Atmospheric and Climate Sciences, 2012, 2, 454-463
http://dx.doi.org/10.4236/acs.2012.24039 Published Online October 2012 (http://www.SciRP.org/journal/acs)
Climate Characteristics over Southern Highlands Tanzania
Yassin Mbululo1,2, Fatuma Nyihirani3
1School of Environmental Studies, China University of Geosciences, Wuhan, China
2Sokoine University of Agriculture, Morogoro, Tanzania
3Mzumbe University, Morogoro, Tanzania
Received June 24, 2012; revised July 28, 2012; accepted August 9, 2012
ABSTRACT
This study was conducted to examine the climate characteristic of southern highland Tanzania (Latitude 6˚S - 12˚S and
Longitude 29˚E - 38˚E). The study findings reveal that rainfall over the region is linked with SST over the Indian Ocean,
where warmer (cooler) western Indian Ocean is accompanied by high (low) amount of rainfall over Tanzania. During
wet (dry) years, weaker (stronger) equatorial westerlies and anticyclone (cyclonic) anomaly over the southern tropics
act to reduce (enhance) the export of equatorial moisture away from East Africa. The wettest (driest) season was found
to be 1978/79 (1999/00) which can be classified as the severely wet (moderate drought). Two different modes of rainfall
have been identified at time scale of 1.5 and 6 years which have been associated with the quasi biennial oscillation
(QBO) and El Nino Southern Oscillation (ENSO), respectively.
Keywords: Tanzania; Climate Characteristics; Dominant Periodicity Mode
1. Introduction
Tanzania has been experiencing unreliable and unpre-
dictable rainfall patterns for the past few decades. This
trend is an alarming problem to agricultural productivity
and wellbeing of the country since the sector is the back-
bone of the country economy. The north and northern coast
of the country tend to exhibit bimodal rainfall pattern
from March to May and October to November. These
patterns are locally known as the “long” or “Masika” and
“short” or “Vuli” rainfall seasons associated with the
northward and southward movement of the ITCZ, re-
spectively [1-3]. The southern, central and western parts
of the country exhibit unimodal rainfall pattern from
November to April, the period which coincides with
southern hemisphere summer. Southern highland Tanza-
nia (latitude 6˚S - 12˚S and longitude 29˚E - 38˚E) which
is studied here is the major cereal producing region.
Since the economy of the country depends mostly on rain
fed agriculture, therefore the country is vulnerable to the
impacts of the rainfall variability.
Over Tanzania, few studies on rainfall variability have
been done, mainly are focusing on the bimodal rainfall
areas (e.g. [1,4,5]) and ignoring the unimodal rainfall
areas. The importance of these unimodal rainfall areas
lies on the fact that, they are the major cereal producing
areas and catchment areas for the rivers which have been
tapped for hydroelectricity power generation [2]. Never-
theless, rainfall variability over the East Africa with rela-
tion to regional and remote atmospheric and oceano-
graphic parameters has been studied by many researchers
(e.g. [3-11]).
Goddard and Graham [6] experiment on Atmospheric
General Circulation Model (AGCM) suggests the Indian
Ocean sea surface temperature (SST) to be exerting a
greater influence over the East and Central Africa rainfall
than the Pacific ocean. They also found a considerable
modification of convective activities over equatorial Af-
rica and the tropical Indian Ocean to be directly linked
by the Pacific Ocean. Saji et al. [12] showed an East-
West dipole mode in the Indian Ocean SST anomalies
that is coupled to the atmospheric zonal circulation. They
also found Indian Ocean Dipole (IOD) to have signifi-
cant correlation with East African rains where the rainfall
is increased during a positive event and decreased during
a negative event. Study by Clark et al. [13] demonstrated
the strong correlation which exists between Indian Ocean
SST and East African short rains. The authors ascribe the
pattern to reoccurring coupled ocean atmosphere phe-
nomena (IOD).
Several other studies (e.g. [7,14]) have tried to deli-
neate homogeneous rainfall region over East Africa and
Tanzania in particular. Basalirwa et al. [14] performed
principal component analysis and delineate Tanzania into
15 homogeneous rainfall regions based on the network of
150 widely distributed rainfall stations. Indeje et al. [7]
performed another principle component analysis and
simple correlation analyses using a network of 136 rain-
C
opyright © 2012 SciRes. ACS

Y. MBULULO, F. NYIHIRANI 455
fall stations over East Africa. Their analyses yielded 8
homogeneous rainfall regions over East Africa, among
them, 5 were found in Tanzania.
