Nile Basin Climates
Pierre Camberlin
To cite this version:
Pierre Camberlin. Nile Basin Climates. Dumont, Henri J. The Nile : Origin, Environments,
Limnology and Human Use, Springer, pp.307-333, 2009, Monographiae Biologicae. <hal00391068>
HAL Id: hal-00391068
https://hal.archives-ouvertes.fr/hal-00391068
Submitted on 3 Dec 2009
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Camberlin P., 2009 : Nile Basin Climates. In “The Nile : Origin, Environments, Limnology and
Human Use”, Dumont, Henri J. (Ed.), Monographiae Biologicae, Springer, 307-333.
NILE BASIN CLIMATES
Pierre Camberlin
Centre de Recherches de Climatologie – UMR 5210 CNRS/Université de Bourgogne
6 Bd Gabriel – 21000 Dijon – France
Tel : (33) 3 80 39 38 21
Fax : (33) 3 80 39 57 41
e-mail : [email protected]
Abstract
The climate of the Nile Basin is characterised by a strong latitudinal wetness gradient.
Whereas the areas north of 18°N remain dry most of the year, to the south there is a gradual
increase of monsoon precipitation amounts. Rainfall regimes can be divided into 9 types,
among which summer peak regimes dominate. In the southern half of the basin, mesoscale
circulation features and associated contrasts in local precipitation patterns develop as a result
of a complex interplay involving topography, lakes and swamps. Precipitation changes and
variability show up as 3 distinct modes of variability. Drying trends since the 1950s are found
in central Sudan and to some extent the Ethiopian Highlands. The equatorial lakes region is
characterised by occasional very wet years (e.g. 1961, 1997). The interannual variations are
strongly, but indirectly influenced by El-Nino / Southern Oscillation. Sea surface temperature
variations over other ocean basins, especially the Indian and South Atlantic Oceans, also play
a significant role. Projections for the late twenty-first century show a 2-4°C temperature
increase over the basin, depending on the scenario, but rainfall projections are more uncertain.
Most models tend to predict a rainfall increase in the equatorial regions, but there is little
consistency between models over the tropical regions.
1. Introduction
It is well known that the river Nile has the world’s longest stretch under arid conditions: along
3000 km of its course, rainfall does not exceed 150 mm annually. However, due to its great
latitudinal and altitudinal extent, the Nile basin displays large variations in precipitation
receipt. These contrasts, which clearly show up in the mean climate fields, also manifest in
time as large year-to-year or longer-term fluctuations.
2. Mean climate
2.1 Drivers of the Nile basin climates
•
General circulation and its forcings
The Nile basin extends over 35 degrees of latitude, from the equatorial zone (4°S) to the
northern subtropics (31°N). This results in highly contrasted climatic conditions, dominated
by the Hadley circulation. The Hadley circulation is fuelled by a north-south energy gradient
between a zone with excess energy (to the south of the basin, shifting to the central part with
the northern summer heating) and a zone with a deficit (to the north of the basin). The excess
energy originates from high solar radiation gains, low terrestrial radiation losses due to an
1
extensive cloud cover, and a high atmospheric moisture content (latent heat). The deficit in
the north, mainly the Sahara desert, is related to lesser solar radiation gains (in the northern
winter), to high terrestrial radiation losses due to cloudless skies, and to a dry atmosphere
(low latent heat content).
The excess surface energy induces relative low pressure and rising motion, as well as a lowlevel wind convergence, the Intertropical Convergence Zone (ITCZ). The ITCZ is located
southward outside the Nile basin in the northern winter, and gradually shifts to the north to
reach the central part of the basin by mid-summer (fig.1). Rising motion within the ITCZ
results into widespread precipitation.
Reciprocally, the energy loss at higher latitudes is accompanied by diverging low-level winds,
subsidence and permanent dryness. The northern part of the Nile basin is therefore capped by
high-pressure systems, namely the Libyan and Arabian Highs, during much of the year (fig.
1). In summer they weaken and are partly replaced by surface heat lows, though subsidence
still prevails higher in the troposphere.
Winds systems reflect the dominant influence of the Hadley circulation. The Libyan high
drives north-westerly winds in Egypt, turning to north-easterly in northern Sudan, during most
of the year (fig. 1). Occasionally, in winter, the northernmost part of the region along the
Mediterranean is affected by disturbances associated with upper troughs in the mid-latitude
westerly circulation. To the south, the dry north-easterly trade winds combine with those
originating from the Indian Ocean to reach southern Sudan and Uganda. From March, the
ITCZ starts shifting northward, and south-westerly monsoon winds, originating from the
South Atlantic and the Congo basin, appear in southern Sudan. The lake Victoria basin
dominantly remains under the influence of easterlies, gradually getting a southerly component
from April onwards. Further north in Egypt, rising temperatures induce surface desert
depressions, which drive extremely hot and dusty southerly khamsin winds across Egypt,
causing sudden heat waves in spring (e.g, 41°C in Cairo in early April 2003). By June, the
monsoon winds reach central Sudan (16°N, fig.2) and the western slopes of the Ethiopian
Highlands. From the end of June, due to the deepening of the Indian monsoon low further
east, the south-westerly monsoon spills over the Ethiopian Highlands to reach the southern
Red Sea, where it is channelled until it joins the main Indian monsoon flow in the Arabian
Sea. However, a distinct feature of the ITCZ in tropical North Africa, including the Nile
valley, is that in summer the surface southerly winds tend to penetrate far to the north due to
the Saharan heat low. Mean July temperatures range between 26 and 33°C in Egypt and
northern Sudan (Shahin, 1985). The surface heat low causes the “Intertropical Front” (ITF),
which separates the monsoon winds from the northerly trade winds, to show a tilt with height
(Hastenrath, 1991). In July-August, at the time of its northernmost location, the ITF is found
around 17-19°N at the surface in the Nile valley (fig.1), and a few hundreds of km further
south at 1500 m. South of lake Victoria, diverging south-easterly winds from the Indian
Ocean induce dry conditions. Starting in September, the retreat of the ITCZ to the south tends
to be faster than its northward shift (Osman & Hastenrath, 1969).
In the upper troposphere, the winds are also characterised by a seasonal reversal. In winter
strong westerlies, forming the subtropical westerly jet (SWJ), are found over much of the Nile
basin (fig. 2). The variations of their latitudinal location and the waves which develop in the
westerly flow affect winter weather conditions in the northern part of the basin. In particular,
cut-off lows in the eastern mediterranean will induce cloudy or possibly rainy conditions over
northern Egypt. Cold fronts sweeping behind Mediterranean depressions result into a
significant temperature drop, and sometimes the instability created by the passage of cold air
2
over the warm desert surface produces widespread dust-storms (Tucker & Pedgley, 1977).
The SWJ persists until May, after which the westerlies are replaced by easterlies (fig.2). They
take the form of a jet (Tropical Easterly Jet, TEJ), which originates from Asia as a result of
the summer monsoon. Maximum winds are found in July-August at 150 hPa near 10-15°N
over Sudan, with velocities decreasing from 25 to 10 m.s-1 from east to west (Hulme &
Tosdevin, 1989; Segele & Lamb, 2005).
•
Regional and local factors
Among the factors which regionally modify the general circulation pattern, the most
important one is topography. The highlands which bound the Nile basin to the east, from
Eritrea to Kenya, restrict the penetration of the easterlies from the Indian Ocean. An exception
is the gap between the Ethiopian Highlands and the Kenya Highlands, where strong easterlies
prevail throughout the year (Turkana Jet; Kinuthia & Asnani, 1982). Mountain ranges also set
up their own circulation and generate their own climate. A strong daytime horizontal flow is
directed towards the heated mass of the Ethiopian Highlands, while it is suggested that the
nearby Sudanese plains exhibit subsidence (Flohn, 1965; Pedgley, 1971).
