Civil and Environmental Research www.iiste.

org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

Hydrodynamics of Selected Ethiopian Rift Lakes

Kassaye Bewketu
Computational Science program, Faculty of Science, Addis Ababa University, Addis Ababa, Ethiopia.

Professor Tenalem Ayenew

Lecturer , Department of Earth Sciences, Faculty of Science, Addis Ababa University, Addis Ababa, Ethiopia.

ABSTRACT

The Main Ethiopian Rift Valley Lakes suffer from water level fluctuations due to several natural and
anthropogenic factors. Lakes located at terminal positions (no surface water outflow) are highly affected by the
fluctuations. These fluctuations are disturbing the stability of the ecosystems, putting very serious impacts on the
lives of many animals and plants around the lakes. Hence, studying the hydrodynamics of the lakes was found to
be very essential. The main purpose of this study is to find the most significant factors that contribute to the
water level fluctuations and also to quantify the fluctuations so as to identify the lakes that need special attention.
The research methodology includes correlation and least squares regression of lake levels on rainfall, discharge
and evaporation, multi-temporal satellite image analysis and land use change assessment. The results of the study
revealed that much of the fluctuations in the lake water levels are caused by human activities especially for the
lakes in the Central Ethiopian Rift. Lakes Abiyata, Chamo, Ziway and Langano are declining while Abaya and
Hawassa are rising. Among the studied lakes, Abiyata is drastically reduced in size (about 28% of its area in
1986) due to both human activities (most dominant ones) and natural factors. The other seriously affected lake is
Chamo with about 11% reduction in its area between 1986 and 2010. Lake Abaya was found to be relatively
stable during this period (showed only a 0.8% increase in its area).

KEYWORDS: Correlation, Fluctuation, Lake Level, Landsat Satellite Images, Land Use Change, Least Squares
Regression.

1. INTRODUCTION
A lake is a large, inland body of standing water that occupies a depression in the land surface.Lakes and lake
shores are attractive places to live and play. Clean, sparkling water, abundant wildlife, beautiful scenery, aquatic
recreation and fresh breezes all come to our mind when we think of going to the lake. Despite their great value,
lakes are fragile and ephemeral (Thompson et al., 2005)

The water-level of a lake changes seasonally and fluctuates annually due to the difference of seasonal or yearly
precipitation and evaporation (Kinshiro, 1974). The levels and sizes of lakes are governed by many natural and
anthropogenic factors. Climatic, hydrological and man-induced factors control lake levels in many ways.
Changes in lake levels result from a shift in the water balance or the net steady-state removal of water via
various surface and subsurface processes. In particular, closed terminal lakes fluctuate significantly in response
to climatic changes but tend to maintain equilibrium between input and output (Tenalem Ayenew, 2002). Only
when we pay attention to the vicinity of a lake, we happen to recognize some terrace which shows the past shore-
line or water level of the lake. In such a case, it is evident that stable and unstable periods existed in the past in
regard to the secular balance of water in the lake. In general, the main factors working on the regulation of such a
lake-level, that is, the reducing factors in the range of fluctuation of water storage are river discharge, ground
water discharge, river inflow, ground water inflow and lake surface area (Kinshiro,1974). Especially, the change
of river discharge (open lake) or ground water discharge (closed lake) greatly affects annual lake-level stability.
In an arid zone, such stability of a closed lake is regulated by the change of evaporation surface, too (Kinshiro,
1974).

The Ethiopian rift is characterized by a chain of lakes of various sizes and hydrological and hydrogeological
settings. The rift lakes and feeder rivers are used for irrigation, soda extraction, commercial fish farming, and
recreation, and they support a wide variety of endemic birds and wild animals. The levels of some of these lakes
have changed dramatically over the last three decades. Lakes that are relatively uninfluenced by human activities
(Langano and Abaya) remain stable except for the usual inter-annual variations, strongly influenced by rainfall.

46
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

Some lakes have shrunk due to excessive abstraction of water (such as Abiyata); others have expanded due to
increases in surface runoff and groundwater flux from percolated irrigation water such as Beseka (Tenalem
Ayenew, 2009). Apart from the various inflow and outflow components, the water balances of lakes are
governed by climate, anthropogenic factors, volcano-tectonism, and sedimentation (Karrow, 1963; Slay, 1973;
Street, 1979 cited in Tenalem Ayenew, 2004).