Nicholson and Entekhabi [15] have demonstrated the
interannual variability of rainfall in much of Africa to be
characterized by strong quasi periodic fluctuations in 2.2 -
2.4, 2.6 - 2.8, 3.3 - 3.8 and 6.0 - 6.3 years spectral bands.
The author acknowledges the presence of other distinct
quasi periodicities which are evident throughout equato-
rial and southern Africa. Study by Indeje and Semazzi
[16] found rainfall oscillates in East Africa to have
dominant periodicity of between 1.7 - 2.5 and 4 - 5 years,
which appear to be related to quasi biennial oscillations
(QBO) and ENSO events respectively. Therefore, it is
the interest of this paper to find out the characteristics of
the rainfall over southern highlands Tanzania and the
reasons for the observed characteristics since the study
area is of significance importance to the economy of the
country.
2. Data and Methodology
2.1. Data
The data used in this study comprises monthly mean
rainfall data from 16 meteorological stations found in
southern highlands Tanzania for the duration of 41 years
from 1970 to 2010. These data were obtained from the
Tanzania Meteorological Agency (TMA) where the name
and geographical positions of these stations have been
presented in Table 1. As the source of the climate rich
data sources, monthly mean zonal and meridional winds
data set stored at 2.5˚ × 2.5˚ grid boxes, from National
Centre for Environmental Prediction (NCEP)/National Cen-
ter for Atmospheric Research (NCAR) reanalysis [17],
Kaplan SST anomaly data stored at 5˚ × 5˚ grid boxes,
Global Ocean Surface Temperature Atlas (GOSTA) data
set [18] were used. NCEP/ NCAR data has been used in
numerous c1imatological studies in Tanzania, such as
those of Mpeta [2]; Mapande and Reason [8,19] among
others and yielded promising results.
Southern highland Tanzania (latitude 6˚S - 12˚S and
longitude 29˚E - 38˚E) which is studied here exhibits uni-
modal rainfall pattern and the rainfall tends to start in
November to April. Four regions (Iringa, Mbeya, Rukwa
and Ruvuma) shown in Figure 1 were used to study the
climate characteristics of the region.
2.2. Methodology
In order to study rainfall characteristics over an area and
at the same time to capture mechanisms which cause that
characteristic, the Standardized Precipitation Index (SPI)
method was used. The SPI is the probability index which
was developed by McKee et al. [20] to give better rep-
Figure 1. Map of Tanzania showing the study area, Iringa,
Mbeya, Rukwa and Ruvuma region.
Table 1. Name and geographical positions of meteorological
stations over southern highlands Tanzania
Station Name Latitude (S) Longitude (E)
Iringa Met Stn (Nduli) 7.60˚ 35.80˚
Ludewa bomani 10.00˚ 30.40˚
Njombe bomani 9.30˚ 34.80˚
Iringa Experimental Stn 7.46˚ 35.41˚
Kyela boma 9.35˚ 33.51˚
Tukuyu Agric 9.10˚ 33.35˚
Mitalula 9.23˚ 33.37˚
Mbimba Coffee Research9.40˚ 32.58˚
Mbarali Irr. Scheme 8.40˚ 34.15˚
Mbeya Met 8.56˚ 33.28˚
Mpanda Boma 6.20˚ 31.50˚
Kisanga hydromet 7.18˚ 36.47˚
Namanyere-Nkansi 7.31˚ 31.30˚
Sumbawanga Agric Stn 7.57˚ 31.36˚
Tunduru Agriculture 11.60˚ 37.22˚
Songea Airfield 10.40˚ 35.35˚
resentation of abnormal wetness and dryness [21]. Since
its development, the index has gained increasing accep-
tance in the United States and other parts of the world as
a valuable tool for monitoring drought. It is currently
being used by the U.S National Drought Mitigation Cen-
ter, the Western Regional Climate Center, as well as the
Colorado Climate Center [22,23]. The index uses only
precipitation data thus making the analysis possible even

Y. MBULULO, F. NYIHIRANI
456
in the absence of other parameters.
The SPI is essentially a standardizing transform of the
probability of the observed precipitation. It can be com-
puted for a precipitation total observed over any duration
desired by a user (e.g. 1 month SPI, 3 month SPI, 6
month SPI, 12 month SPI, 24 month SPI etc.); short term
durations of the order of months may be important to
agricultural interests while long term durations spanning
years may be important to water resources management
purpose because of the slow inherent responses in water
bodies to rainfall changes [22-24]. The method is ca-
pable of returning essential parameters after the analysis
such as severity, magnitude, and frequency of the
drought. Heim [24] reveals that, even though index was
developed purposely for use in Colorado, it can be ap-
plied universally to any location. Furthermore, Manatsa
et al. [23] found the index to be temporarily and spatially
comparable, independent of geographical and topog-
raphical differences, and even relevant in regions with
diverse rainfall patterns.
