intensities are of the same order of magnitude throughout the Sudan.

Marked diurnal variation in the incidence of rainfall is a feature of the rainy season (Pedgley, 1969). At Khartoum rain falls most frequently during the hours of darkness, particularly between sunset and midnight. Since rain clouds are formed by the ascent of moist air, it follows that the diurnal incidence of rainfall must be determined by the diurnal variations both of upcurrents and of moisture content. Pedgley suggests that an extensive plume of warm air probably develops daily in midtroposphere over the Ethiopian highlands, streaming downwind across the central Sudan and suppressing daytime convection. Rainfall at El Obeid was found to be well distributed throughout the day, so it seems probable that in the project area diurnal variation may be weaker than that observed at Khartoum. The value of rainfall for crops and natural plant cover is likely to be greater if most rain falls at night, since evaporation rates will be less at this time of day.

A given annual precipitation total may assume different significance and effectiveness according to the precise distribution of the small number of rain-days in the wet season. Amongst possible patterns Oliver (1965) notes: early commencement of rains followed by partial or complete failure; late commencement of rains; extended distribution of small rainfalls throughout the wet season; and well-distributed but infrequent heavy falls. Better knowledge of the characteristics of the rainfall regimes is required to make assessment of agricultural potential more realistic. To this end more information is needed about rainfall behaviour over periods that are shorter than a calendar month (the normal unit of time for which data is available), and in particular for pentades (five-day periods). Unfortunately, there seems to be no correlation between June precipitation and subsequent July to September falls, as arid zone rainfall is too erratic to permit such long-range forecasts. Nevertheless, efforts to find useful precursors of wet-season behaviour need to be actively pursued.

FIG. 3.5. White Nile rainfall, 1921 - 1980

Hammer (1972) has provided some evidence that daily distributional rainfall patterns can provide a means of evaluating precipitation mechanisms. Spatial evidence suggests that storm rainfall is often random, with extensive and heavy precipitation in limited areas. This analysis also revealed the existence of a possible four to five-day cycle in Sudanese rainfall, with a definite spacing of rainfall peaks at this interval. If regularities in the data of this type can be shown to be widespread and reliable they could be of useful forecasting value.

Recently Hulme and Walsh (1984) have begun to investigate the structure of the wet season using the Ed Dueim record. A tentative definition of the wet season was devised as follows. The start of the wet season was taken as the first daily rainfall of at least 10 mm, which is succeeded by at least 10 mm more rain in the following 10 days. The end of the wet season was defined as the last day with at least 10 mm of rain. Breaks in the wet season, so damaging from the agricultural point of view, were defined as any period of at least 15 days' duration within the wet season with no single daily fall of more than 2 mm. Figure 3.6 shows the wet season at Ed Dueim, using these definitions, since 1902. A virtual collapse of the wet season in length, timing, and reliability from about 1960 onwards is quite clear. Changes have also occurred in daily rainfall magnitude frequencies between the periods 1921-1950 and 1951-1980. Rainfall frequency (the number of days with at least 1 mm of rain) has declined from 28 to 20 days per annum, with a steep decline in the frequency of minor 11-24.9 mm) and very heavy (over 50 mm) daily rainfalls.

Moderately heavy falls in the 25.0-49.9 mm category have remained almost unchanged. The marked decline in the frequency of heavy falls over these two periods is shown in figure 3.7. The recurrence interval of a 50 mm daily rainfall was around one year between 1921 and 1950, but had declined to 1.7 years in 1951-1980, whilst the figures for a 70 mm fall have decreased from three years to over five years.

A Daily Rainfall Estimation Model

As most of the rainfall in the central Sudan is of convective origin and derives from showers and thunderstorms, it exhibits an extreme spatial variation which the present rain gauge network cannot monitor adequately. Indeed, it could be argued that no likely or conceivable network would be adequate for the task. In these circumstances the best existing means of monitoring rainfall are radar and satellite imagery. Radar is expensive to install and maintain, but satellite weather coverage is good and available virtually on demand. Accordingly it is appropriate to review briefly the applicability of rainfall estimation models, based upon satellite information, to the project area.

