1

Evaluation of Climate Change Impacts on the Blue Nile flows using SWAT

Imtiaz Bashir1, Jiri Nossent

1, Willy Bauwens

1, Okke Batelaan

1,2

[1] {Earth System Sciences Group and Dept. of Hydrology and Hydraulic Engineering,

Vrije Universiteit Brussel (VUB), 1050 Brussels, Belgium}

[2] {Dept. of Earth and Environmental Sciences, Katholieke Universiteit Leuven, Celestijnenlaan

200E, 3001 Heverlee, Belgium}

Correspondence to: Imtiaz Bashir (imtiaz.bashir@vub.ac.be)

Abstract

This contribution aims at the quantitative assessment of the impacts of climate change on the water

resources and hydrological flow pattern of the Blue Nile River basin using hydrological modeling

technique. The Blue Nile River contributes around 60% of the total mean annual flow of the Nile,

and most of this flow is generated during the flooding season (July – October) by high precipitation

received in the Upper Blue Nile catchment. Large irrigation schemes in Sudan and Egypt are mainly

dependent on the flooding in the Blue Nile, which get water from several reservoirs built along the

system. However, recent observations show that the climate has changed at an unprecedented rate

during the last decade, with an evident increase in extreme precipitation events and temperature in

eastern parts of Africa. This increases the vulnerability of downstream located countries for

probable effects on their widespread agricultural based economic activities.

This research analyzed the climate change impacts on the Blue Nile River basin by the year 2025,

through use of the hydrological model SWAT, with spatial maps and synthetic weather data input.

The model was calibrated at a monthly time scale against observed discharge series of four gauging

stations of the Blue Nile. The climate change scenarios were constructed using outcomes of two

General Circulation Models (MIROC and INMCM) for emission scenarios A2 and A1B, by adjusting

the baseline climatic variables that represents the current precipitation and temperature patterns.

The constructed climate scenarios were applied to the calibrated and validated hydrological model,

to generate runoff and investigate climate change impacts on water yield and hydrological flow

patterns.

The results show that the climate is likely to become wetter and warmer in 2025 (2010-2039) in most

of the Blue Nile basin; annual water yield is expected to increase at basin scale and low flows

become higher, while, the outflows from the Rosieres reservoir show substantial increase in

magnitude under combined pressure of climate change and loss of the reservoir storage due to

sedimentation. However, uncertainties on the results are high on account of the quality of the

baseline climate data, the discharge series and the accuracy of climate models. Overall, the study

suggests that the water resources of the Blue Nile River basin will not be adversely affected by

climate change depending on future actions. The increase in precipitation and resulting water

resources may help to meet future water needs in the region.

Key words: SWAT, climate change, hydrological model, the Blue Nile, water yield.

2

1. Introduction

A greater tendency for maximizing the utilization of transboundary freshwater resources within

territorial limits resulted in instigating and/or escalating conflicts among communities in almost

every region of the world. Climate change, population growth, demographic changes and economic

development are leading factors producing pronounced quantitative and qualitative impacts on

transboundary water resources. Scientists and researchers are on high alerts, particularly, to

anticipate natural variability and uncertainty in weather patterns, and to human induced climate

changes. Consequently, climate change related studies emerged on the top of research agenda

worldwide (IPCC, 2007). The Fourth Assessment Report by the Intergovernmental Panel on Climate

Change, previous assessment reports and related documents (IPCC, 2001 and 2007) present evidence

of global climate change, with particular focus on issues likely to be faced by the water resources

managers. According to the IPCC, the climate change has the potential to affect many sectors in

which water resource managers play active role, including water availability and quality, flood risk

reduction, coastal areas, navigation and hydropower.