Some of the Great Lakes of East Africa also tend to develop their own circulation, in the form
of lake breezes induced by the small diurnal temperature variations of the lakes (around 25°C
for lake Victoria) compared to the surrounding areas. This is most evident for lake Victoria,
whose circular geometry encourages daytime breezes diverging from the lake to the warmer
surrounding land, and night-time land breezes converging to the warmer middle part of the
lake (Fraedrich, 1972). These breezes interact with slope circulation, especially to the northeast of the lake which is bordered by the Western Kenya Highlands (Okeyo, 1987). The joint
effect of lake and upslope breezes enhances afternoon convection. The strong updraughts are
responsible for a high frequency of hail around Kericho and Nandi Hills in Kenya (Alusa,
1986). An important aspect is also the interaction with the large-scale circulation (Asnani &
Kinuthia, 1979; Anyah et al., 2006). The latter in the middle and upper troposphere is
predominantly easterly, thus the convective clusters associated with low-level convergence
and instability tend to drift westwards. This explains asymmetries in the rainfall distribution
on either side of the mountain ranges or lakes.
The influence of regional features (topography, waterbodies, land cover) over atmospheric
circulation is best studied by numerical modelling. The overall influence of the East African
Great Lakes was assessed by Bonan (1995) using a general circulation model, but regional
models are needed to adequately resolve the local topographical features. For East Africa, Sun
et al. (1999a-b) assessed the ability of the NCAR RegCM2 model to simulate the region’s
climate. It was subsequently shown that taking into account three-dimensional lake dynamics
is necessary to satisfactorily reproduce climate conditions over lake Victoria itself (Song et
al., 2004). Anyah et al. (2006) then performed sensitivity experiments showing the key role
played by the Kenya Highlands on the diurnal circulation and associated rainfall distribution
over the lake Victoria basin. For the Nile basin as a whole, Mohamed et al. (2005) used the
RACMO regional model to evaluate the components of the water cycle. The study enabled to
confirm that in June-September the low-level moisture advection towards the Ethiopian
Highlands and Bahr-el-Ghazal region is mainly of Atlantic origin, whereas in Uganda and
east of the Bahr-el-Jebel (White Nile) the Indian Ocean is the major moisture source.
However, regional models still have some difficulties in reproducing some aspects of climate
variation. For instance, while moisture convergence and associated rainfall is well simulated
in June-September over the Ethiopian Highlands and Uganda north of lake Victoria in the
3
RACMO model (Mohamed et al., 2005), the March-May rains over East Africa are strongly
underestimated.
2.2 Spatial and temporal distribution of rainfall
•
Mean annual rainfall
The mean annual rainfall for the Nile basin is low (630 mm over the period 1961-1990), but it
is spatially very contrasted. As much as 28% of the basin receives less than 100 mm annually
(fig.3), and part of it experiences hyper-arid conditions. However, a substantial area exhibits
sub-humid conditions (34% between 700-1300 mm), as displayed on the frequency
distribution plot.
Spatially, there is a very gradual decrease of rainfall amounts from south to south in the
central part of the basin (about 100-140 mm per degree of latitude; fig.4). North of about
18°N, from northern Sudan all across Egypt, rainfall is negligible (below 50 mm a year),
except for a small increase along the Mediterranean coast (Alexandria 180 mm). Precipitation
in excess of 1000 mm is restricted to two areas: the equatorial region from south-western
Sudan to most of the lake Victoria basin, and the Ethiopian Highlands. Even in these two
areas, precipitation amounts are contrasted, with maxima around 2100-2300 mm near Gore,
south-western Ethiopia, 2200 mm over the western part of lake Victoria, 2000 mm on the
western slopes of Mt Elgon. Localised values over 2500 mm are found on Mt Rwenzori
(Osmaston & Kaser, 2001), and perhaps over 3000 mm on the western slopes (Leroux, 1983).
Parts of the Western Rift Valley (lake Edward, lake Albert), north-eastern Uganda, and areas
to the west and south of lake Victoria get less than 900 mm (fig.4).
This general distribution reflects the latitudinal movement of the ITCZ, which never reaches
Egypt and northernmost Sudan, stays only briefly in central Sudan but longer further south. In
wet areas, regional and local enhancements of rainfall are found on the west-facing slopes,
mainly as a result of orographic lifting of the moist south-westerlies or westerlies. In the Great
Lakes region in the south, local wind systems associated with the lakes and the complex
topography of the Rift system explain the variable rainfall amounts. Rain shadows are found
over lakes Albert and Edward in the Western Rift, while the maximum over the western part
of lake Victoria results from the nocturnal convergence which takes place over the lake,
generating thunderstorms which subsequently drift westward. The increase of rainfall with
elevation is far from being universal; moisture convergence or orographic uplift (the latter
possibly resulting into mid-slope maxima or windward / leeward rainfall asymmetries) are
actually of much greater importance in both Ethiopia and the Equatorial Highlands.
•
Seasonal rainfall distribution
The seasonal distribution of the rains is illustrated by the typology shown in fig. 5, which is
based on mean monthly rainfall for the period 1961-1990. The classic division of the basin
into three rainfall regimes (arid, tropical and equatorial, from north to south) remains valid,
but a further separation into nine types is useful. The northernmost part of Egypt receives
some winter rains characteristic of the Mediterranean margins (type 1). The rest of the
country, as well as northern Sudan, are totally dry throughout the year (type 2). At around
18°N a summer rainfall peak starts to appear (type 3), but the mean rainfall for the wettest
month (August) remains very low. It increases in types 4 and 5, while the rainy season
becomes longer (for type 5, 4 to 6 months above 50 mm at 10-13°N in central Sudan). In
Southern Sudan (type 6), the rainy season lengthens further, but though the peak is still in
4
August, rainfall tends to level off during summer. This contrasts with type 7, covering much
of western Ethiopia, where the rainfall regime is very similar to type 6, except for a strong
increase in the middle of the rainy season, in July-August. This increase is likely to be related
to the moist monsoon south-westerlies becoming thicker (Segele & Lamb, 2005), and
overflowing the warm surface of the Ethiopian plateau, which results into enhanced
convective instability. Equatorial regimes are found further south, with two peaks (the main
one in April) separated by two drier seasons, in connection with the twice-a-year passage of
the ITCZ over the region (Nicholson, 1996). In much of Uganda and Western Kenya (type 8),
the northern winter is the driest season. After the March-May rains, rainfall remains quite
high for 6 consecutive months till November. Over lake Victoria and its catchment in
Tanzania, Rwanda and Burundi (type 9), it is the northern summer which becomes the driest
season, whereas there is only a relative decrease between the “short rains”, peaking in
November, and the “long rains”, peaking in April.
•
Small-scale temporal organisation of rainfall
The modalities of within-season variations in rainfall depend on the region. Apart from
northern Egypt, the effect of extra-tropical disturbances on rainfall in the Nile basin is limited.
Upper troughs remain generally rainless over Sudan, though over eastern equatorial Africa,
cases of interactions with moist easterlies have been reported, giving enhanced rainfall. In
Sudan during the summer rains, a pronounced periodicity around 4-5 days, except in dry
years, is found in daily precipitation (Hammer, 1972, 1976), and is suggested to be associated
with easterly waves activity. As recently demonstrated by Mekonnen et al. (2006), dynamic
components of easterly waves are weak in the region, compared to West Africa. However, a
2-6 day convective variance as strong as in West Africa suggests an initiation of the waves by
convective activity, on the western side of the mountains of Sudan (Darfur) and Ethiopia (see
also an example in Desbois et al., 1988). Though most storms developed over Ethiopia drift
westward due to the Tropical Easterly Jet of the upper troposphere, many decay before
reaching the Nile plains (Pedgley, 1969). In equatorial East Africa, organised weather systems
are even more difficult to detect. Synoptic disturbances have been identified, but little is still
known about their behaviour and relationship with rainfall, partly because the effects of
topography make them difficult to follow (Nicholson 1996). Emphasis has been given to
incursions of moist westerlies from the Congo basin, to explain wet spells in the region.