For the purpose of brevity of this study, a “black box”
software package was downloaded
(http://www.drought.un l.edu/MonitoringToo ls/Download
ableSPIProgram.aspx) and used for computation for
which the input is a rainfall data time series and the out-
put is the SPI. Details on the mathematical computation
of SPI may be found in Guttman [21]; Bordi et al. [22]
and Lloyd-Hughes and Saunders [25] among others.
2.3. Regional SPI
To calculate the Regional SPI, aerial average stations
6-month scale SPI values for the month of April was
taken. This 6 month scale counts from November of the
previous year to April of the current year. Specifically,
Regional SPI is defined by
Regional SPIy = N1
1SPI N
y
i



 (1)
where N is the number of the regional stations operating
in the year y (in this case N = 16 and y = 1971, 1972…
2010). SPI values, Percentage of occurrence and nominal
class descriptions have been presented in Table 2.
3. Results and Discussion
3.1. Region Standardized Precipitation Index
Figures 2(a)-(d) show the Standardized Precipitation
Index (SPI) for one month for Iringa, Mbeya, Rukwa and
Ruvuma region respectively. Observational study on
these SPI figures shows substantial interannual variabi-
lity of rainfall for each station. Moreover, some degree of
persistence in rainfall anomalies can be seen from each
station. The positive (negative) SPI value shows the
rainfall of above (below) normal and the number is the
magnitude of the departure from normal.
(a)
(b)
(c)
(d)
Figure 2. Schematic representation of Standardized Pre-
cipitation Index (SPI) for one month period from January
1970 to December 2010. The x-axis represents Years and
y-axis represents Amplitude of SPI value. (a) Iringa (b)
Mbeya (c) Rukwa (d) Ruvuma region.
3.2. Regional Standardized Precipitation Index
In order to relate rainfall demand with practical applicat-
ions, 6-month time steps of SPI for the November to

Y. MBULULO, F. NYIHIRANI 457
April rain season have been applied in this study. During
these months is when the agricultural activities are taking
place such as plowing and seed planting so the soil
moisture demand is at maximum level, therefore relation
of this period is done to mean rainfall amount. Most of
the annual rainfall experienced in southern highland
Tanzania occurs during this period, and as such, water
availability for vegetation is determined primarily by the
amount of seasonal rainfall alone. So this seasonal rain-
fall is the most important single factor for water avail-
ability in agricultural activities. The 6-month scale is not
only suitable in this research work, but it is also the most
common used time scale for regional agricultural drought.
Other research works which used this time scale are like
those of Manatsa et al. [18]; Ntale and Gan [25] among
others.
Moreover, other months which fall outside the grow-
ing period (May to October) will not be analyzed because
the region during this period is normally virtual dry. In
any case, the rainfall that falls within this period mostly
goes to waste since crop growing is not undertaken dur-
ing this period. Thus, 6-month SPI analyzed during this
dry period may be misleading in the sense that large
negative or positive SPI values may be associated with
rainfall not very different from the mean. This is because
during the dry periods, the mean total will be small, and
hence, relatively small deviations on either side of the
mean could have large negative or positive SPI values.
Figure 3 represents the regional SPI (Iringa, Mbeya,
Rukwa and Ruvuma) for the duration of 6-month. The
plot shows the departures in November to April from
1970/71 to 2009/10 seasons for the region where the year
axis refers to the April month (i.e. For 1970/71 season, it
represent November of 1970 to April of 1971) of a given
season. Although there are some variations in magnitude
of the anomalies for each year, the region is experiencing
the normal rainfall as the SPI values are at the range of
−0.523 to 0.523 values as indicated in Table 2.
In order to shed more light on the regional SPI, the
Mexican hat wavelet method was used to reveal the rain-
fall characteristics of the region. Wavelet representation
for the regional SPI presented in Figure 4 as contour
maps shows the largest power at the time scale of 4 to 8
years particularly in the 1970s to 1990s, and time scale of
2 to 4 years in the 1990s to 2005s. Figure 5 shows the
series of purple (shaded) and white (unshaded) bands
with numbers representing powers. The purple band
represents rainfall of above normal while the white band
represents rainfall of below normal. The dominant perio-
dicity mode is seen at the time scale of 1.5 and 6 years
which may be associated with the quasi biennial oscilla-
tion (QBO) and El Nino Southern Oscillation (ENSO).