FIG. 3.6.Rainy season at Ed Dueim,1902-1982

Methods of rainfall estimation using satellite data have been reviewed by Barrett (1970) and incorporate two basic principles.
- The design of an equation to yield for a given station over a chosen period of time a rainfall coefficient, from features of the cloud field.
- The transformation of the coefficient into rainfall estimates, based on empirical relationships between coefficient values and rainfall recordings from the station in question.

These two stages constitute the framework of the whole methodology. A limited investigation of the application of the basic model to Khartoum has been attempted by Mustafa (1975). His results showed that the model estimated heavy rainfall quite well, but was less reliable with smaller sporadic rainfalls.

Rainfall, Drought, and Desertification

Rainfall Deficiancies as a Contributing Factor in the Desartification Process

Desertification is a process of desert encroachment and expansion, and although the term has recently come to be used principally in connection with the desert margin of Africa there is abundant and conclusive geological proof of contractions and expansions of dry climates. In order to put recent climatic fluctuations in perspective, it needs to be stressed that in semi-arid areas that are prone to desertification rainfall is highly erratic in both its temporal and spatial distribution, and comes during heavy but brief and unevenly spaced showers during the rainy season. Recurrent droughts are in fact a part of such climates and should not be viewed as unexpected events. It is misleading to view "average" in the context of rainfall as a "normal" quantity with the implication that a departure is "abnormal." In the 25 per cent of the earth's land surface classified as semiarid (average ppt. 250-500 mm and average potential evapotranspiration in excess of 800 mm), "normal" has no predictive value (Glantz and Katz, 1977). The mistaken idea that drought is an unexpected event has often been used to excuse the fact that long-range planning has failed to take rainfall variability into account. The climate is frequently blamed for agricultural failures and made a scapegoat for faulty population and agricultural policies.

FIG. 3.7. Changes in the magnitude and frequency of heavy daily rainfalls at Ed Dueim

Rapp ( 1974) has suggested that a common mechanism of desertification is the following: - Expansion and intensification of land use during wet years. These actions include increased grazing, ploughing, and cultivation of new land together with increased wood collection. - Wind erosion during the next dry year, or water erosion during the next heavy rainstorms. The physical climate, bioclimate, and soil climate all change as vegetation is degraded or removed and as the soil is blown or washed away. More research is needed into the relationship between climate and the desertification pro cess, but it is already clear that feedback processes between the mechanisms described above and the largescale atmospheric motion system are quite possible, particularly as albedo and direct solar heating at the surface will be altered with vegetation changes.

It is clear that these processes ought to be incorporated into a system of simulation models. A model of the desertification process should include:

- the surface physical climate, including the hydrological regime;
- a reasonable representation of the surface vegetation cover and animal life;
- land-use practices, both pastoral and arable, including the use of fire;
- the resulting geomorphic disturbance.

Unfortunately no comprehensive model of the process yet exists, though attempts are being made in this direction in other parts of the world. In Australia first results suggest a model of the type described could be related to an atmospheric model of the general circulation type. Not until such modelling is far advanced will it be possible to say to what degree unwise human behaviour and technology or changes of climate over varying time scales are the prime cause of the desertification process. In the meantime the main climatological contribution to desertifi cation studies should be a realistic statistical appraisal of drought incidence and of prolonged excessive or unusually favourable weather. It is obvious that governments, economists, herdsman, and cultivators alike have dangerously short weather memories. It is crucial that decision-makers supplement such terms as "the mean" with more informative statistical measures to characterize adequately the variability of the climate. Before realistic advice can be given about the future climate, the historic record must be studied.