However, this is challenged by Anyah et al. (2006) who demonstrated through numerical
modelling that rainfall over the lake Victoria basin is much less sensitive to changes in
moisture entering the region from the west than from the east. It is likely that it is more the
zonal wind convergence associated with westerly wind anomalies propagating from the
Congo basin, rather than the moisture amount, that is instrumental in rainfall variability. At
longer time-scales, the 30-60 day “Madden-Julian” oscillation (MJO), which strongly
modulates convection over the Indo-Pacific region, has also a moderate effect on intraseasonal rainfall variability in Eastern Africa (Pohl & Camberlin, 2006).
Rainfall intensities may be high, though not as extreme as in many parts of the tropics.
Maximum 24-hour precipitation, though an imperfect indicator, shows that highland locations
do not necessarily tend to have the highest intensities, despite their high mean rainfall
amounts (Potts, 1971; Griffiths, 1972). A generalised map of 100-yr 24-hour extreme rainfall
for the lake Victoria catchment area, based on Gumbel-I distribution, provides values
generally comprised between 110 and 150 mm (WMO, 1974). As an illustration of the poor
relationship between mean and extreme rainfall, in Uganda, the return period for a 100 mm
daily fall is 4 years at the dry station of Moroto, and between 10 and 30 years for most wetter
stations across the country, while at the highland station of Kabale (1800 m) it is even as long
5
as 170 years (Potts, 1971). The differences result from the respective parts played by
convective and orographic rainfall, and the lesser precipitable moisture over high ground. In
the northern part of the Nile basin, the computation of extreme rainfall probabilities is
difficult due to rainfall events becoming much less frequent. However, very high intensities
are possible. In the Nile Plains at Khartoum, a 201 mm 24-hour record, resulting into
extensive damages and human deaths, was recorded on 4 August 1988 (Hulme & Trilsbach,
1989), while the record of the previous 90 years was only 88 mm. On shorter time spans, very
high peak rainfall rates may be found throughout the basin.
As in most other continental regions, rainfall generally peaks in the afternoon, especially over
mountain areas (Desbois et al., 1988; Ba & Nicholson, 1998), but quite contrasted patterns are
found across the basin. These diurnal regimes reflect the regional wind systems and their
daytime evolution. The afternoon maximum is sharp over the Ethiopian Highlands and in
northern Uganda and western Kenya (fig.6). Asmara (Eritrea), on the plateau, gets one third
of its rainfall between 1500 and 1800 local time. This maximum shifts to the evening hours
further west, as the main convective cloud clusters move westward in the direction of the
prevailing upper easterlies. An early morning maximum is even found in the central plains of
Sudan. At Kosti 52% of the August rainfall occurrences occur between 0000 and 0900 LT
(Pedgley, 1969). The weakness of the afternoon maximum is suggested to be due to
subsidence, as a result of the daytime circulation which develops between the Nile Plains and
the Ethiopian Highlands (Pedgley, 1971). Further south over the western part of lake Victoria,
the morning peak is even more pronounced (fig.6). It results from the build up of the nighttime convergence in the middle of the lake, the associated thunderstorms then drifting
westward (Asnani & Kinuthia, 1979, Ba & Nicholson, 1998; Anyah et al., 2006).
3. Climate variability and change: present and future
3.1 Space-time patterns of rainfall variability
The region shows relatively coherent areas of rainfall variability. For the period 1951-2000,
44 % of the space-time variance (for all areas receiving at least 100 mm annual rainfall) can
be accounted for by 3 regional modes of variability, as deduced from principal component
analysis (fig.7). Region 1 corresponds to the northern part of the summer rainfall belt (centred
around 15-16°N), from northern Darfur to Kordofan and the Ethiopian-Eritrean border, with a
southward extension along the Nile Valley. It is characterised by heavy rains in the 1950s and
early 1960s, followed by much drier conditions throughout the last three decades of the
twentieth century, except for relatively isolated wet years like 1980, 1988 and 1998-1999.
The second region is located further south, describing the latitudes 7-14°N but not
continuously from Sudan to Ethiopia. Areas best represented are Darfur and the EthiopianSudanese border area. Like for region 1, a marked rainfall decrease is evident, but the shift
from wetter to drier conditions is delayed to the late 1970s, and a partial recovery occurs in
the late 1990s. Region 3 comprises much of the equatorial regions, as well as the southwesternmost part of Ethiopia. It is dominated by a skewed distribution of annual rainfall, with
some occasional very wet years like 1961, followed by 1951, 1963, 1982 or 1997. There is a
tendency for the 1960s to stand out as wetter than average.
This description of space-time rainfall variations leaves out smaller size areas, like the central
part of the Ethiopian Highlands and parts of southern Sudan. This is partly because the timescale retained (annual) combines different seasons and is not fully relevant for those areas
where different climate signals, with a different spatial extension, are consecutively found
during the year. For instance, part of the Ethiopian Highlands displays spring rains which are
6
unrelated to those found at the same time in equatorial regions. Additionally, the main
Ethiopian rains in summer show similarities with both the sudano-sahelian belt further west,
and the equatorial regions, at this time of the year. However, on an annual basis, the
equatorial region is dominated by the interannual variations of the northern autumn “short
rains”. On average the amount of rain during the latter is equal or less than that of the spring
“long rains”. However, the “short rains” display a much larger amplitude in their interannual
variability than the rains falling during the rest of the year (Nicholson, 1996; Black et al.
2003). In addition, it is spatially much more coherent than the “long rains” (Camberlin &
Philippon, 2002). Therefore, the “short rains” dominate the interannual signal in Equatorial
rainfall and lake levels (Bergonzini et al., 2004).
A key feature of the last decades over part of the Nile basin is the downward rainfall trend,
whose spatial pattern is shown in fig.8. Over the period 1951-2000, the linear trend is
particularly pronounced in most of northern Sudan (around 50% precipitation decrease). It is
also highly significant in Darfur (15-50% decrease). Other regions of significant decrease
include central Sudan (10-40%) and south-western Ethiopia (20-35%). The equatorial region
exhibits localised areas of rainfall decrease, like in West Nile (Uganda), but the overall
rainfall is stable during 1951-2000. Small increases are restricted to isolated pockets around
Addis-Ababa and the lake Victoria shorelines.
In Sudan, the rainfall decrease which started in the second half of the 1960s resulted into a
much shorter rainy season in the 1970s and 1980s, particularly due to an earlier end, while all
classes of daily rainfall amounts have seen a lesser occurrence (Hulme, 1988). As in the West
African Sahel, the rainfall recovery in the 1990s is only a very partial one, this decade
remaining drier than the 1950s and early 1960s. In Ethiopia, previous evidence of decreasing
rainfall is mixed (Conway & Hulme, 1993; Conway, 2000; Funk et al., 2003; Seleshi &
Zanke, 2004) : several studies failed to detect any systematic downward trend while others
did. Conway et al. (2004) noted that Addis-Ababa rainfall shows little trends, but this station
was found to be unrepresentative of the rest of the Ethiopian Highlands. In the latter, a drying
trend is actually shown if only the summer monsoon rains are considered, and the 1950s and
1960s included. Several wet years in the 1990s additionally contribute to dampen the trend,
but it remains very significant in south-western Ethiopia. The coherence between Central
Sudan rainfall and that recorded over part of the Ethiopian Highlands (Atbara basin), as
shown by region 1, replicates the correlation found in annual rainfall for the corresponding
sub-basins in Conway and Hulme (1993). Further south, the association between Western
Ethiopia and south-western Sudan precipitation (region 2), corroborates the correlation found
by Sutcliffe & Parks (1999) between the river flows of the Blue Nile and the Jur at Wau
(Sudan).