1.1 General overview of the study area

The study area (the Rift Valley Basin) covers a total area of 53034 km2 and lies between 04O23’N and 08O28’N
latitude and 36O37’E and 39O22’E longitude (Fig 1).The Ethiopian Rift Valley is part of the Great East African
Rift which is the largest, longest and most conspicuous feature of its kind on earth, stretching for nearly 5,600km
from the Red Sea into the mouth of the Zambezi River in Mozambique. The elevation, width and tectonic setting
of the Ethiopian Rift Valley are extremely variable. The interesting feature of the northern and central sectors is
the existence of open and closed lakes situated within large depressions. The major lakes are located within the
central Main Ethiopian Rift with relatively higher elevations as compared to Afar and Chew Bahir Rift bordering
Kenya. These lakes occupy an enclave of internal drainage basins separating the tributaries of the Nile and
Wabishebele River Basins (Tenalem Ayenew, 2009).

Figure 1. The study area

The Rift Valley of Southern Ethiopia runs NNE from the Kenya frontier of 600Km to the Koka Dam on the
Awash River where the rift begins to open out into the Afar and Danakil depressions (Grove et al., 1975). The
Ethiopian rift valley basin has three physiographic regions: the rift, escarpments and highlands. The rift where
large lakes exist starts in the neighborhood of Lake Abhe and extends some 1000km southwards into northern
Kenya (Tenalem Ayenew, 1998). The altitude ranges from 1600 m a.m.s.l. in the rift to over 4000 m a.m.s.l. in
the large volcanic peaks of the Eastern Highlands. There are highly elevated volcanic mountains both within the
rift and the highlands. The highest elevation is 4245 m.a.s.l, which is the peak of Mount Kaka located on the
eastern boundary of the Ziway-Shala lake basin (Tenalem Ayenew 1998).

47
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

Figure2 .Location map;2:Chamo; 3: Abaya; 4:Hawassa;6:Abiyata;7:Langano; 8:Ziway

2. MATRIALS and METHODS

2.1 Collection and Organization of Data
This study is based on an assessment of existing hydrometeorological records of lake level, rainfall, river
discharge and evaporation for the period of 1980-2007. The rainfall stations used for estimating the areal depth
of the rainfalls corresponding to the lakes were: Arba Minch, Chencha, Hagere Selam and Hossana for Lakes
Abaya and Chamo; Assela, Butajira, Kersa, Langano, Meki, Meraro, and Shashemene for the Lakes Abiyata,
Langano and Ziway; Hawassa and Wondo Genet (incomplete data) for Lake Hawassa. The research is also based
on the interpretation of multi-temporal satellite images acquired during 1986, 2000 and 2010. The data required
for the study have been collected from various sources. Meteorological data for twenty eight years (incomplete
most of the time) have been collected from the Ethiopian Meteorological Agency and the hydrological data for
seven rivers (Bilate, Horakelo, Katar, Kerkersitu, Kulfo , Lipis, Meki and Tikur wuha) and six lakes(Abaya,
Abiyata, Hawassa, Chamo, Langano and Ziway) have been collected from the Ministry of Water Resources.

The lake level records were used to reconstruct the recent changes and to correlate the lake levels with other
catchment hydrometeorological factors. The required softwares (Microsoft Excel, ArcGIS, ERDAS IMAGINE,
SPSS and Matlab) that are anticipated to support the study have been collected from different sources and
utilized. Moreover, Land Sat Satellite images for three different years but the same month have been collected
from Addis Ababa University department of Earth Sciences and other sources. After collecting all the required
hydrologic and meteorological data, it was arranged and reorganized. Some of the required data were averaged
(lake levels) and some of them were summed up (rainfall, discharge and evaporation) and used for the analysis
of the hydrodynamics of the lakes. The average rainfall corresponding to each lake has been calculated and
analysis on the change of lake level and size has been made by looking the historically recorded data of rainfall,
lake levels, river discharges and evaporation rates.

2.2 Correlation of lake water level with precipitation, river discharge and evaporation

To see the significance of the hydroclimatic, geologic and anthropogenic factors on lake water level fluctuations,
correlation with most important hydrometeorological factors (precipitation, river discharge and lake evaporation)

48
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

were made. For lakes with low correlation coefficients, some possible explanations regarding the reason for the
fluctuation of the water levels were presented.