A number of meteorological measures have been suggested by Hare et al. (1977) to mitigate the effects of climatic variability. These include:
- making maximum use of existing climatic records to ensure that the element of surprise and strategic vulnerability is minimized;
- searches for seasonal and inter-annual predictability using sophisticated time-series analysis techniques;
- attempts in conjunction with agriculturists to recognize some concept of sustained yield for different cultural and environmental circumstances.

Precipitation Variability and Drought

An important distinction should be drawn between aridity and drought. Although both are characterized by a lack of water, aridity carries the connotation of a more or less permanent climate condition, while drought is definable as a temporary condition. Drought is often widespread over a large area simultaneously, and in an area like the Sudan it is a recurrent problem. By its very nature it is difficult to define because clearly the definition must vary for different purposes; thus, meteorological, hydrological, and agricultural drought have quite different thresholds. Even such general definitions as "severe and prolonged water shortage" are inadequate as they require the further definition of "shortage." Water needs depend on the type and number of animal and plant communities using the water and thus the concept of drought cannot be divorced entirely from the use to which water is put. However, because water availability depends largely on rainfall, this is often used as the best single index of drought. If the pattern of rainfall occurrence were known, it would be possible to adjust water use so that drought would never occur.

TABLE 3.1. Rainfall extremes: Khartoum 1905-1974 (mm)

  Mar. Apr. May June July Aug. Sept. Oct. Year
Lowest on record 0 0 0 0 2 7 0 0 45
Highest on record 14 15 35 337 204 263 104 28 414
1st decile 0 0 0 0 3 12 1 0 76
9th decile 0 0 8 62 112 171 52 12 332

Because rainfall totals in the project area are not normally distributed, common statistical measures such as the arithmetic mean are poor indicators of rainfall occurrence. Rainfall occurrence is best described by quoting the limits of a certain proportion of the occurrences. One method of describing the distribution is to state the limits of each ten per cent (or decile) of the distribution. Thus the first decile is that rainfall amount which is not exceeded by the lowest 10 per cent of totals, the second decile is the amount not exceeded by 20 per cent of totals, and so on. The fifth decile is the median. By stating the values of the nine deciles it is possible to have a reasonably complete picture of a particular rainfall distribution, and a useful indication of departure from the average can be obtained. The use of such a methodology, based on the largest possible runs of precipitation data, would provide extremely useful information for planning purposes. Table 3.1 illustrates its applicability in the context of the Khartoum record.

Gross variations in rainfall over short distances is related to the patchy nature of semi-arid rainfall, which arises from individual storm occurrences, often over very small areas. Area-averaged figures have little value, as rainfall statistics for any station are representative of that station only Illustration of this point can be seen by considering a northsouth transect down the White Nile in 1967, when the following sequence occurred: Khartoum, very wet; Jebel Aulia, dry; El Geteina, wet; Shabasha, dry. This variation can be still further exacerbated by an apparent tendency for a number of dry (or wet) years to follow each other (fig. 3.8).

Precipitation Decline and Global Climatic Change

Several research workers have attempted to relate changes in rainfall in the Sahel-Sudan area to changes in the general circulation, but the results are not particularly convincing. Winstanley (1973) attributed rainfall decline to an expansion of the circumpolar vortex and a corresponding shift of the climatic zones in the Northern Hemisphere towards the equator. This suggestion is strongly contested by Miles and Folland (1974), who can find no evidence of such a shift and indeed find that the subtropical high over the North Atlantic has rather moved slightly polewards. More recently Kidson (1977) has shown that the downward trend in rainfall at 15 N (the latitude of the project area) since the early 1950s is not present at other latitudes and he attributes this to a large scale perturbation in the upper air circulation over Africa. Low rainfall is found to be associated with the virtual disappearance of the 850 mb trough near 8 N and a weaker easterly jet at 200 mb. These changes appear to be part of a global trend towards a weaker circulation in the middle and upper troposphere in the Northern Hemisphere.