3.2 Causes of rainfall variability
One of the main causes of climate variability in the tropics, the El-Nino Southern Oscillation
(ENSO), has for long been found to affect the Nile basin, as demonstrated by Nile flows
interannual variations (Bliss, 1925). However, studies on the regional rainfall response to
ENSO did not emerge until the late 1980s. This response is seasonally dependant (fig. 9).
The main ENSO signal is found during the northern summer, at which time a negative
correlation is found with the Nino 3+4 index, depicting lower than normal rainfall in years of
higher sea-surface temperatures (SST) in the eastern equatorial Pacific (i.e., El Nino years).
The summer signal is consistent over most parts of the basin where rains are found during this
season (fig.10). This includes most of Sudan and Ethiopia, but the largest regional
7
correlations (-0.71 for July-September 1951-2000) are found further south in Uganda and
Western Kenya. Significant correlations between ENSO and rainfall have been documented
by Osman & Shamseldin (2002) for Sudan (annual rainfall), by Haile (1990), Beltrando &
Camberlin (1993), Seleshi & Demarée (1995), Seleshi & Zanke (2004), Segele & Lamb
(2005) and Korecha & Barnston (2007) for Ethiopia (main July-September rains as well as
rainy season length), and by Ogallo (1988), Camberlin (1995), Indeje et al. (2000) and
Phillips & McIntyre (2000) for Uganda and Western Kenya (July-September). All conclude to
some summer rainfall reduction in El-Nino years.
However in October-December, the correlation with ENSO switches signs, becoming positive
(fig.9). At this time of the year, precipitation, mostly restricted to the equatorial regions, is
higher than normal in El-Nino years. This pattern is typical of Eastern Equatorial Africa
during the “short rains” of October-December (Ogallo, 1988; Hastenrath et al., 1993;
Nicholson & Kim, 1997; Nicholson & Selato, 2000; Mutai & Ward, 2000; Indeje et al.,
2000). Following Saji et al. (1999) and Behera et al. (2005), the ENSO signal in the East
African “short rains” may only be an indirect manifestation of anomalous Walker-type (eastwest) circulation in the Indian Ocean, associated with zonal SST gradients. Though most of
the years which record abundant “short rains” in Eastern Africa are El-Nino years, like 1965,
1972, 1977, 1982, 1997 and 2006 in the last 4 decades, there exist a few cases of high rainfall
which can solely be explained by anomalous SST and circulation in the Indian Ocean. This is
best shown in the year 1961 (Flohn, 1987; Kapala et al., 1994; Hastenrath & Polzin, 2003),
whose exceptionally heavy rains resulted from a strong warming (cooling) in the western
(eastern) Indian Ocean, in the absence of El-Nino conditions. Thus the October-December
rains exhibit a higher correlation with the zonal SST gradient in the Indian Ocean than with
the Nino 3+4 index (fig.9-10), in agreement with Saji &Yamagata (2003), though in many
cases it is ENSO which originally triggers off the Indian Ocean dipolar events. During other
seasons (including summer), zonal SST gradients in the Indian Ocean have virtually no role in
rainfall variability over the Nile basin.
Though the summer rains in the Nile basin are partly dependant on ENSO, a stronger
connection exists with the Indian Monsoon (Camberlin, 1995, 1997). Averaged over Ethiopia,
Eritrea, Sudan, Uganda and Western Kenya (i.e., roughly the part of the Nile basin receiving
summer rains), July-September rainfall during 1953-1988 is correlated at -0.89 with Bombay
pressure, an indicator of the Indian monsoon intensity. Droughts (heavy rains), and not only
those occurring at times of El-Nino (La-Nina) events, therefore tend to occur simultaneously
in the Nile basin and in India. This is because a deepened monsoon low increases the pressure
gradient with the South Atlantic Ocean, resulting into strengthened moist south-westerlies /
westerlies in Ethiopia, Sudan and Uganda. An additional factor is the strengthened Tropical
Easterly Jet associated with an intense Indian monsoon, known to enhance rainfall in the
sahelian belt (Hulme & Tosdevin, 1989; Fontaine & Janicot, 1992; Grist & Nicholson, 2001).
Both teleconnections with ENSO and the Indian Monsoon fail to account for the downward
rainfall trend experienced in the central part of the basin in the 1970s and 1980s. This trend is
actually similar to that found further west in the sudano-sahelian belt of West Africa. Land
cover changes, associated in particular with overgrazing, and which was earlier thought as
being a possible trigger of the drought (Charney, 1975), are now considered only as an
amplifier of ocean-induced rainfall variations. Inter-hemispheric SST variations are actually
considered as the key player in the long-term rainfall evolution (e.g., Folland et al., 1986;
Ward, 1998; Giannini et al., 2003). In particular, the cooling of the North Atlantic ocean in
the late 1960s, combined with the warming of the Southern Hemisphere oceans (including the
Indian Ocean) is likely to have resulted into a reduced northward excursion of the ITCZ
8
and/or more rainfall over the tropical oceans than over the African continent. A negative
correlation was noted between and Indian Ocean SST and both Sudan rainfall (Osman &
Shamseldin, 2002) and Ethiopia rainfall (Seleshi & Demarée, 1995; Gissila et al., 2004;
Korecha & Barnston, 2007). This prompted these authors to define seasonal prediction
models, including Indian Ocean SST as well as ENSO as predictors. However it should be
noted that the negative correlation with Indian Ocean SST is mainly a manifestation of
common trends, as evidenced for the Sahel as a whole (Giannini et al., 2003). South Atlantic
SST also exhibits a warming trend. A small contribution of this basin to summer rainfall
variability is demonstrated for Sudan and Ethiopia. The South Atlantic warming is not only a
component of that of the southern hemisphere, but it also occasionally has a separate
incidence on monsoon depth, like in 1984 where the devastating drought which affected
Ethiopia and Sudan can be related to exceptionally high SST in the Gulf of Guinea. The main
evaporative regions that contribute to summer rainfall in the eastern Sahel are actually central
Africa and the Gulf of Guinea, and not the Indian Ocean (Druyan & Koster, 1989). The
higher moisture content which prevails in warm years is largely counterbalanced by the
reduced northward excursion / monsoon depth which results from the unfavourable
meridional energy gradients.
Other causes of rainfall variability are not well known. This particularly applies to the
northern spring (March-May) rainy season in the equatorial part of the basin. Its interannual
variability is spatially poorly organised, and the above-mentioned sources of climate
variability account for only a small part of rainfall fluctuations (Camberlin & Philippon,
2002). The zonal gradient between the Atlantic and Indian oceans and the associated
anomalous westerlies play a part, and so do interactions with the upper tropospheric
extratropical circulation. A North Atlantic Oscillation (NAO) signal is found over the
southernmost part of the basin during the December-February dry season (McHugh and
Rogers, 2001), suggesting unseasonable rainfall in negative NAO years, though the
teleconnection mechanisms are not well understood. In Ethiopia, an 11-yr cycle was
previously identified in rainfall and associated with sunspot numbers and solar activity
(Wood, 1977; Seleshi et al., 1994). Evidence of similar oscillations in the equatorial lakes
region is still inconclusive.
3.3 Projected climate changes
The drying trend which affected much of the Nile basin in the 1970s and 1980s, added to
potential climate effects associated with increasing greenhouse gases concentrations in the
global atmosphere, have prompted speculations about the future of the Nile basin climate.