Scientists and researchers argue that the future changes in overall magnitude and variability of the

river flows, and the timing of the peak flow events are amongst the frequently debated hydrologic

issues (Frederick 2002; Wurbs et al., 2005). Observed records of climatic and hydrological variables

during the twentieth century, show that many natural systems are being greatly affected by the

regional climate changes, particularly, by the increase in temperature and the variation of

precipitation events. The increasing temperature is expected to change the mix of precipitation

toward more rain and less snow. Regional shifts in precipitation are expected, with some areas

receiving more and others receiving less. Changes in seasonal flow patterns and precipitation

extremes may lead to enhanced chances of droughts and floods occurrences. These are likely to

occur more frequently and/or be more severe under future climatic conditions (IPCC, 2007). Similar

trends can be followed in the African continent reporting increase in the frequency of heavy

precipitation events over most of the land areas and widespread changes in extreme temperatures

over the past 50 years. Recent trends show a tendency towards greater extremes: arid or semi-arid

areas in northern, western, eastern and parts of southern Africa are becoming steadily drier, and

increased variability of precipitation and storms is observed.

A typical example of a transboundary river basin in Africa is the Nile River basin, which is the

longest river basin in the world, draining almost 1/10th

of the whole continent. It comprises of two

major tributaries: the White Nile and Blue Nile. The latter is the source of most of the water and

fertile soil in the Nile basin area. Historical footprints along the river banks in Egypt, Ethiopia and

Sudan show established water uses of the Nile River since ancient times and indicate its importance

for agricultural, economical (Whittington, 2005), social and cultural aspects of human life. Most of

the parts of the Nile River basin are found sensitive to climatic variations (Conway and Hulme 1996;

Conway 2005; Kim et al 2008). Researchers who examined the impacts of climate change on water

resources in the Nile Basin include Gleick (1991); Conway and Hulme (1996); Strzepek and Yates

(1996); Yates and Strzepek (1996, 1998a,b); Sene et al. (2001); Conway (2005). They primarily

focus on the changes in runoff and their consequences on the economies of riparian countries. Rising

conflicts on account of increasing water demands by the riparian countries have asserted great

realization upon water resources managers to investigate further likely climate impacts for

sustainable river basin management to assure peace and tranquility in the region.

The Blue Nile, major tributary of the Nile River, drains around 10% of the entire Nile basin area in

Ethiopia and Sudan, but its flow is of vital importance for downstream usage in Sudan and Egypt.

The river flows contribute around 60% of total mean annual flow measured at High Aswan Dam

3

(Waterbury 1979; Sutcliffe and Parks 1999; Conway 2005). The long-term variability of the Blue

Nile flows become of particular interest for the riparian countries. Especially to Ethiopia, wherein

most of the Blue Nile flow is generated, but it doesn’t have any established rights on the use of river

water. People have small scale irrigated and/or rain-fed agriculture based livelihood in this area. On

the other hand, the agricultural based economies of Sudan and Egypt is greatly dependent on the

flow generation in Ethiopia. Therefore, efforts for understanding changes in temporal and spatial

patterns of stream flows on account of future climate changes are crucial for food security in the

region (NMSA, 2001; MOWR, 2002; UNESCO, 2004; Conway, 2005). Elshamy et al. (2009)

assessed the water balance of Blue Nile for future climate using outputs of 17 General Circulation

Models (GCMs) and reported moisture constrained Blue Nile by 2100. Similarly, Kim et al. (2008)

used outputs of six GCMs to assess climate impacts on hydrology and water resources of the Upper

Blue Nile river basin and reported an increase of low flows.

The Soil and Water Assessment Tool (SWAT) has been widely used for evaluating the influence of

climate variability on watersheds (Neitsch et al., 2005). This is done in the form of sensitivity

analysis where baseline climatic conditions are used to simulate stream flow and then compared with

simulation for the future changes in precipitation, temperature and other climate variables. This

analysis provides information on the direction and the change of magnitude of the stream flow and

gives insight into which variables are most significant in predicting these changes. Stonefelt et al.

(2000) and Fontaine et al. (2001) applied SWAT to assess climate change impacts on annual water

yield and stream flow predictions of some river basins in North America, and used arbitrary changes

in weather inputs based on GCM/RCM projections. The study concluded that coupled variable

analyses represent a more realistic climate change regime and reflect the combined response of the

basin. Githui et al. (2008) used SWAT to assess future climate impacts on the Nzoia River in Kenya

and reported higher expected stream flows in the coming decade. Other researchers like Van Liew

and Garbrecht (2001) and Varanou et al. (2002) used SWAT to investigate the impacts of climate

change on the water yield.