FIG. 3.8. Annual rainfall at Ed Dueim and Geteina, 1946-1979

Some climatologists have suggested that desertification is a visible component of climatic change, but whether drought is the result of climatic change must remain uncertain unless the magnitude (or sequence) of the occurrence of rainfall deficiencies is such as to lie outside the possibility of expectation based on previous climatic records. As climatic change is likely to be a gradual and insidious process, its detection will be extremely difficult, particularly if, as is the case with rainfall, the natural variability of the element is considerable. A 30-year "normal" can be dangerously misleading because differences between successive 30year "normals" are likely to be no more than sampling defects.

Movement of the ITCZ is the process by which rain is brought to the project area in the White Nile, and so anomalies in its movement might be expected to be a cause of drought years. Very recent work (Nicholson and Chervin, 1983) suggests that although an anomalous displacement of the ITCZ is probably associated with wetter years, it is unlikely to be associated with drought years. The rainy season appears to be of its usual length and onset, but less intense during drought years. There is also evidence of a markedly synchronous occurrence of droughts in the African subtropics in both the Northern and Southern Hemispheres, and this perhaps suggests that major changes in the Hadley circulation produce African droughts.

Recent Rainfall Trends

Deficiencies in the number, quality, and length of rainfall records make it extremely difficult to construct an accurate chronology of precipitation behaviour in the project area as a whole. Nevertheless, it is possible to make some obser vations about rainfall behaviour during this century.

In the first half of the century, and especially in the first decade, the climate of the Sudan was probably drier than it had been during the second half of the nineteenth century. The single driest five-year period in the Khartoum record so far this century was centred on 1912 (101 mm).

Within this dry period Grove (1973) has pointed out that 1913 stands out as a particularly dry year, with the isohyets 150 km or more south of their mean positions; the Nile flood was the poorest for which authentic modern records exist. The available data suggests that annual rainfall totals below 100 mm must have been general in the project area at this time. From 1920 (fig. 3.3) the climate became noticeably wetter, with more reliable rainfall from year to year and several sequences of years when rainfall exceeded the recent 30year normals. At Khartoum the five year period centred on 1922, with 289.8 mm, was the wettest in the record, but at Ed Dueim the wettest period occurred later, with a mean of 428 mm in the fiveyear period centred on 1938. In 27 out of 39 years between 1920 and 1956 general precipitation totals were above the long period normals. Undoubtedly, agricultural practices and planning decisions that were made during this favourably wet period need to be reviewed in the light of precipitation behaviour in the last two decades.

Numerous papers have dealt at length with the drought years which have plagued many parts of the African Sahel and which seem to have been particularly marked between 1968 and 1973. Detailed analysis of the records suggests that since about 1960 a generally dry episode has prevailed, setting in progressively from north to south [Nicholson, 1980). At Ed Dueim the lowest five-year mean period was centred on 1975 (156.9 mm), while the 1966-1982 mean of 210 mm was only 37 per cent of the 1921-1950 mean. The plot of five-year running means of rainfall at this station and also at Geteina (fig. 3.3) show clearly the general decline of precipitation since the mid 1930s and the acceleration of this decline during the 1970s. Expressed as deviations of annual rainfall from the 1931-1960 period mean average (fig. 3.8), the long run of dry years is striking, especially at Ed Dueim. Both diagrams reinforce the point that at all locations in the study area rainfall exhibits a typical noise pattern of alternating ups and downs. Despite this restlessness there are conspicuous intervals of shorter or longer duration when the departure from average seems to be one-sided. Such a persistent process is capable of producing non cyclic fluctuations of rainfall which in short records may appear as spurious trends or pseudo cycles. Hence great care is needed when searching the long period records for any form of cycle, and healthy scepticism should prevail at any claim to have recognized such a cycle.