In the 1990s, twenty-first century projected climate changes for the Nile basin were focused
upon in a series of studies dedicated to the assessment of water resources, based on a direct
use of General Circulation Models (GCM) experiments. Strzepek et al. (1996) used 3 GCM
(UKMO, GISS and GFDL) to evaluate future changes in the Nile water resources, under a
doubled CO2 hypothesis. A 15-17% increase in the basin-averaged rainfall was found for the
first two models, and a 5% increase for the latter. Expected temperature rises ranged from 3.1
to 4.7°C. A more detailed assessment, based on interpolated data from 6 GCMs, provided
contrasted projections (Yates & Strzepek, 1998). While for the equatorial lake region almost
all the models gave an increase in annual precipitation (0 to 26%), the projections for the Blue
Nile catchment were between -9% and +55%. Temperature rises were more consistent among
the models, as well as spatially. Inter-model differences in precipitation were also pointed out
by Conway and Hulme (1996), who used the GFDL and GISS models as well as a weighted
ensemble mean from 7 GCM experiments. Whereas all the experiments showed a temperature
9
increase slightly under 1°C by 2025 with respect to 1961-1990, and fairly identical in the Blue
Nile and lake Victoria areas as well as in all the four trimesters considered, precipitation
projections were more contrasted. Increases dominated over lake Victoria, but large
disparities in the Blue Nile region highlighted the difficulty for most models to simulate
precipitation. More recently, McHugh (2005), based on the 4 models out of 19 which best
simulate East African rainfall, showed a projected rainfall increase in the equatorial regions,
mainly due to a rise in DJF and MAM rainfall.
Other studies considered more diverse greenhouse gases emission scenarios (SRES
scenarios), but the outcome is not markedly different from that derived from earlier studies.
Hulme et al. (2001), analysing 10 experiments from 7 coupled models, obtained a median
temperature rise of 0.12-0.18°C per decade under scenario B1 (low emission scenario) to 0.30.6°C per decade under scenario A2 (high emission scenario) for the Nile basin region. In
each case the rise is less in the equatorial region, and higher in northern Sudan. Inter-model
differences are moderate for temperatures, but not for precipitation. This is particularly so for
the summer rains. While a majority of decreases (exceeding natural variability) are projected
for much of eastern Africa, in Ethiopia the trend is negative in some models, and positive in
others (Hulme et al., 2001). In the northern winter (DJF), whatever the scenario, there is a
more robust tendency for increased rainfall in Ethiopia and Uganda, but this is during the low
rainfall season in most parts of the region.
The projections shown in the third assessment report (IPCC, 2001), and confirmed in the 2007
fourth assessment report, give an average temperature increase for the Nile basin region of
around 2°C (B1 scenario) to 4°C (A2 scenario) for the years 2090-2099. Confidence in the
projections remains lower for precipitation. The only two consistent trends projected using
ensemble model experiments correspond to an increase in northern winter (DJF) rainfall in
eastern equatorial Africa, and a decrease in the Eastern Mediterranean (in Egypt, from already
very low rainfall amounts).
The expected temperature rise is to be compared with the actual changes during the twentieth
century. Studying a greater eastern Africa region, from Sudan to Mozambique, King’uyu et
al. (2000) found a general minimum (night-time) temperature warming, especially in the
northern part of the region, which corresponds to the Nile basin. The geographical patterns of
the trends are very complex. However, for East Africa as a whole, a rise of about 0.4°C is
found in the last four decades of the twentieth century (Hulme et al., 2001), though it should
be recalled that the 1960s stand out as an unusually wet and cold decade. A significant rise
(decrease) in hot (cold) extremes is also found in Uganda (New et al., 2006). For Egypt,
decreasing temperature trends were found over the observation period 1941-2000 for annual,
maximum, winter and autumn temperatures, and increasing trends for minimum, winter and
spring temperatures (Domroes & El-Tantawi, 2005). Increasing trends prevailed in summer
(Hasanean & Abdel Basset, 2006). For the recent period 1971-2000 all trends were positive
except maximum temperature (Domroes & El-Tantawi, 2005). In Addis-Ababa in Ethiopia,
Conway et al. (2004) showed increasing trends in annual minimum and maximum
temperatures from 1951 to 2002 (0.4°C/decade and 0.2°C/decade, respectively). On the
whole, while a majority of warming trends are found in the region in the last 50 years, the
patterns are spatially and temporally variable, though becoming increasingly clear in the last
decades. Observed rainfall trends have been discussed above. They are characterised by the
pronounced decrease of the 1970s and 1980s in Sudan and parts of the Ethiopian Highlands,
with a partial recovery since then, and have been shown to result from hemispheric SST
variations.
10
The fact that the region is subject to large natural climate variability makes the observation of
climatic trends of the last few decades a poor indicator of the climate conditions expected for
the Nile basin during the twenty-first century. This can only be assessed through numerical
modelling. However, the problem with GCM projections is that they are made at coarse
resolutions (typically at a scale of 2.5° latitude and longitude) so that, in a region like the Nile
basin which has a complex topography and contrasted land surface properties, the associated
regional circulation features are very poorly resolved. To date, there have been few attempts
to address this problem, though the current development of a regional climate model tuned to
North-East Africa (e.g., Mohamed et al., 2005; Anyah et al., 2006) will be an important step
towards higher resolution climate projections. Alternatives are the use of statistical
downscaling methods, which rely on the definition of empirical models relating present-day
rainfall and temperature variations to large-scale atmospheric patterns. The latter are generally
better simulated than precipitation by the GCMs, hence by using the projections of dynamical
variables under enhanced greenhouse conditions, it might be feasible to estimate rainfall
changes and their spatial distribution. However, this requires a prior understanding of the
factors which determine current rainfall variations in the Nile basin region.
References
Alusa A. L., 1986. Some Synoptic Patterns and Wind fields Associated with Hail Occurrence in Kericho and
Nandi Hills Area of Kenya. IMTR Research Report, Nairobi, Kenya 3/1986, 28pp.
Anyah R. O., F. H. M. Semazzi & Lian Xie, 2006. Simulated physical mechanisms associated with multi-scale
climate variability over lake Victoria Basin in East Africa. Monthly Weather Review: 3588-3609.
Asnani G. C. & J. H. Kinuthia, 1979. Diurnal variation of precipitation in East Africa. KMD Research Report
8/79: 58 pp.
Ba M. B. & S. E. Nicholson, 1998. Analysis of convective activity and its relationship to the rainfall over the
Rift Valley lakes of East Africa during 1983-90 using the METEOSAT infrared channel. Journal of
Applied Meteorology 37: 1250-1264.
Behera S. K., J. J. Luo, S. Masson, T. Yamagata, P. Delecluse, S. Gualdi & A. Navarra, 2005. Paramount Impact
of the Indian Ocean Dipole on the East African Short Rains: A CGCM Study. Journal of Climate 18:
4514-4530.
Beltrando G. & P.Camberlin, 1993. Interannual variability of rainfall in the Eastern Horn of Africa and
indicators of atmospheric circulation. International Journal of Climatology 13: 533-546.
Bergonzini L., Y. Richard, L. Petit & P. Camberlin, 2004. Zonal Circulations over the Indian and Pacific Oceans
and the Level of Lakes Victoria and Tanganyika. International Journal of Climatology 24: 1613-1624
Black E., J. Slingo & K. R. Sperber, 2003. An observational study of the relationship between excessively strong
short rains in Coastal East Africa and Indian Ocean SST. Monthly Weather Review 131: 74-94.
Bliss E. W., 1925. The Nile flood and world weather. Memoirs of the Royal Meteorological Society 4: 53-84.
Bonan G. B., 1995. Sensitivity of a GCM simulation to inclusion of inland water surfaces. Journal of Climate 8:
2691-2704.
Camberlin P., 1995. June-September rainfall in North-Eastern Africa and atmospheric signals over the tropics: a
zonal perspective. International Journal of Climatology 15: 773-783.
Camberlin P., 1997. Rainfall anomalies in the source region of the Nile and their connection with the Indian
summer monsoon. Journal of Climate 10: 1380-1392.
Camberlin P. & N. Philippon, 2002. The east African March-May rainy season : associated atmospheric
dynamics and predictability over the 1968-97 period. Journal of Climate 15: 1002-1019.
Charney J. G., 1975. Dynamics of desert and drought in Sahel. Quarterly Journal of the Royal Meteorological
Society 101: 193-202.
Conway D., 2000. Some aspects of climate variability in the north east Ethiopian highlands - Wollo and Tigray.