2.3 Multiple regression of lake water level on precipitation, river discharge and open lake evaporation

To identify the relative importance of precipitation, river discharge and evaporation on lake level fluctuation,
regression of lake level on these factors was made. Least squares regression of lake level on rainfall, discharge
and evaporation was made via curve fitting methods (linear, quadratic, cubic, exponential, logarithmic, inverse
etc) .By looking the level of significance (0.01,0.05 and 0.1) in each curve fitting method, the importance of each
factor in lake level fluctuation was determined.

2.4 Trend analysis

After rearranging the hydrometeorological data (lake level, precipitation, river discharge and evaporation rate),
the five year moving averages were calculated and plotted against time. The rainfalls, discharges and open lake
evaporations were also plotted against time. The long-term trend analysis of lake water level, precipitation, river
discharge and lake evaporation rates were made.

2.5 Multi-temporal satellite image analysis

By zooming in the images and then digitizing their boundaries (perimeters), shape files for each lake
corresponding to each year were created. By stacking the layers and then clipping the lake images with the
corresponding shape files, the clipped images for each lake were presented for better visualizations. The areas of
the lakes for each year were then obtained by calculating the areas of the corresponding shape files using the
ArcGIS and ERDAS IMAGINE softwares.

2.6 Land use/cover change analysis

Based on the land use/cover data obtained from FAO (1997) and Ministry of Water Resources (2008), the
change in the proportion of the land cover types to the total area of the Rift Valley Basin has been assessed (only
the end results of the land use land cover classifications done by the above two organizations was used). This is
done to relate the depletion of vegetation with lake level fluctuations.

3. RESULTS and DISCUSSIONS

3.1 Data Consistency
Hydrometeorological data are vital instruments to assess the hydroclimatic and anthropogenic contributions to
the water level fluctuations of the Ethiopian Rift Valley lakes. In spite of this fact, there are problems of data
inconsistency with regard to the hydrometeorogical records. Due to this fact, only similar rainfall patterns
corresponding to each lake were averaged, plotted together with the lake levels and then analyzed.

Figure 1.The average rainfalls and the discharge from Bilate River corresponding to Lake Abaya

49
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

Figure 4. The average rainfalls and the discharge from Kulfo River corresponding to Lake Chamo.

Figure5. The average rainfall for all stations and the discharges corresponding to Lake Abiyata, Langano and
Ziway.

Figures 3,4 and 5 clearly show that the average rainfalls and the discharges do not show similar trends especially
for Lake Chamo indicating data inconsistency. Due to this problem, more emphasis has been given to the
analysis of the satellite data.

Figure 6. Yearly trends of lake level, rainfall and discharge (Lake Abiyata)

50
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

The figure 6 clearly indicates how drastically the level of Lake Abiyata is declining.

Figure 7 . Yearly trends of lake level, rainfall and discharge (Lake Chamo)

It can be clear from figure 7 that the level of Lake Chamo is not associated with rainfall and discharge indicating
inconsistency of hydrometeorological data. Due to this, special emphasis has been given to the analysis of
satellite images for the fluctuation of the level of the lake.

Figure 2. Yearly trends of lake level, rainfall and discharge (Lake Hawassa)

Figure8 shows that the level of Lake Hawassa is largely affected by ground water discharge since the overall
rainfall pattern shows a constant trend and also due to the relatively small correlation coefficient between rainfall
and river discharge.

Table 1. Correlations of rainfalls with the discharges (to check for consistency of hydrometeorological data)

Lake Abaya Abiyata Chamo Hawassa Langano Ziway

Correlation 0.131825 0.379668 - 0.00709 0.319599 0.306506 0.052279

coefficient

51
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

As we can see from Table 1, the associations of rainfalls with the discharges have the smallest correlation
coefficients for Lakes Chamo (unrealistic correlation coefficient), Ziway and Abaya which confirm some
degrees of data inconsistency. As we can see from table 1, the hydrometeorological data inconsistency is much
series especially for Lake Chamo.