Similar results to this quasi periodic oscillation over the
region have been presented by Mpeta [2]; Nicholson and
Figure 3. Schematic representation of Regional (Iringa,
Mbeya, Rukwa and Ruvuma) Standardized Precipitation
Index (SPI) for six month, November to April from 1970 to
2010. The period which fall outside this six month, May to
October are not shown as they are normally dry. The x-axis
is Year and y-axis is the amplitude deviation of SPI from
normal.
Table 2. Table showing classification of SPI value and their
corresponding event probability which is commonly used in
southern African region adopted from Manatsa et al. [23].
SPI value
occurrence % Occurrence Nominal SPI class
>1.645 ≤5 Extremely wet
1.644 to 1.282 6 - 10 Severely wet
0.842 to 1.281 11 - 20 Moderate wet
0.524 to 0.841 21 - 33 Slightly wet
−0.523 to 0.523 34 - 50 Normal
−0.841 to −0.524 21 - 33 Slightly drought
−1.281 to −0.842 11 - 20 Moderate drought
−1.644 to −1.282 6 - 10 Severely drought
<−1.645 ≤5 Extremely drought
Figure 4. Wavelet power spectrum of Standardized Pre-
cipitation Index (SPI) for six month, November to April
period from January 1970 to December 2010 showing the
largest powers found within the region (Iringa, Mbeya,
Ruvuma and Rukwa) during the study period.

Y. MBULULO, F. NYIHIRANI
458
Entekhabi [15] and Indeje and Semazzi [16] among others.
Furthermore, Figure 5 shows above and below normal
rainfall to be dominant between the year 1978s to 1991s
and the year 1999s to 2010s respectively; consistence
with results presented in Table 3.
Inspection from the regional SPI value obtained re-
veals that, the wettest season in record was that of
1978/1979 which can be classified as Severely wet (SPI
value of 1.62) and the driest season was that of 1999/
2000 which can be classified as Moderate drought (SPI
value of −1.26) according to Manatsa et al. [23] classifi-
cation presented in Table 2. Table 3 shows five (5) wet-
test and driest years in record for the region during the
study period. Similar results have been shown by Ogallo
[3] and Mapande and Reason [8,19] for the period in
between 1970 to 1999 and 1970 to 1993 respectively, the
period which compromise with this research work.
3.3. Wind Flow Pattern at 850 hPa Level and
SST Anomaly
The results presented to this point indicate that Atmos-
Figure 5. Wavelet representation of Standardized Precipi-
tation Index (SPI) for six month, November to April period
from January 1970 to December 2010 showing the domi-
nant periodicity modes found within the region (Iringa,
Mbeya, Ruvuma and Rukwa) during the study period.
Table 3. Table showing the wettest and driest years in re-
cords for southern highland Tanzania during the study
period of 41 year, from 1970 to 2010.
No. Wettest years Driest years
1 1977/78 1976/77
2 1978/79 1987/88
3 1984/85 1999/00
4 1988/89 2002/03
5 1997/98 2005/06
pheric circulations over the Indian Ocean act as an
important factor in influencing rainfall over East Africa.
Study by Goddard and Graham [6] reveal 850 hPa level
to be most representative of the behaviour of the
vertically integrated moisture flux over East Africa. This
level was proposed for the purpose of reducing the
effects of high grounds in some areas. Indeed, low level
winds are of significant importance as the convergence
of these winds may result in the development of clouds
and precipitation when enough moisture content is
present [2]. Thus, the data from 850 hPa were used as
most representative of the behavior of the low level
winds in this region. Wettest (1997/1998) and driest
(1999/2000) years were selected from Table 3 to study
the dominant moisture patterns and SST anomaly.
Monthly plots corresponding to the approximate begin-
ning, peak and end periods of the rainy season over
southern highland Tanzania in November, February and
April are shown. Other months which tend to show
anomalies that are a transition between those for Novem-
ber and February, and between February and April res-
pectively, are not shown for brevity.
In November, the beginning of the rain season of the
wet year indicates anomalously positive SST anomalies
over the western Indian Ocean, particularly at the
northeastern part of Madagascar (Figure 6(a)). Study by
Webster et al. [11] reveals the same trend and suggest
that, the development of this anomalous warm SST
anomaly in western Indian Ocean started in June 1997
and reached its maximum in February 1998 (Figure
6(b)). At about the same time, the eastern Indian Ocean
developed a strong negative SST anomaly starting in July
1997 and reached a maximum in November 1997
(Figure 6(a)). This SST anomaly pattern is evocative of
the subtropical South Indian Ocean SST dipole pattern
suggested by Saji et al. [12]; Behera and Yamagata [26]
and tends to influence southern African summer rainfall.