Sinet - Ethiopian Journal of Science 23: 139 – 161.
Conway D. & M. Hulme, 1993. Recent fluctuations in precipitation and runoff over the Nile sub-basins and their
impact on main Nile discharge. Climatic Change 25: 127-151.
Conway D., C. Mould & W. Bewket, 2004. Over one century of rainfall and temperature observations in Addis
Ababa, Ethiopia. International Journal of Climatology 24: 77-91
Conway D. & M. Hulme, 1996. The impacts of climate variability and future climate change in the Nile Basin on
water resources in Egypt. International Journal of Water Resources Development 12: 261-280.
11
Desbois M., T. Kayiranga, B. Gnamien, S. Guessous & L. Picon, 1988. Characterisation of some elements of the
Sahelian climate and their interannual variations for July 1983, 1984 and 1985 from the analysis of
METEOSAT ISCCP data. Journal of Climate 9: 867-904.
Domroes M. & A. El-Tantawi, 2005. Recent temporal and spatial temperature changes in Egypt. International
Journal of Climatology 25: 51-63.
Flohn H., 1965. Contributions to the synoptic climatology of North-East Africa. WMO Technical notes, 69, 236244.
Flohn H., 1987. East African rains of 1961/62 and the abrupt change of the White Nile discharge. Palaeoecology
of Africa 18: 3-18.
Folland C. K., T. N. Palmer & D. E. Parker, 1986. Sahel rainfall and worldwide sea temperatures (1901-1985).
Nature 320: 602-607.
Fontaine B. & S. Janicot, 1992. Wind-field coherence and its variations over West Africa. Journal of Climate 5:
512-524.
Fraedrich K., 1972. A simple climatological model of the dynamics and energetics of the nocturnal circulation at
lake Victoria. Quarterly Journal of the Royal Meteorological Society 98: 332-335.
Funk C., A. Asfaw, P. Steffen, G. Senay, J. Rowland & J. Verdin, 2003. Estimating Meher crop production using
rainfall in the 'Long Cycle' region of Ethiopia. FEWS-NET special report, October 2003, 4 pp.
Giannini A., R. Saravanan & P. Chang, 2003. Oceanic forcing of Sahel rainfall on interannual to interdecadal
time scales. Science 302: 1027-1030.
Gissila T., E. Black, D. I. F. Grimes & J. M. Slingo, 2004. Seasonal forecasting of the Ethiopian summer rains.
International Journal of Climatology 24: 1345-1358.
Griffiths J. F., 1972. Eastern Africa; Ethiopian Highlands. In World Survey of Climatology, vol.10, Africa, 313388.
Grist J. P. & S. Nicholson, 2001. A Study of the Dynamic Forces Influencing Rainfall Variability in the West
Africa Sahel. Journal of Climate 14: 1337-1359.
Haile T., 1990. Drought in Ethiopia. In 3rd WMO Symposium on Meteorological aspects of Tropical Droughts,
Niamey, 30 April- 4 May 1990, WMO TMRP Series 36: 273-278.
Hammer R. M., 1972. Rainfall patterns in the Sudan. Journal of Tropical Geography 35: 40-50.
Hammer R. M., 1976. Rainfall characteristics in Eastern Sahel. Nature 263: 48-50.
Hasanean H. M. & H. Abdel Basset, 2006. Variability of summer temperature over Egypt. International Journal
of Climatology 26: 1619-1634.
Hastenrath S., 1991. Climate Dynamics of the Tropics. Kluwer, Dordrecht, Netherlands, 488 pp.
Hastenrath S., A. Nicklis & L. Greischar 1993. Atmospheric-hydrospheric mechanisms of climate anomalies in
the western equatorial Indian Ocean. Journal of Geophysical Research 98: 20 219-20 235.
Hastenrath S. & D. Polzin, 2003. Circulation mechanisms of climate anomalies in the equatorial Indian Ocean.
Meteorologische Zeitschrift 12: 81- 93.
Hulme M., 1988: Changes in wet season structure in Central Sudan, 1900-86. In Recent climatic change, S.
Gregory, Ed., Belhaven Press, London & New-York, 179-192.
Hulme M., R. Doherty, T. Ngara, M. New & D. Lister, 2001: African Climate Change: 1900-2100. Climate
Research 17: 145-168.
Hulme M. & N. Tosdevin, 1989. The tropical easterly jet and Sudan rainfall: a review. Theoretical and Applied
Climatology 39: 179-187.
Hulme M. & A. Trilsbach, 1989. The August 1988 storm over Khartoum: its climatology and impact. Weather
44: 82-90.
Indeje M., F. H. M. Semazzi & L. J. Ogallo, 2000. ENSO signals in East African rainfall seasons. International
Journal of Climatology 20: 19-46.
IPCC, 2001. Climate change 2001. The Scientific Basis. Contribution of Working Group I to the Third
Assessment Report of the Intergovernmental Panel on Climate Change. Ed. by J.T. Houghton, Y. Ding,
D.J. Griggs, M. Noguer, P.J. van der Linden, X. Dai, K. Maskell, C.A. Johnson. Cambridge University
Press, Cambridge, UK, and New York, NY, USA, 881 pp.
Kalnay E. et al., 1996. The NCEP/NCAR 40-year Reanalysis Project. Bulletin of the American Meteorological
Society 77: 437-471.
Kapala A., K. Born & H. Flohn, 1994. Monsoon anomaly or an El-Nino event at the equatorial Indian Ocean ?
Catastrophic rains 1961/62 in East Africa and their teleconnections. Proceedings of the International
Conference on Monsoon Variability and Prediction, Trieste, Italy, 9-13 May 1994. WMO/TD 619: 119126.
King’uyu S. M., L. J. Ogallo & E. K. Anyamba, 2000. Recent trends of minimum and maximum surface
temperatures over Eastern Africa. Journal of Climate 13 (16): 2876-2886.
Kinuthia J. H. & G. C. Asnani, 1982. A newly found jet in North Kenya (Turkana Channel). Monthly Weather
Review 110: 1722-1727.
Korecha D. & A. G. Barnston, 2007. Predictability of June–September Rainfall in Ethiopia. Monthly Weather
Review 135: 628–650.
12
Leroux M., 1983. Le climat de l'Afrique Tropicale. Champion, Paris. 636 pp.
McHugh M. J. & J. C. Rogers, 2001. North Atlantic Oscillation influence on precipitation variability around the
Southeast African Convergence Zone. Journal of Climate 14: 3631-3642.
McHugh M. J., 2005. Multi-model trends in East African rainfall associated with increased CO2. Geophysical
Research Letters 32, L01707, doi:10.1029/2004GL021632
Mekonnen A., C. D. Thorncroft & A. R. Aiyyer, 2006. Analysis of Convection and Its Association with African
Easterly Waves. Journal of Climate 19: 5405-5421
Mohamed Y.A., B. J. J. M. van den Hurk, W. G. M. Bastiaanssen & H. Savenije, 2005. Hydroclimatology of the
Nile: Results from a regional climate model. Hydrology and Earth System Sciences 9: 263-278.
Mutai C. C. & M. N. Ward, 2000. East African rainfall and the tropical circulation / convection on intraseasonal
to interannual timescales. Journal of Climate 13: 3915-3939.
New M. et al., 2006. Evidence of trends in daily climate extremes over southern and west Africa, Journal of
Geophysical Research 111, D14102, doi:10.1029/2005JD006289.
New M., D. Lister, M. Hulme & I. Makin, 2002. A high-resolution data set of surface climate over global land
areas. Climate Research 21: 1-25.
Nicholson S. E., 1996. A review of climate dynamics and climate variability in Eastern Africa. In The
limnology, climatology and paleoclimatology of the east african lakes, Johnson T.C. & Odada E.O., Eds.
Gordon and Breach, 25-56.
Nicholson S. E. & J. Kim, 1997. The relationship of the El Nino-Southern Oscillation to African rainfall.