Considering the abundance of evidence that exists con cerning the prevailing dry regime of recent years, 1978 comes as a surprise. At both Ed Dueim and Geteina totals for that year were the highest recorded this century. The total of 532 mm at Ed Dueim exceeded the previous highest annual total by almost 60 mm and serves as a good illustration of the highly capricious nature of Sudanese rainfall in general. It is still too early to say for certain whether 1978 represented an isolated anomalous year from the general trend of lower rainfall or whether it marks a return to the more plentiful rains which seem to have been the rule in the 19201950 period.

If, as Nicholson (1980) claims, there exists a marked coherence of rainfall variation in Africa south of the Sahara, then the recent findings of Faure and Gac (1981) could be important. They believe that annual river discharge data indicates that a series of short periods of drastic rainfall minima are followed by a slow return to longer periods of relatively wet conditions. If this is applicable in the Sudan then by extrapolation the present drought should come to an end in about 1985, with a full reestablishmment of normal rainfall by 1992.

Climate and Rural Change (H. R. J. Davies)

In the White Nile man's room for manoeuvre in the agricultural sector under rain-fed conditions is clearly circumscribed. The basic food crops most likely to succeed are sorghum and millet. The system of cultivation is one that works and has been derived by trial and error over generations of cultivators. Inevitably, therefore, the farmer is reluctant to give up a system that he feels he can rely on, as the margin for error for any change is small. This means that any real change will require something "revolutionary," such as irrigation. "Revolutionary" change, unless one is desperate, is more difficult to come to terms with than "evolutionary" change. The recent period of drought, coming after a rather wetter period, means that on the one hand cultivators will want to hang on desperately to what they know, but, on the other hand, if the situation becomes too bad they may be more willing to change than before.

It is clear that during the recent past both wetter years and a rising population have exerted new pressures on these arid marginal lands. It is clear that a withdrawal of cultivation from the desert margins, together with a more reliable and intensive form of agriculture, is required unless alternative livelihoods can be provided. This points towards an expansion of irrigation. However, the dilemma is plain. A larger population in drought years needs more land to provide for itself; this encourages a desertward expansion, and people will cling to their old and tried systems even more than before. The question remains as to whether the crisis has been sufficient to convince people that a different way of life is desirable. Evidence in later chapters will suggest that there are other factors for change which are reinforcing the climatic need for life-style adjustment in this area of the White Nile.


Barbour, K. M. 1961. "The Diversity of Sudanese Climate." The Republic of the Sudan. University of London Press, London.

Barrett, E. C. 1970. Climetology from Satellites. Methuen, London.

Davies, H. R. J. 1973. Tropical Africa: An At/as for Rural Development. University of Wales Press, Cardiff.

Faure, H., and J. V. Gac.1981. "Will the Sahelian Drought End in 1985?" Nature, 291: 475-478.

Glantz, M H., and R. W. Katz. 1977. "When is a Drought a Drought? " Nature, 267: 192- 194.

Grove, A. T. 1973. "A Note on the Remarkably Low Rainfall of the Sudan Zone in 1913."Savanna, 2: 133-138.

Hammer, R. N. 1972. "Rainfall Patterns in the Sudan." Trop. Geog., 35: 40-50.

Hare, F. K., et al. 1977."The Making of Deserts, Climate, Ecology and Society." Econ. Geog., 53: 332-346.

Hulme, M., and R. P. D. Walsh. 1984. "Hydrological Consequences of Recent Climatic Change in the West central Sudan and Some Suggestions for Future Monitoring and Assessment." In Procedings of Workshop on Monitoring and Controlling Desertification in the Sudan. Clark University, Worcester, Mess.

Kidson, J. W 1977."African Rainfall and Its Relation to the Upper Air Circulation." O. J. B. Met. Soc., 103: 441-456.

Miles, N. K., and C. F Folland 1974 "Changes in the Latitude of the Climate Zones in the Northern Hemisphere." Nature, 252: 616617.