3.2 Correlation coefficients and their interpretations

There are always time lags between precipitation events and lake level fluctuations. Moreover, most of the
rainfall stations are located on the Graben Shoulder of the Highlands, only few stations are situated in the Rift
valley (Stefan et al., 2004) and hence the rainfall records can be less realistic. For this reason, poor correlation
coefficients between lake levels and rainfalls have been observed except for Lake Langano with a value of 0.68.
Due to the same reason, the correlation coefficients between lake levels and discharges are also relatively very
small or negligible except for Lakes Hawassa, Langano and Ziway with values 0.76, 0.57 and 0.40 respectively.
There is a significant correlation between water level and evaporation for Lake Abiyata only. The correlation of
lake level with rainfall, discharge and evaporation is summarized in Table 2.

3.3 The significance of the multiple regression equation

Least squares regression equations can be used to predict the values of a dependent variable (lake level in our
case) from one or more independent variables (rainfall, discharge and evaporation) with coefficients of multiple
correlations indicating the strength of the relationships between the dependent and independent variables.
Relatively poor or negligible fits between lake level and rainfall, discharge and evaporation have been observed.
The possible explanation for this may be due to the time lags between precipitation events and lake level
fluctuations or questionable data or both.

3.4 Trend analysis for lake levels, precipitations, discharges and lake evaporations.
Figure 6 shows that the water level of Lake Abiyata is declining drastically while those of Abaya and Hawassa
are slightly rising. This is in agreement with the results obtained using Landsat satellite images. The levels of
Langano and Ziway seem to have constant trends and that of Chamo shows a rising trend. However, the results
obtained using the satellite images show contrasting trends. As shown in table 3, the water level of Lake Abaya
was below its long term average between1980 and 1993, above the long term average between 1994 and 2007.
This could be due to increased deforestation of its catchment as indicated by the land use/cover maps (Figure 13
and 14).
The level of Lake Abiyata showed variable trends between 1980 and 2000, and a constant trend (declining trend)
after 2000 onwards. More specifically, the lake level has been decreased by about 1.6m from its long term
average (about 4.3m) and this shows that very large amount of water has been abstracted from the lake in
addition to the absence of the overflow from Ziway.

Figure9. Five Years’ moving averages showing the lake levels trends

52
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

The level of Lake Hawassa showed declining trends between 1980 and 1993, rising trends after 1994 onwards
which may be due to recent neotectonic activities resulting in ground water inflow into the lake. The levels of
Lakes Chamo and Ziway showed increasing trends between 2001 and 2007. This result contradicts with the one
obtained using Landsat satellite image analysis. The possible explanation for these contrasting results can be data
inconsistency as explained in the previous subsection. The level of Lake Langano is slightly lower than its long
term average indicating relatively less anthropogenic contributions to the lake level fluctuation.

Table 2.The correlation matrix indicating the association of lake levels with rainfall, discharge and evaporation

Lake Rainfall Discharge Evaporation

Abaya 0.08 0.10 -0.77*
Abiyata 0.37 0.29 -0.74*
Hawassa 0.23 0.77* -0.37*
Chamo 0.17 0.43 -0.74*
Langano 0.68* 0.57* -0.01
Ziway 0.26 0.56* -0.1

The level of Lake Abiyata showed variable trends between 1980 and 2000, and a constant trend (declining trend)
after 2000 onwards. More specifically, the lake level has been decreased by about 1.6m from its long term
average (about 4.3m) and this shows that very large amount of water has been abstracted from the lake in
addition to the absence of the inflow from Ziway.
Table 3. Seven years’ water level trends for the studied lakes

No Lake Long term 1980-1986 1987-1993 1994- 2001- Main reason for the
(1980-2007) (average) (average) 2000 2007 fluctuation
average(m) (average) (average)
1 Abaya 0.699261 1.059803 2.613429 1.465826 Increased surface
(Above (Above runoff, diversion
1.45958 (Below average) (Below average) average) average) or high
groundwater
2 Abiyata 5.914493 3.860681 4.648305 2.710621 Large-scale
(Above (Above (Below abstraction of
4.283525 (Below average) average) average) water
3 Hawassa 1.932398 average)
1.272631 1.744318 (Below 2.476688 2.235955 Diversion from
average) (Above ( Above nearby water
(Below average) average) average) bodies,
4 Chamo 1.265267 0.725389(Below 1.743975(Above 1.092512 1.49919 groundwater
ground water
average) average) (Below (Above outflow, diversion
average) average) of feeder rivers,
5 Langano 1.069915 1.389686 1.55971 1.134429 increased and
Diversion
(Below average) (Above average) (Above (Below abstraction of
1.288435 average) average) water
6 Ziway 0.934722 1.064988 (Below 1.244583 1.10928 abstraction of
(Below average) average) (Above (Above water from feeder
1.088393 average) average) rivers for
irrigation