International Journal of Climatology 17: 117-135.
Nicholson S. E. & J. C. Selato, 2000. The Influence of La Nina on African Rainfall. International Journal of
Climatology 20: 1761-1776.
Nicholson S. & X. Yin, 2002. Mesoscale patterns of rainfall, cloudiness and evaporation over the Great Lakes of
Africa. In The East African Great Lakes: Limnology, Palaeolimnology and Biodiversity. E.O. Odada &
D. O. Olago, eds. Kluwer Academic Publishers, Dordrecht, Netherlands, 93-120
Ogallo L. J., 1988. Relationships between seasonal rainfall in East Africa and the Southern Oscillation. Journal
of Climatology 8: 31-43.
Okeyo A. E., 1987. The influence of lake Victoria on the convective activities over the Kenya highlands. Journal
of the Meteorological Society of Japan 65, special volume, 689-695.
Osman O. E. & S. L. Hastenrath, 1969. On the synoptic climatology of summer rainfall over Central Sudan.
Archiv für Meteorologie, Geophysik und Bioklimatologie, Ser. B, 17: 297-324.
Osman Y. Z. & A. Y. Shamseldin, 2002. Qualitative rainfall prediction models for central and southern Sudan
using El Niño-southern oscillation and Indian Ocean sea surface temperature Indices. International
Journal of Climatology 22 (15): 1861-1878
Osmaston H. & G. Kaser, 2001. Rwenzori Mountains National Park, Uganda. Glaciers and Glaciations. E.
Gärtner & C. Georger. Bartholomew, Glasgow. Folded map.
Pedgley D. E., 1969. Diurnal variation of the incidence of monsoon rainfall over the Sudan. Meteorological
Magazine 98: 97-106 & 129-134.
Pedgley D. E., 1971. Diurnal incidence of rain and thunder at Asmara and Addis-Ababa. Meteorological
Magazine 100: 66-71.
Philipps J. & B. McIntyre, 2000. ENSO and interannual rainfall variability in Uganda : implications for
agricultural management. International Journal of Climatology 20 (2): 171-182.
Pohl B. & P. Camberlin, 2006. Influence of the Madden-Julian Oscillation on East African rainfall. Part I:
Intraseasonal variability and regional dependency. Quarterly Journal of the Royal Meteorological Society
132: 2521-2539
Potts A. S., 1971. Maximum daily point rainfall in Uganda: A sample survey. East African Geographical
Review 9: 25-34.
Saji N. H., B. N. Goswami, P. N. Vinayachandran & T. Yamagata, 1999. A dipole mode in the tropical Indian
Ocean. Nature 401: 360-363.
Saji N. H. & T. Yamagata, 2003. Possible impacts of Indian Ocean Dipole mode events on global climate.
Climate Research 25 (2): 151-169.
Segele Z. T. & P. J. Lamb, 2005. Characterization and variability of Kiremt rainy season over Ethiopia.
Meteorology and Atmospheric Physics 89: 153-180.
Seleshi Y., Demarée G. R., 1995. Rainfall variability in the Ethiopian and Eritrean highlands and its links with
the Southern Oscillation Index. Journal of Biogeography, 22: 945-952.
Seleshi Y., G. R. Demarée & J. W. Delleur, 1994. Sunspot numbers as a possible indicator of annual rainfall at
Addis-Ababa, Ethiopia. International Journal of Climatology 14: 911-923.
Seleshi Y. & U. Zanke, 2004. Recent changes in rainfall and rainy days in Ethiopia. International Journal of
Climatology 24: 973-983.
Shahin M., 1985. Hydrology of the Nile Basin. Elsevier, Amsterdam, 575 pp.
Song Y., F. H. M. Semazzi, L. Xie & L. J. Ogallo, 2004. A coupled regional climate model for the lake Victoria
Basin of East Africa. International Journal of Climatology 24: 57-75.
13
Strzepek K. M., D. N. Yates & D. E. El Quosy, 1996. Vulnerability assessment of water resources in Egypt to
climatic change in the Nile Basin. Climate Research 6: 89–95.
Sun L., F. H. M. Semazzi, F. Giorgi & L. J. Ogallo, 1999a. Application of the NCAR regional climate model to
eastern Africa. Part I : simulations of autumn rains of 1988. Journal of Geophysical Research 104 (D6):
6529-6548.
Sun L., F. H. M. Semazzi, F. Giorgi & L. J. Ogallo, 1999b. Application of the NCAR regional climate model to
eastern Africa. Part II : simulations of interannual variability. Journal of Geophysical Research 104 (D6):
6549-6562.
Sutcliffe J. V. & Y. P. Parks, 1999. The Hydrology of the Nile. International Association of Hydrological
Sciences, Wallingford, Grande-Bretagne. IAHS Special Publication no. 5, 179pp.
Tilahun K., 2006. Analysis of rainfall climate and evapo-transpiration in arid and semi-arid regions of Ethiopia
using data over the last half a century. Journal of Arid Environments 64 (3): 474-487.
Tomsett J. E., 1975. The diurnal variation of precipitation in East Africa. East African Met Dept Technical
Memorandum n° 25, 66 pp.
Tucker M.R. & D. E. Pedgley, 1977. Synoptic disturbances over the Nile Valley north of the Intertropical
Convergence Zone. Tellus 29 (1): 17-24.
Ward M.N., 1998. Diagnosis and short-lead time prediction of summer rainfall in tropical north Africa at
interannual and multidecadal timescales. Journal of Climate 11: 3167-3191.
WMO (World Meteorological Organisation), 1974. Hydrometeorological survey of the catchments of Lakes
Victoria, Kyoga and Albert. RAF/66-025 Technical Rep. 1. WMO, Geneva.
Wood C. A., 1977. A preliminary chronology of Ethiopian droughts. In Drought in Africa, Dalby D. et al., Eds.
International African Institute, London. 68-73.
Yates D. N. & K. M. Strzepek, 1998. Modeling the Nile basin under climatic change. Journal of Hydrologic
Engineering 3 (2): 98-108.
14
January
July
35N
1018
06
4
1016
101
101
6
12
20N
10
06
1008
15N
1010
10
100
8
10
18
25N
20N
15N
1010
10
1020
10N
1010
10N
5N
12
5N
0
101
1014
1010
Equ
5S
10
16
5S
12
10
1012
12
10
Equ
12
10 1014
10
1014
101
2
1016
25N
30N
2
100
30N
1004
1012
35N
10S
20E
25E
30E
35E
40E
45E
10S
20E
25E
30E
35E
40E
45E
Fig.1 : Mean surface winds and sea-level pressure over North-East Africa in January and July.
Data : Long-term mean (1968-1996) NCEP-NCAR reanalysis (Kalnay et al., 1996). Data
over areas of complex topography should be considered as approximate. Bold line : July
InterTropical Front surface location as deduced from the shift from northerly to southerly
winds. Shading : regions of mean ascending motion at 400 hPa.
15
U wind
100
2
−2 −5
−10
0
−2
20
2
5
20
10
−10
200
5
−5
10
400
0
0
−2
300
10
5
500
5
−5
−5
700
2
−2
2
600
0
800
0
−2
−2
−2
0
2
900
0
−2
−2
1000
J
F
M
A
M
J
J
A
S
O
N
D
J
Fig.2 : Time-height distribution of mean zonal winds above central Sudan (15°N, 32.5°E), in
m.s-1. Data : Long-term mean (1968-1996) NCEP-NCAR reanalysis. Positive (negative)
values indicate westerlies (easterlies).
16
Mean annual rainfall distribution in the Nile Basin
30
Mean = 630 mm
25
% surface
20
15
10
5
0
0
300
600
900
1200
mm
1500
1800
2100
Fig.3 : Frequency distribution of mean annual rainfall in the Nile basin (1961-1990 average),
as a percentage of the total surface area. Data : Climate Research Unit CRU CL 2.0
climatology at a 10’ resolution (New et al., 2002).