Mustafa, E. N. 1975. "Estimating Daily Rainfall from Satellite Data in the Central Sudan." M.Sc. thesis. University of Birmingham',B irmingham.

Nicholson, S. E 1980. "The Nature of Rainfall Fluctuations in Subtropical West Africa," Man. Weath. Rev., 108: 473-487.

Nicholson, S. E., and R. M. Chervin. 1983. "Recent Rainfall Fluctuations in African Interhemisphere Teleconnections." In A. Street Perrot and M. Beren, Variations in the Global Water Budget.

Oliver, J. 1965. "The Climate of Khartoum Province." In Guide to the Natural History of Khartoum Province, Part 11. Sudan Notes and Records, 46.

Osman and Hastenrath.1969. "On the Synoptic Climatology of Summer Rainfall over Central Sudan." Arch. Met Geophys. Siaklinn,Series B, 17: 297-324.

Pedgley, D. E. 1969. "Diurnal Variation of the Incidence of Mon. soon Rainfall over the Sudan." Met. Mag., 98: 97-134.

Rapp, A., et al. 1974. "Can Desert Encroachment Be Stopped?" Ecolog. Bulletin, 24.

Street-Perrot, A., and M. Beren, eds. 1983. Variations in the Global Water Budget D. Reidel, Dordrecht.

Sudan Metereological Service. 1963. "Meteorology of Sudan." Memoir no,6. Khartoum.

Tanaka, M., et al. 1975. "Recent African Rainfall Patterns." Nature, 255: 201-203.

Thompson, J. D. 1978. "Ocean Deserts and Ocean Oasis." Climatic Change, 1: 205-230.

El Tom, M. A. 1975. The Rains of the Sudan. University of Khartoum Press, Khartoum.

Winstanley, D. 1973. "Rainfall Patterns and General Atmospheric Circulation." Nature, 245: 190- 194.

4. A note on vegetation

I.O. Alam el Din

In 1961 the Sudan Survey Department produced its vegetation map of the Sudan. The project area was almost entirely classified as semi-desert, apart from the area around Ed Dueim, which was classified as woodland savanna with low rainfall. The semi-desert region from north to south was divided into an Acacia tortilis-Maerua crassifolia desert scrub north of Wad Nimr and an area of Acacia mellifera Commiphora desert scrub in the more sandy areas to the west. South of Wad Nimr, almost as far as Ed Dueim, was classified as semi-desert grassland on sand to the west (i.e. largely perennial grasses) and on clay (i.e. largely annual grasses) along the river. A similar division between sands and clays was noticed for the woodland savanna with low rainfall, where sands were characterized by hashab (Acacia Senegal) and clays by thickets of Acacia mellifera with alternating grassy areas.

Inevitably a general map of the Sudan compiled in the 1950s must contain generalizations of often doubtful validity, but one cannot attribute to this alone the general impression one has on visiting this area-namely, that these boundaries, especially the woodland savanna one, are rather optimistic. Contemporary vegetation, it seems, is more degraded than these statements suggest. Furthermore, Smith, writing at about the same time (1950), stated that on datum soils the critical line for Acacia Senegal was 350 mm of annual average rainfall on clays and 250 mm on sands. Harrison (1955) gives 400 mm annual average rainfall as a rough guide for the desertward limit of woodland savanna. Bearing these figures in mind, figure 3.5 suggests that our region is too dry for Acacia Senegal to flourish today, and it is markedly absent except in areas of concentrated groundwater, usually near wadis and khors. The northernmost limit of woodland savanna today must be placed at Kosti.

Another factor is man's increased activity in the area; cultivation and grazing both destroy the vegetation, the first by clearance and the second by selective grazing. Harrison (1955) refers to the effects of overgrazing in the Butana, which included the eating out of various high-quality grazing species such as Blepharis. Reports from White Nile today suggest that in most areas Blepharis either has disappeared or is about to disappear through overgrazing.