Over the Atlantic Ocean, positive SST anomalies are
observed in the Gulf of Guinea, Namibia and Angola
coast. Several studies (e.g. [8,27]) suggest that these
positive SST anomalies are evocative of Benguela Nino
in the tropical and southeast Atlantic Ocean. The
Benguela Nino is often associated with floods in Angola
and Namibia and abundant rainfall in the usually arid
Namib Desert [27].
Simulated moisture patterns in November, the begin-
ning of rain season of the wet year indicate that there is
enhanced northeasterly and easterly moisture flux
convergence hence increasing moisture penetration in
Tanzania (Figure 7(a)). Also there is presence of
enhanced cyclonic flow over northern Namibia which
induces an increased flow of moist air. As a result,
increased low level moisture convergence is expected
over northern Zambia and southern highland Tanzania.

Y. MBULULO, F. NYIHIRANI 459
(a)
(b)
(c)
Figure 6. Composite SST anomaly during the rainy season
of wet year of the 1997/98 in the Atlantic and Indian Ocean
in different time. (a) November, 1997 during the beginning
of rain season (b) February, 1998 during the peak of the
rain season (c) April, 1998 near the end of rain season.
Another moisture source for East Africa is the tropical
East Atlantic Ocean. Anomalous westerlies from East
Atlantic and Congo basin oppose the easterly flux that
exists at this time of year from the Indian Ocean (Figure
7(a)). As a result, relative convergence of low level
moisture occurs over western Tanzania, Uganda and
northern Zambia.
In February, during the peak of the rain season of the
wet year anomalously positive SST anomalies is en-
hanced all over the tropical western Indian Ocean (Fig-
ure 6(b)). Warmer SSTs near Tanzania coast act to
increase the local evaporation, and hence the low level
moisture over the land. Simulated moisture pattern during
this period indicates that the northeasterly wind has en-
hanced and it penetrates further south (Figure 7(b)). A
second source of moisture which is westerly wind from
Atlantic ocean and Congo basin have also strengthened
therefore transporting more moisture into great lake
region and northern Mozambique (Figure 7(b)). A
cyclonic center can be seen over northern Namibia due to
recurving of northeasterlies to south-westerlies. Obser-
vations show that, the lower tropospheric anticyclonic
anomaly over the tropical south Indian Ocean is stronger
than November when the rain season started. Study by
Mapande and Reason [8] suggest this anticyclonic cir-
culation feature in November and February to be pos-
sibly part of the local atmospheric response of cool SST
anomaly in the tropical south Indian Ocean. Moreover,
tropical Indian Ocean is seen to be dominated by east-
erlies during this peak season as the result of warming of
the SST in the west and cooling in the East Indian Ocean
[28]. Taken together with this, the extreme continental
southward location of ITCZ between Tanzania and
central Mozambique during December and February [3,4,
6,7], suggests a substantial rainfall over southern Tan-
zania during this period.
Towards the end of the rainy season in April, cyclonic
moisture flux anomalies (Figure 7(c)) emerge over the
tropical equatorial western Indian Ocean consistent with
the small warm SST anomalies (Figure 6(c)) in those
regions. The anomalous positive SST anomaly at Indian
Ocean is seen to shift toward the East of the Indian
Ocean. The presence of strong southeasterly moisture
flux which is fed by south western Indian Ocean region
imply strong moisture is fetch toward East Africa and
Congo basin. Such conditions are favorable for ongoing
agricultural activities which are taking place during this
time of the year.
During dry year of 1999/2000, simulation of 850 hPa
level wind pattern show the equatorial westerlies and
south Indian Ocean trade winds are reversed from those
of wet scenario described above for 1997/1998. In
November, during the beginning of the rain season of the
dry year indicates anomalously positive SST anomalies
in eastern Indian Ocean, while the western part is
dominated by near normal SST (Figure 8(a)). The same
scenario can be seen over eastern Atlantic Ocean where
near normal SST is observed. The inverse is true for the
wet scenario discussed above (Figure 6(a)). Simulated
moisture pattern during the beginning of rainy season of
the dry year is dominated by northeasterly and weak
easterly wind (Figure 9(a)). A weak cyclonic flow is
seen to develop at the western Indian Ocean due to the
cool SST at this region of the ocean. Also there is
westerly wind observed at the tropical center Indian
Ocean which denies moisture transport to Tanzania as
the result of turning southeasterly wind.
In February, during the peak of the rain season of the
dry year (Figure 8(b)), anomalously negative SST

Y. MBULULO, F. NYIHIRANI
460
(a)
(b)
(c)
Figure 7. Simulated moisture pattern during the rainy sea-
son of the wet year of 1997/98 at 850 hPa level (ms−1). (a)
November, 1997 during the beginning of rain season (b)
February, 1998 during the peak of the rain season (c) April,
1998 near the end of rain season.
anomalies is developed over south western Indian Ocean.