53
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

Table 4. Long-term average annual water balance of the studied lakes (mcm)

Pl: precipitation on the lake; Ri: inflow from rivers; Gi: groundwater inflow; Sr: inflow from surface runoff; El:
lake evaporation; Ro: outflow in river outlets; Go: groundwater outflow; A: abstraction; Ng: negligible; VH:
very high; H: high; M: medium; RO: rare outflow; Mcm: million cubic meter

inflow outflow Difference

Lake Pl Ri Gi Sr El Ro Go A (106m3)
Abaya 556 VH VH VH 1900 RO M - -
Abiyata 113 230 26.8 15 372 closed 1.2 13 -1.4
Hawassa 106 83.1 H 83.7 132 closed 58 Ng +82.8
Chamo 406 H H H 900.9 RO H Ng -
Langano 186 212 135.4 VH 463 46 18.9 - +5.5
Ziway 323 656.5 80.5 48 890 184 14.6 28 -8.6
Source: Tenalem Ayenew (2004)

3.5 Analysis of multi-temporal satellite images of the lakes

The analysis of the Land sat images (with acquisition years: 1986, 2000 and 2010) indicated that Abiyata,
Chamo, Ziway and Langano are declining while Hawassa and Abaya are rising slightly. The decline of Abiyata
is very drastic (nearly 28% of its area in 1986) and the decline of Chamo is about 11% of its area in 1986. Table
5 summarizes the fluctuation in the size (area) of the lakes under consideration.

Table 5.Areas of the lakes (km2) from Land Sat images

Lake Year Change (%)

name b/n 1986 & 2010
1986 (jan-21) 2000 (jan-05) 2010 (jan-
31)
Abaya 1102 1111 1111 +0.8 → Rising Trend
Abiyata 162 164 117 -27.8 → Declining Trend
Hawassa 91 96 94 +3.3 → Rising Trend
Chamo 331 312 295 -10.9 → Declining Trend
Langano 229 228 224 -2.2 → Declining Trend
Ziway 424 420 412 -2.8 → Declining Trend
Note:-acquisition dates are given in the brackets

The reductions in the sizes of Lakes Abiyata and Chamo are shown pictorially in figure 10 and figure 11
respectively. The size of Lake Hawassa has increased by 3.3% of its area in 1986. This is pictorially shown in
Figure 12. The following three figures are created by overlapping the lake images corresponding to 1986, 2000
and 2010.

54
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

Figure 10. The size of Lake Abiyata during 1986, 2000 and 2010

Figure 11. The size of Lake Chamo during 1986, 2000 and 2010

55
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

Figure 12. The size of Lake Hawassa during 1986, 2000 and 2010

3.6 Analysis of land use/cover changes in the Rift Valley Basin

The lakes under consideration are included in the Rift Valley Basin. The major land use/cover types in the basin
are indicated in Table 6. Figures 13 and 14 show the change in the land use/cover of the basin and the
quantitative description is summarized in Table 6. Cultivated land and bareland have been increased by 90% and
75%, vegetation cover and grassland have been decreased by about 80% and 68% respectively between 1997 and
2008. This could imply slight rise in the level of the lakes. However, other factors also contributing to the
fluctuations and the levels are actually declining except for Lakes Abaya and Hawassa.

Figure 13. Land use land cover map of the Rift Valley Basin (FAO, 1997)

56
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

Figure 14. Land use land cover map of the Rift Valley Basin (Ministry of Water Resources, 2008)

Table 6. Change in area (km2) of the land use/cover types for the Rift Valley Basin

Land use type Data Data From Ministry of

No From FAO (1997) Water Resources (2008 )
Change (%)
Value(km2) % Value(km2) %
1 Cultivated land 22938 43.25 43796 82.58 +90.94
2 Vegetation 20772 39.18 4142 7.81 -80.07
3 Grassland 6291 11.86 1596 3.00 -67.76
4 Water body 3023 5.70 2619 4.94 -0.13
5 Bareland 9 0.02 808 1.52 +75
6 Urban area 1 0.002 73 0.14 +69
total 53034 100 53034 100
Table6 shows how fast the vegetation is being depleted in the basin which can results in slight rises in the lake
levels.