Above all demonstrate the wide application of SWAT hydrological model for investigating the

climate change impact assessment on hydrology in different parts of the world; however, few

researchers applied SWAT for climate change impact assessment in the Blue Nile River basin. Most

of the researchers applied conceptual hydrological and stochastic modeling techniques to smaller

river catchments of the Blue Nile River basins. Researchers who used physically based SWAT

hydrological model; they focused their motivation either on developing and calibrating a

hydrological model or studying sediment transportation in the Upper Blue Nile River basin

seemingly for its hydrological importance to the river flows. In this study, we developed a SWAT

hydrological model for the entire Blue Nile river basin to assess climate change impacts by the year

2025. The simplest application of the direct use of nearest GCMs grid output to the considered

weather station was adopted in this study. It is one of the IPCC proposed methods to apply the GCM

outcomes to regional level studies (IPCC, 1999). We were able to incorporate predictions of future

climate variability of the GCMs using this method, and found it reasonably workable in bigger

catchments for climate change impact assessment.

2. Study area

This study covers the entire Blue Nile river basin including Lake Tana and its drainage area. The

Blue Nile river basin has a drainage area of about 310,000 km2 just upstream of Khartoum (Sudan),

where it eventually joins the White Nile. This area covers most of Ethiopia and part of Sudan

between longitude 32oE and 40

oE and latitude 9

oN and 16

oN, shown in Figure 1.

4

$

$

$

0 300 Kilometers

Elevation

374 - 803

804 - 1232

1233 - 1661

1662 - 2090

2091 - 2519

2520 - 2948

2949 - 3377

3378 - 3806

3807 - 4236

No Data

Streams

$ Lakes / ReservoirsN

Figure 1- The Blue Nile River in Africa and representation of topography in the basin

2.1 Topography, landuse and soil

The Blue Nile and all of its tributaries originate in Ethiopian Plateau at an elevation of 2000 m to

3000 m amsl. About 35 km downstream of Lake Tana exit, the river drops approximately 50 m into

Tis Issat Falls and starts flowing in a deep gorge about 1200

m below the ground level through the plateau. The Blue Nile

emerges from the plateau close to the western border of

Ethiopia, where it turns in north-west direction and enters

Sudan at an altitude of

490 m amsl. Just

before crossing the

frontier in Sudan, the

river enters a clay

plain where it flows to

Khartoum to

eventually merge into

the White Nile to form the Nile River. Downstream of

Rosieres, the Blue Nile is a mild stream with a slope of about

0.12x10-3

, 1/10th

the slope of the torrential stream which

prevails from the Lake Tana exit to Rosieres, shown in Figure

2.

Most of the Ethiopian Plateau is hilly with grassy downs,

swamp valleys, and scattered trees (Shahin, M., 1985). Figure

3 shows the dominant landuse/landcover type in the basin.

Figure 2: Longitudinal Profile of Blue Nile River

0 300 Kilometers

Landuse / Landcover classes

Barren or Sparsely Vegetated

Dryland Cropland and Pasture

Cropland/Grassland Mosaic

Cropland/ Woodland Mosaic

Deciduous Broadleaf Forest

Evergreen Broadleaf Forest

Mixed Forest

Grassland

Savanna

Shrubland

Residential - Medium Density

Water Body

StreamsN

Figure 3: Spatial representation of major

Landuse/landcover types in the area

Kessie

Rosieres

Sennar

Khartoum

Lake Tana

5

Savanna covers around 71% of the basin area with concentrated growth in central part, and

diminishing trends of this can be noticed towards the basin outlet at Khartoum. Other main land

cover types are dryland cropland and pasture; grassland; and cropland/woodland mosaic covering

18%, 4% and 3% respectively of the basin area. The basin area is generally characterized by clayey

soils, which covers about 52% of the basin area. Table 1 tabulates top seven dominant soil classes

according to FAO-UNESCO soils classification along with respective soil composition in 1st and 2

nd

layers of 300 mm and 1000 mm thickness, respectively.