Simulation moisture flux at this period indicates a lower
tropospheric anticyclone flow over western Indian Ocean
right at Tanzania coast and cyclonic flow near northern
Namibia (Figure 9(b)). The northeasterly moisture flux
anomaly over the western Indian Ocean turns more north
and south due to the presence of anticyclone thereby
denying inland penetration of moisture. Not only that,
but also strong westerlies wind is seen over western
Indian Ocean, as the result more moisture is transport
moisture away from mainland Tanzania. In comparison
to the scenario of the wet year (Figure 7(b)), there is a
stronger divergence of moisture from the Indian Ocean
during this period. Over the Congo Basin, the easterly
moisture flux is also turning northward different from the
case seen during wet year (Figure 7(b)) thus, further
export of moisture away from the country. Little
moisture found in Tanzania is also exported to Madaga-
scar by northeasterly and westerlies wind. As it was the
case for the wet scenario, a cyclonic center can be seen
over northern Namibia as the result of recurving of
northeasterlies to southwesterlies.
In April, near the end of the rainy season of the dry
year (Figure 8(c)), positive SST anomaly emanate at
southwestern Indian Ocean which led to development of
anticyclone flow. Simulation of the moisture flux
indicates the easterly over southwestern Indian Ocean
(a)
(b)
(c)
Figure 8. Composite SST anomaly during the rainy season
of dry year of the 1999/00 in the Atlantic and Indian Ocean
at different time. (a) November, 1999 during the beginning
of rain season (b) February, 2000 during the peak of the
rain season (c) April, 2000 near the end of rain season.

Y. MBULULO, F. NYIHIRANI 461
(a)
(b)
(c)
Figure 9. Simulated moisture pattern during the rainy sea-
son of the dry year of 1999/00 at 850 hPa level (ms-1). (a)
November, 1999 during the beginning of rain season (b)
February, 2000 during the peak of the rain season (c) April,
2000 near the end of rain season.
(Figure 9(c)) turns more southerly due to development
of this anticyclone. On the other hand the northeasterly
turn more north over the East Africa coast due to de-
velopment of weak cyclone and cross equatorial westerly
over the Indian Ocean thereby denying inland penetration
of moisture to Tanzania. It can also be seen that, there is
weak westerly from Congo basin converge with the
easterly from Kenya at Lake Victoria basin.
In general, for wet year the moisture flux pattern and
SST anomaly plots suggest enhanced convective precipi-
tation over Tanzania to be associated with low level
moisture flux from the northeast and southeast Indian
Ocean followed by tropical Atlantic Ocean and Congo
basin. For the case of dry years, simulation of moisture
flux pattern suggest weak easterly propagating wind to
be more apparent over Tanzania coast followed by weak
low level moisture flux from Atlantic Ocean and Congo
basin due to the dominance of cool SST. Since the pat-
tern of winds during the wet years are opposite to that of
the dry years, their reversal provide credence to the sug-
gestion by Goddard and Graham [6]; Mapande and Rea-
son [8] and Mpeta and Jury [9] that the modulation of the
atmospheric circulation over the Indian ocean is impor-
tant factor in influencing rainfall over Tanzania regard-
less of how much incoming moisture content turns in
Mozambique Channel at the beginning of rain season.
4. Conclusions
The climate characteristic of southern highland Tanzania
was studied. It was found that rainfall over the region is
linked with the sea surface temperature over the Indian
Ocean, where warmer (cooler) western Indian Ocean is
accompanied by high (low) amount of rainfall over most
part of the Tanzania. During wet (dry) years, weaker
(stronger) equatorial westerlies and anticyclone (cyclonic)
anomaly over the southern tropics act to reduce (enhance)
the export of equatorial moisture away from East Africa.
Not only that, but also moisture influx from the northeast
Indian monsoon has significant influence on the rainfall
over the region. During the wet years, strong northeast-
erly Indian monsoon is evident over most of Tanzania
while during the dry year the northeasterly is seen to turn
north hence denying moisture influx over Tanzania. In
addition, increased (decreased) low level moisture influx
from gulf of Guinea and Congo basin tend to occur dur-
ing the wet (dry) seasons, leading to enhanced (reduced)
low level moisture convergence over western part of
Tanzania.
Although there are some variations in magnitude of
the anomalies for each year, the region is experiencing
normal rainfall. The results suggest the wettest season in
record during the study period to be 1978/1979 which
can be classified as severely wet and the driest season to
be 1999/2000 which can be classified as moderate
drought. Furthermore, analysis of rainfall amount re-
ceived throughout the study period shows that domi-
nance of above normal rainfall condition during 1978s to
1991s and below normal condition during 2000s to 2010s.