3.7 Synthesis
The results of the correlations and least squares regressions indicated that lake level fluctuations are not
associated with rainfall variability for most of the studied lakes (except Langano). The reasons for these weak
associations can be time lags between rainfall events and lake level fluctuations, less realistic lake level or
rainfall data (few rainfall stations available near the lakes), and dominant anthropogenic contributions to lake
level fluctuations among many others.

The associations of the lake levels with the discharges are significant only for Lakes Hawassa, Langano and
Ziway. According to the study made by Tenalem Ayenew (2006), the level of Hawassa is rising (increased
discharge) due to neotectonism and considerable deforestation of the lake basin which in turn resulted in
increased surface runoff. The change in land use in the Rift Valley Basin, as summarized in Table 6, also
confirms this result. The decline in the water level of Lake Ziway can be attributed to the abstraction of water for
irrigation (Tenalem Ayenew et al., 2006; Huib and Herco, 2006) and the decline in the level of Lake Langano
can be related to the decrease in discharge caused by neotectonic activities and diversion of the tributary rivers in
the Arsi Highlands for small scale irrigation (Tenalem Ayenew, 2009) in addition to the slight decline in average
annual rainfalls ( after 1998) as indicated in table 3.

The fluctuation (decline) in the water level of Lake Abiyata is not significantly associated with rainfall and
discharge variability indicating extremely high anthropogenic contributions (large-scale abstraction of water for
soda ash extraction and diversion of the tributary rivers for irrigation). The level of the lake showed a declining
trend between 1980 and 1998 while the rainfall and discharge showed increasing trends during the same period
of time. This clearly shows very significant anthropogenic contributions towards the shrinking of the lake. As

57
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

indicated in Figure 5, the rainfall and the discharge show declining trends between 1998 and 2004. This
situation, together with the large-scale abstraction of water, resulted in serious decline in the level of Lake
Abiyata. Other reasons for the decline may include the increased evaporation rates (although there are no
sufficient evaporation data) and the decrease in overflow from Lake Ziway (Tenalem Ayenew, 2004; Huib and
Herco, 2006).

The decline in the level of Lake Chamo can be connected to the increase in evaporation rates and may be to the
diversion of feeder rivers for irrigation. Another possible reason may be due to groundwater outflow from the
lake to the surrounding aquifers. This reason seems realistic since the lake water is more saline as compared to
that of Lake Abaya.

The slight rise in the level of Lake Abaya can be related to very high groundwater inflow from the surrounding
aquifers (Tenalem Ayenew, 2009) and may be to considerable deforestation of the lake catchment leading to
increased sedimentation and siltation. This is evidenced by the different color in the northern portion of the
image of the lake. The following Table summarizes the most significant factors that led to the lake level
fluctuations.

Table7. Ordering the hydrometeorological, geological and anthropogenic factors

Lake Most significant factor for lake level fluctuation.

Abaya Groundwater inflow, evaporation, sedimentation and siltation (Land use change)
Abiyata Abstraction of water, evaporation, decreased overflow from Lake Ziway, Rainfall
Hawassa Increased discharge due to neotectonic activities(ground water inflow) and land use changes,
evaporation
Chamo Evaporation, diversion of feeder rivers, groundwater outflow.
Langano diversion of tributary rivers, neotectonic activities, rainfall, river discharge
Ziway Abstraction of water, river discharge.

4. CONCLUSIONS and RECOMMENDATIONS

4.1 Conclusions
The results of the study revealed that much of the fluctuations in the lake water levels are caused by human
activities especially for the lakes in the Central Ethiopian Rift (CER). Lakes Abiyata, Chamo, Ziway and
Langano are declining while Abaya and Hawassa are rising. This is also confirmed by the long-term average
annual water balance of the studied lakes as indicated in Table 4, except for Lake Langano with a positive
difference between inflows and outflows. Among the studied lakes, Abiyata is drastically reduced in size
(about 28% of its area in 1986) due to both human activities (most dominant ones) and natural factors. The
other seriously affected lake is Chamo with about 11% reduction in its area between 1986 and 2010.
Lake Abaya was found to be relatively stable during the indicated period (showed only a 0.8% increase in its
area).Groundwater flow into Lake Abaya is very high and the flow into Lake Chamo is relatively low. Regarding
Lake Abaya, surface water and groundwater flow into the lake exceeds the increase in evaporation rates plus the
effects of anthropogenic factors (the variation of rainfall and discharge being negligible) leading to the slight rise
in the level of the lake. As to Lake Chamo, the increases in the lake evaporation rates plus abstraction of water
dominate the groundwater inflow which resulted in the decreasing trend of the level of the lake.