Climate, soils, topography, groundwater, and man's activities all contribute to contemporary vegetation patterns. The following notes provide sketches of the vegetation under various contemporary conditions at latitude 14 30'N and from the White Nile westwards to Esh Shuqeiq.

Along the river, around the islands, and close to the banks, considerable quantities of floating vegetation originating in the Sudd and associated marshes are to be found. The most important component today is water hyacinth (Eichhornia crassipes). This vegetation drifts around the Jebel Aulia reservoir following the wind direction, and mats of it become stranded. Water hyacinth in particular constitutes a serious hazard for pump schemes and other irrigation works throughout the project area (Davies, 1959), though at present very little has managed to pass north of the Jebel Aulia Dam itself.

The mud flats along the edge of the river are largely given over to gerf cultivation, but other parts support annual grasses after the river has fallen and provide useful dryseason grazing. Other silty clay areas near the river and among the palaeo-channels of the White Nile have been utilized for irrigation schemes. In some areas, which are periodically flooded to some extent (maiyas), sunt forest Acacia nilotica) is to be found, whilst in other ungrazed areas a thorny scrub of various acacias may be seen.

Some areas near the river are slightly higher than the existing clays, and these include a number of small qoz areas such as Qoz Umm Sufi. These sites are favoured for settlements, and thus their vegetation is often seriously degraded to a few odd patches of tufts, though potentially they are capable of supporting dense thickets. The most common vegetation today includes low bushes of marakh (Leptadenia spartium), senna mekka (Cassia acutifolia), and mesquite.

The rainland to the west is characterized by a thin scatter of bushes interspersed with cultivation and patches of annual grasses. The bushes include senna mekka, tundub (Capparis decidua), occasional stunted talh (Acacia seyal), and heglig (Balanites aegyptiaca). Near the wadis the vegetation improves in density.

Further west the qoz areas are encountered. These are intensively cultivated and grazed, and so again are denuded of vegetation. The grasses tend to be tufted, perennial, and deep-rooted, whereas those on the clays are annuals which form mats. Marakh with scattered hashab and talh are also found. Where wadis, such as Idd Umm Qantar near Shuqeiq, pass through the qoz, the vegetation improves markedly in density and variety, with the haraz (Acacia albida) in particular making its appearance.

The inselbergs, such as Jebel Arashkol, tend to be bare of vegetation apart from a few thorny bushes and a sparse scatter of annual grasses. Near their foot and along the drainage lines the vegetation improves dramatically unless a village is nearby. All villages in this area are characterized by a desert perimeter of bare earth with no vegetation left at all, unless it be the useless Sodom apple (Solanum sodomoum), which not even the goat can use.

This depressing picture of the vegetation is essentially a dryseason one. In years of good rainfall a green grass carpet soon appears with a display of small wild desert flowers. Unfortunately, however, occasional heavy rain storms, particularly if they follow a previous poor rainy season, have a deleterious effect on the soil, washing large quantities of it away and so making it progressively more difficult for the vegetation of regenerate itself. Man, with his cultivation and his animals, makes things worse: he lays large areas bare. The soil is blown or washed away or buried under fine sands. The problem with these lands is that vegetation is being denuded at a faster rate than it can possibly regenerate itself. The result is that a given number of animals requires a progressively larger area for grazing and a downward spiral in the vegetation cover is entered upon.


Davies, H. R. J. 1859. ".The Effects of Water Hyacinth (Eichhornis crasfipes) in the Nile Valley." Nature, 184: 1085-1086.

Harrison, M. N. 1955. "A Report on a Grazing Survey of the Sudan." Ministry of Animal Resources, Khartoum.

Smith, J. 1950. "Distribution of Tree Species in the Sudan in Relation to Rainfall and Soil Texture." Ministry of Agriculture Bulletin 4. Ministry of Agriculture, Khartoum.

Contents - Previous - Next