Different dominant periodicity modes have been ob-
served over the study period, but two of them seem to be
more dominant over the whole study period. These
modes of rainfall have been identified at time scale of 1.5
and 6 years which may be associated with the quasi bi-
ennial oscillation (QBO) and El Nino Southern Oscilla-
tion (ENSO) respectively.
REFERENCES
[1] T. A. Kabanda and M. R. Jury, “Interannual Variability of

Y. MBULULO, F. NYIHIRANI
462
Short Rains over Northern Tanzania,” Climate Research,
Vol. 13, No. 3, 1999, pp. 231-241. doi:10.3354/cr013231
[2] E. J. Mpeta, “Mechanisms of Interannual Rainfall Vari-
ability over Tropical Highlands of Africa and Its Predict-
ability Potential,” Ph.D. Thesis, Department of Geogra-
phy and Environmental Studies, University of Zululand,
Empangeni, 2002.
[3] L. A. Ogallo, “Dynamics of the East African Climate,”
Journal of Earth System Science, Vol. 102, 1993, pp.
203-217.
[4] S. K. Behera, J. Luo, S. Masson, T. Yamagata, P. Dele-
cluse, S. Gualdi and A. Navarra, “Paramount Impact of
the Indian Ocean Dipole on the East African Short Rains:
A CGCM Study,” Journal of Climate, Vol. 18, No. 21,
2005, pp. 4514-4530. doi:10.1175/JCLI3541.1
[5] M. Latif, D. Dommenget, M. Dima and A. Grotzner, “The
Role of Indian Ocean Sea Surface Temperature in Forcing
East African Rainfall Anomalies during December-Ja-
nuary 1997/98,” Journal of Climate, Vol. 12, No. 12,
1999, pp. 3497-3504.
[6] L. Goddard and N. E. Graham, “Importance of the Indian
Ocean for Simulating Rainfall Anomalies over Eastern
and Southern Africa,” Journal of Geophysical Research,
Vol. 104, No. d16, 1999, pp.19,099-19,116.
[7] M. Indeje, F. H. M. Semazzi and L. J. Ogallo, “ENSO
Signals in East African Rainfall Seasons,” International
Journal of Climatology, Vol. 20, No. 1, 2000, pp. 19-46.
doi:10.1002/(SICI)1097-0088(200001)20:1<19::AID-JO
C449>3.0.CO;2-0
[8] A. T. Mapande and C. J. C. Reason, “Links between
Rainfall Variability on Intraseasonal and Interannual
Scales over Western Tanzania and Regionacirculation and
SST Patterns,” Meteorology and Atmospheric Physics,
Vol. 89, No. 1-4, 2005, pp. 215-234.
doi:10.1007/s00703-005-0130-2
[9] E. J. Mpeta and M. R. Jury, “Intraseasonal Convective
Structure and Evolution over Tropical East Africa,” Cli-
mate Research, Vol. 17, No. 1, 2001, pp. 83-92.
doi:10.3354/cr017083
[10] H. K. Ntale and T. Y. Gan, “East African Rainfall anom-
aly Patterns in Association with El Nino Southern Oscil-
lation,” Journal of Hydrologic Engineering, Vol. 9, No. 4,
2004, pp. 257-268.
doi:10.1061/(ASCE)1084-0699(2004)9:4(257)
[11] P. J. Webster, J. P. Loschnigg, A. M. Moore and R. R.
Leben, “Coupled Ocean-Atmosphere Dynamics in the In-
dian Ocean during 1997-98,” Nature, Vol. 401, No. 6751,
1999, pp. 356-360. doi:10.1038/43848
[12] N. H. Saji, B. N. Goswami, P. N. Vinayachandran and T.
Yamagata, “A Dipole Mode in the Tropical Indian
Ocean,” Nature, Vol. 401, No. 6751, 1999, pp. 360-363.
doi:10.1038/43855
[13] C. O. Clark, P. J. Webster and J. E. Cole, “Interdecadal
Variability of the Relationship between the Indian Ocean
Zonal Mode and East African Coastal Rainfall Anoma-
lies,” Journal of Climate, Vol. 16, No. 3, 2003, pp.
548-554.
doi:10.1175/1520-0442(2003)016<0548:IVOTRB>2.0.C
O;2
[14] C. P. K. Basalirwa, J. O. Odiyo, R. J. Mngodo and E. J.