Lake Hawassa is rising due to the increased discharge and slight decline in evaporation. About 59% of the
variation in the lake level is due to the variation in the discharge and about 14% of the variation is due to the
variation in evaporation rates.

The water level of Lake Abiyata is declining drastically due to large scale-water abstraction, decreased overflow
from Ziway and increased evaporation. The significant association of the level of Lake Abiyata with evaporation
seems to be realistic since the lake is closed topographically and is located at a terminal position in the CER.

The size of Lake Ziway is also reduced slightly due to abstraction of water for irrigation and the variation in
discharge (about 16% of the lake level fluctuation is due to the variation in the river discharge). The lake level
does not associate significantly to rainfall and evaporation indicating considerable anthropogenic contributions
to the level fluctuation.

The size of Lake Langano is reduced slightly. The lake level is associated with rainfall and river discharge.
About 61% of the lake level fluctuation is caused by the variation in both rainfall and discharge. This

58
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

comparatively large value could imply relatively less (as compared to the other lakes in the CER) man-induced
contributions to the fluctuation in the level of Lake Langano.

The lakes under consideration in order of decreasing magnitude of the degree of fluctuation are: Abiyata,
Chamo, Hawassa, Ziway, Langano and Abaya.

4.2 Recommendations
Concerned bodies need to work in harmony to manage water use in the Ethiopian Rift Valley lakes,
particularly Abiyata and Chamo.

Since Lake Abiyata is seriously affected, special attention is needed to replenish the water in the lake
by controlling the use of water and save the lives of many endemic animals including the flora and
fauna.

Further studies need to be made on the Ethiopian Rift Lakes to accurately identify the prevailing
factors that are causing lake water level fluctuations, especially in the Lake Abaya-Chamo Basin.

Educations on the consequences of excess water use and on how to protect the water resources need to
be given to some representatives of the community living around the lakes.

AKNOWLEDGEMENTS

I am grateful to the almighty God, the merciful and compassionate that helped me in all endeavors of my life.
My second deepest acknowledgement goes to my advisor Prof. Tenalem Ayenew for his effective and
intellectual guidance, consistent follow-up, supervision, provision of some necessary materials for the research,
encouragement and support throughout the work devoting his valuable time and effort .He was also always
approachable and patient with me and gave me constructive suggestions and confidence towards finalizing this
thesis work.

I would also like to thank Dr. Seifu Kebede( lecturer at Addis Ababa University, Department of Earth Sciences)
for his constructive suggestions, provision of relevant books and encouragement.

My special thanks go to the Ministry of Water Resources and Ethiopian Meteorological Agency for providing
the data and allowing me to use their libraries. Many special thanks to the GIS and Hydrology units in the
Ministry of Water Resources.

Finally, I would like to thank my family, relatives, friends and colleagues who contributed helpful advice and
assistance in providing some materials for accomplishing this thesis.

REFERENCES
Alexander, S. and Habteselassie Redeat (2010). Fish faunas of the Chamo-Abaya and Chew Bahir basins in
southern portion of the Ethiopian Rift Valley: origin and prospects for survival, Aquatic Ecosystem Health &
Management. Russian Academy of Sciences, Institute of Ecology and Evolution, Moscow.
Azeb Belete (2009). Climate Change Impact on Lake Abaya Water Level. Unpublished MSc. Thesis, Addis
Ababa University, Addis Ababa.
Ebrahim Esa (2008). Analysis of the Impacts of Soda Ash Mining on Abiyata and Shalla Lakes Subwatershed:
The Main Ethiopian Rift System. Unpublished MSc. Thesis, Addis Ababa University, Addis Ababa.
Grove, A.T., Alayne, F.S. and Goudie, A.S.(1975). Former Lake levels and climatic changes in the rift valley of
Southern Ethiopia. The Geographical Journal, 141:177-194. Blackwell Publishing, England.
Heathwaite, A. L. and Burt, T.P. (1991). Predicting the Effect of Land Use on Stream Water Quality in the Uk.
University of Sheffield, Department of Geography, Uk.
Huib, H. and Herco, J. (2006). Agricultural development in the Central Ethiopian Rift valley: A desk-study on
water-related issues and knowledge to support a policy dialogue. Plant Research International B.V.,
Wageningen, the Netherlands. Keddy, P.A. and Reznicek, A.A.(1986). Great Lakes Vegetation Dynamics:
The Role of Fluctuating Water Levels and Buried Seeds. Department of Biology, University of Ottawa,
Ottawa.