17
100
Cai
30N
50
50
50
27N
50
Asw
24N
50
50
21N
100
18N
50
100
100
15N
200
300
400
Kha
300
600
12N
0
0
200
Asm
40
30
200
300
400
600
80
0
0 800
60
400
1300
16
00
800
00
Mal
9N
13
800
80
0
600
AA
2000
1600
00
Wau
13
1300
6N
800
Jub
600
3N
13
0
0
800
1300
Equ
1300
Kam
1300
1600
1300
16
00
1300
0
80060
3S
24E
27E
30E
33E
36E
39E
Fig.4 : Mean annual rainfall map of the Nile basin (1961-1990). Data : Climate Research
Unit CRU CL 2.0 climatology at a 10’ resolution (New et al., 2002). Isohyets in mm.
18
mm
mm
mm
mm
mm
mm
mm
mm
mm
Rainfall regimes
250
1
200
150
100
50 11
0
250
2
200
150
100
50
3
0
250
3
200
150
100
27
50
0
250
4
200
150
92
100
50
0
250
5
200
169
150
100
50
0
250
183
6
200
150
100
50
0
301
250
7
200
150
100
50
0
250
8
200
162
150
100
50
0
250
206
9
200
150
100
50
0
J F M A M J J A S O N D
1
30N
27N
24N
2
21N
18N
3
15N
4
12N
5
7
9N
6
6N
3N
8
Equ
9
3S
24E
27E
30E
33E
36E
39E
Fig.5 : Typology of rainfall regimes for the Nile basin. Based on a hierarchical cluster
analysis of mean monthly rainfall for 1961-1990 (CRU CL 2.0 data). Figures in each
panel indicate the maximum monthly rainfall. For type 7, type 6 is shown as a dashed line
for comparison.
19
Fig.6 : Mean diurnal distribution of precipitation along west-east cross-sections. Top : crosssection from the Nile plains to the Ethiopian Highlands, around 15°N (3-hourly rainfall
amount or occurrence as percentage of the total for May-October). Bottom : cross-section
from the lake Victoria west coast to the Indian Ocean, around 0-5°S (3-hourly rainfall
occurrence as percentage of the annual total). Data from Pedgley (1969 ; 1971), Tomsett
(1975).
20
2000
1980
1960
−1
2000
0
1
3
24E
1980
1960
−2
−2
2000
−1
2
−1
36E
1980
1960
0
1
24E
3S
Equ
3N
6N
9N
12N
15N
18N
21N
2
28E
0.5
0
0
PC1 − 19.6% var.
36E
0.3
32E
0
0 0
0.7
0.5
0
0.3
0
40E
3S
Equ
3N
6N
9N
12N
15N
0
0
18N
1
0.3
21N
PC2 − 13.4% var.
0.7
24E
28E
32E
0
0
0
0.5
0
0.
0.7 5
0 0.3
0.3
0
0
0
40E
3S
Equ
3N
6N
9N
2
28E
0
32E
5
0. 7
0.
0
0
0.3
12N
15N
18N
21N
0.5
0
0.5
0
7
PC3 − 11.2% var.
40E
36E
0
0
0.
0.3
Fig.7 Space-time patterns of the interannual variability of annual rainfall amounts, for 19512000. First 3 principal components (with a varimax rotation) of the correlation matrix.
Map shadings : areas whose rainfall is significantly correlated (5% confidence level) with
the corresponding time-series (bottom). Figures are correlation values.
21
21N
−50
18N
−50
−50
Kha
−50
−20
−50
0
−15
−15
−10
−2
0
Mal
−5
−10
0
−15
−10
0
−10 Wau−1 −5
−5
−1
−15
AA
−30
−20
5
6N
5
−5
9N
5
0
−3
−2
12N
−20
−15
−10 0
−5
−30
−30
Asm
−30
−10
15N
−10
−10
Jub
3N
0
−5
0
−5 0
−1
0
150 Kam
0
5
5
−10
−5
−5
500
−−1
0
15
−
0
0
−5
5
5
−1
0
10
Equ
5
3S
24E
27E
30E
33E
36E
39E
Fig.8 Percentage change in annual rainfall between 1951 and 2000 (slopes of the linear trend
as a percentage of the mean rainfall). Data : interpolated rainfall series based on the CRC
rainfall data base and CRU rainfall climatology. Dashed : rainfall decrease ; solid :
rainfall increase. Dots : significant decrease ; cross-hatching : significant increase
(Spearman’s ranks correlation, 95% confidence level).
22
Correlation with NINO34 / DMI
0.6
0.4
0.2
0
−0.2
30N
−0.4
27N
−0.6
0.6
24N
0.4
21N
0.2
0
18N
−0.2
−0.4
15N
−0.6
12N
0.6
0.4
9N
0.2
6N
0
−0.2
3N
−0.4
−0.6
Equ
0.6
3S
0.4
24E 27E 30E 33E 36E 39E 0.2
0
−0.2
−0.4
−0.6
0.22
1
1
2
3
4
5
3
−0.24
7
6
8
9
4
−0.37
5
−0.52
J F M A M J J A S O N D
0.6
0.4
0.2
0
−0.2
−0.4
−0.6
0.6
0.4
0.2
0
−0.2
−0.4
−0.6
0.6
0.4
0.2
0
−0.2
−0.4
−0.6
0.6
0.4
0.2
0
−0.2
−0.4
−0.6
0.42
6
−0.65
7
−0.63
0.59
8
−0.71
0.67
9
−0.54
J F M A M J J A S O N D
Fig.9 : Correlations between rainfall in the Nile basin and the Nino 3+4 index (solid line) and
the Indian Ocean Dipole Mode Index (dashed line). Period : 1951-2000. The computation
uses 3-mth averages (e.g., the correlation for July stands for June-August rainfall vs JuneAugust Nino 3+4), for all months with a mean rainfall above 5 mm. Figures indicates the
maximum correlation. Stars : significant correlations at the 99% level.
23
Corr. with NINO34 JAS 1951−2000
24N
Corr. with NINO34 OND 1951−2000
24N
Asw
21N
18N
Corr. with DMI OND 1951−2000
24N
Asw
21N
21N
18N
18N
Asw
−0.2
Kha
Asm
−0
−0.2
0.2
0.2
0.6
Jub
0.6
0.4
4
0.4
0.6
0.
0
−0.2
6
0.4
0.2
0
3S
3S
33E
Kam
Equ
2
0.
.2
−0
0
0.6
0.
0.4
Kam
0.4
0.2
.4
−0
Equ
0.4
0.4
Kam
(a)
0.4
0.2
0.2
3N
.4
30E
0.2
Jub0
.4
Equ
−0.2
AA
0.4
Wau
6N
0.2
3N
−0
9N
0
Wau
0.4
Mal
AA
0.2
0.4
0.2
0.
2
−0
.2
6N
0.2
0.
2
−0.4
0
−0
.2
−0.2
−0.4
0.
2
9N
0.2
−0
.2
0.2
Mal
AA
12N
2
0.
.2
Jub
27E
Asm
0
−0.2
−0
0
0.2
3N
24E
Kha
15N
0.4
−0.2
12N
.2
−0
−0.2
Mal
9N
6N
0.2
0
−0.2
−0.4
−0.2
Wau
Asm
.2
0
12N
Kha
15N
0.2
15N
36E
39E
0.4
3S
24E
27E
30E
(b)
33E
36E
39E
24E
27E
30E
33E
36E
39E
(c)
Fig.10 : Correlation maps between interpolated rainfall and climate indices. (a) Nino 3+4 and
July-September rainfall ; (b) Nino 3+4 and October-December rainfall ; (c) Indian Ocean
Dipole Mode Index and October-December rainfall. Period : 1951-2000. Dashed (solid)
lines : negative (positive) correlations. Shading : correlations significant at the 95% level.
24