Mpeta, “The Climatological Regions of Tanzania Based
on the Rainfall Characteristics,” International Journal of
Climatology, Vol. 19, No. 1, 1999, pp. 69-80.
doi:10.1002/(SICI)1097-0088(199901)19:1<69::AID-JO
C343>3.0.CO;2-M
[15] S. E. Nicholson and D. Entekhabi, “The Quasi Periodic
Behavior of Rainfall Variability in Africa and Its Rela-
tionship to the Southern Oscillation,” Archives for Mete-
orology, Geophysics and Bioclimatology, Vol. 34, No.
3-4, 1986, pp. 311-348.
[16] M. Indeje and F. H. M. Semazzi, “Relationships between
QBO in the Lower Equatorial Stratospheric Zonal Winds
and East African Seasonal Rainfall,” Meteorology and
Atmospheric Physics, Vol. 73, No. 3-4, 2000, pp. 227-244.
doi:10.1007/s007030050075
[17] E. Kalnay, M. Kanamitsu, R. Kistler, W. Collins, D.
Deaven, L. Gandin, M. Iredell, S. Saha, G. White, J.
Woollen, Y. Zhu, M. Chelliah, W. Ebisuzaki, W. Higgins,
J. Janowiak, K. C. Mo, C. Ropelewski, J. Wang, A.
Leetmaa, R. Reynolds, R. Jenne and D. Joseph, “The
NCEP/NCAR 40 Year Reanalysis Project,” Bulletin of
American Meteorology Society, Vol. 77, No. 3, 1996, pp.
437-471.
doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO
;2
[18] A. Kaplan, M. A. Cane, Y. Kushnir, A. C. Clement, M. B.
Blumenthal and B. Rajagopalan, “Analyses of Global Sea
Surface Temperature 1856-1991,” Journal of Geophysical
Research, Vol. 103, No. C9, 1998, pp. 18,567-18,589.
doi:10.1029/97JC01736
[19] A. T. Mapande, and C. J. C. Reason, “Interannual Rain-
fall Variability over Western Tanzania,” International
Journal of Climatology, Vol. 25, No. 10, 2005, pp. 1355-
1368. doi:10.1002/joc.1193
[20] T. B. McKee, N. J. Doesken and J. Kleist, “The Rela-
tionship of Drought Frequency and Duration to Time
Scales,” 8th Conference on Applied Climatology, Ana-
heim, 1993, pp. 179-184.
[21] N. B. Guttman, “Accepting the Standardized Precipitation
Index, A Calculation Algorithm,” Journal of the Ameri-
can Water Resources Association, Vol. 35, No. 2, 1999,
pp. 311-322. doi:10.1111/j.1752-1688.1999.tb03592.x
[22] I. Bordi, S. Frigio, P. Parenti, A. Speranza and A. Sutera,
“The Analysis of the Standardized Precipitation Index in
the Mediterranean Area; Large Scale Patterns,” Annal
Sgeofisica, Vol. 44, No. 5-6, 2001, pp. 965-978.
[23] D. Manatsa, G. Mukwada, E. Siziba and T. Chinyanganya,
“Analysis of Multidimensional Aspects of Agricultural
Droughts in Zimbabwe Using the Standardized Precipita-
tion Index (SPI),” Theoretical and Applied Climatology,
Vol. 102, No. 3-4, 2010, pp. 287-305.
doi:10.1007/s00704-010-0262-2
[24] R. R. Heim, “A Review of Twentieth Century Drought
Indices Used in the United States,” American Meteoro-
logical Society, Vol. 83, No. 8, 2002, pp. 1149-1165.
[25] B. Lloyd-Hughes and M. A. Saunders, “A Drought Cli-
matology for Europe,” International Journal of Clima-
tology, Vol. 22, No. 13, 2002, pp. 1571-1592.

Y. MBULULO, F. NYIHIRANI
463
doi:10.1002/joc.846
[26] S. K. Behera and T. Yamagata, “Subtropical SST Dipole
Events in the Southern Indian Ocean,” Geophysical Re-
search Letters, Vol. 28, No. 2, 2001, pp. 327-330.
doi:10.1029/2000GL011451
[27] P. Florenchie, J. R. E. Lutjeharms, C. J. C. Reason, S.
Masson and M. Rouault, “The Source of Benguela Ninos
in the South Atlantic Ocean,” Geophysical Research Let-
ters, Vol. 30, No. 10, 2003, 4 pp.
doi:10.1029/2003GL017172
[28] S. K. Behera and T. Yamagata, “Influence of the Indian
Ocean Dipole on the Southern Oscillation,” Journal of the
Meteorological Society of Japan, Vol. 81, No. 1, 2003, pp.
169-177. doi:10.2151/jmsj.81.169