59
Civil and Environmental Research www.iiste.org
ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online)
Vol.7, No.12, 2015

Kinshiro, N. (1974). Hydrological Regime on the Regulation o f Lake-level, A Contribution to the

Paleolimnological Study of Lake Biwa. Department of Geophysics, faculty of Science, Hokkaido University,
Hokkaido.
Murimi, S.K.(1994). Falling Water-levels in Saline Lakes of the Central Rift Valley of Kenya: the case of Lake
Elmenteita. Geography Department, Kenyatta University, Nairobi. International Journal of Salt Lake
Research 3:65-74.
Niko, P. and Bardhyl, A.(2009). A Method to Estimate Buna River Discharge, Albania. Tirana University,
Faculty of History, Philology and Geography, Tirana.
Qin, B.(1993).The Hydrological Characteristics and Water Budget of Qinghai Lake Drainage Basin, China.
Nanjing Institute of Geography and Limnology, Academia, Sinica, Nanjing.
Ramsis, S., Elise, B., Daniel, P., Lorraine, B., John, B., Wayne, H., Gerald, W. and Greg, B.(2002). Hydrological
Response Units of the Ord Stage I Irrigation Area and the Dynamic Filling of the Aquifers of the Ivanhoe and
Packsaddle Plains. Technical Report 7/02, CSIRO Land and Water Private Bag No 5, Western Australia.
Shewadeg Molla (2008). Decision Support Tool for Estimating Lake Water Balance with Particular Emphasis to
Lake Abiyata. MSc Thesis, Addis Ababa University, Addis Ababa, Ethiopia.
Sousounis, P.(2000). The Potential Impacts of Global Warming on the Great Lakes Region. University of
Michigan and Patty Glick (National Wildlife Federation), Michigan.
Stefan, T. and Gerd, F. (2004). Water Resources Assessment in the Bilate River Catchment-Precipitation
Variability. Lake Abaya Research Symposium 2004-Proceedings.
Susanne, B. and Brigitta, S. (2004).Character of Lake Floor Sediments From Central Lake Abaya, South
Ethiopia. Lake Abaya Research Symposium-2004 Proceedings.
Tang, Z., Engel, B.A., Pijanowski, B.C. and Lim, K.J. (2004). Forecasting land use change and its environmental
impact at a watershed scale. Journal of Environmental Management, Department of Agricultural and
Biological Engineering, Purdue University, Purdue.
Tenalem Ayenew (1998). The hydrogeological system of the Lake District basin: Central Main Ethiopian Rift.
Ph.D. Dissertation, Free University of Amsterdam, Amsterdam.
-----------------.(2002). Recent changes in the level of Lake Abiyata, central main Ethiopian Rift. Department of
Geology and Geophysics, Addis Ababa University, Addis Ababa.
-----------------.(2004).Environmental implications of changes in the levels of lakes in the Ethiopian Rift since
1970. Department of Geology and Geophysics, Addis Ababa University, Addis Ababa.
-------------------.(2006). Vulnerability of selected Ethiopian lakes to climatic variability, tectonism and water use.
Global Change System for Analysis, Research and Training.
------------------. (2009).Natural Lakes Of Ethiopia. Addis Ababa University Press, Addis Ababa.
Tenalem Ayenew and Yemene Gebreegziabher (2006). Application of a spreadsheet hydrological model for
computing the long-term water balance of Lake Hawassa, Ethiopia. Department of Earth Sciences, Addis
Ababa University, Addis Ababa.
Thompson, F. and Turk, J. (2005).Earth Science and the Environment, 3rd edition. Books/Cole, a division of
Thompson Learning Inc., 10 Davis Drive, Belmont.

60