Table 1: Dominant soil types in the Blue Nile River basin

FAO classification

soil ID

Classification % area

coverage

1st layer composition 2

nd layer composition

Clay Silt Sand Clay Silt Sand

Vc36-3a#277 Clay 26 49 28 22 54 26 20

Be9-3c#26 Clay 11 43 28 29 46 28 26

Bh12-3c#31 Clay loam 9 36 27 37 37 29 35

Ne12-3b#156 Clay 9 43 31 26 54 25 21

Re59-2c#246 Loam 7 22 31 47 22 28 49

Ne13-3b#158 Clay 6 47 26 27 47 26 27

Be51-2a#22 Clay loam 5 35 36 30 35 36 28

Qc2-1bc#176 Sandy loam 5 12 19 69 12 14 72

2.2 Climate

The climate of the Blue Nile basin varies from humid in Ethiopian Plateau to semi arid in the north

of Sudan. Climate in the highlands is strongly influenced by the effects of elevation and is generally

temperate at higher elevations and tropical at lower elevations. The southern parts in the highlands of

Ethiopia experience heavy rain (more than 1520 mm during the summer). Tropical climate with

well-distributed rainfall is found in parts of the Lake Tana region and southwestern Ethiopia. About

70% of the annual precipitation falls during the period between June to September (Kim et al. 2008).

In the Ethiopian Plateau, there is little variation in the mean daily temperature throughout the year,

which ranges from 14°C to 27°C depending on locality and altitude. Annual mean daily relative

humidity is about 68% on the average, which also varies with altitude.

Similar climatic conditions prevail in the Southern parts of Sudan, which receive around 1270 mm

rain over nine-month period between March and November, with the maximum occurring in August

(Shahin, M., 1985). Maximum and minimum temperatures are recorded between March and June,

and December or January respectively, while the annual average relative humidity remains around

55%. The rainy season, which occurs in the South from March to November, gets shortened and

confined to July and August in the Northern part of central Sudan with around 250 mm rainfall

annually and average annual relative humidity of 31%. The minimum temperature occurs in January

and the maximum in May or June, when it rises to a daily average of 37°C in Khartoum. The annual

free surface water evaporation varies from 3.0 mm/day for Lake Tana to 6.5 mm/day in Sennar and

eventually to 7.8 mm/day around Khartoum. The potential evapotranspiration, based on Penman’s

method, in the basin varies from 1150 mm around Lake Tana to 2000-2100 mm around Rosieres and

2300-2400 mm around Khartoum (Shahin, M., 1985).

6

2.3 Hydrology

The catchment area of Lake Tana is around 14,500 km2 and its own surface area ranges from 3,000

to 3,500 km² depending on the season and rainfall. Because of the large storage capacity and the

regulation at its outlet, the peak outflows occur in September, two months after the maximum rainfall

in July (USBR, 1964b). Hurst et al. (1950) estimated the flow at station Kilo330 downstream of lake

exit, about 2.2x106 m

3/day during the low stage period and about 220x10

6 m

3/day in the flood

season, which is about 10 times as much as its initial value at lake exit. About 935 km below Lake

Tana exit, the river discharge at Rosieres is about 7x106 m

3/day in the low stage period, which

indicates reach contribution between Kilo330 and Rosieres. The average flow volume at Rosieres for

the period 1912-1973 is 49.2x109 m

3/year with a peak flow in August, and that at Sennar is 47.2x10

9

m3/year. Total transmission losses in the reach are around 2x10

9 m

3/year mainly accounted for

evaporation in the reach and from the reservoir (Shahin, M., 1985).

Usually little additional runoff is generated in the North of Rosieres except for the two tributaries,

the Dinder and the Rahad. Both the tributaries join the Blue Nile downstream of Sennar close to Wad

Medani and also have their headwaters in the Ethiopian Highlands. Both the rivers hardly carry any

discharge in the period from January to May, having a triangular shape of the hydrograph with a base

width of about 200 days. The peak discharges in Dinder and Rahad have been found to be around

480 m3/sec and 160 m

3/sec, respectively. The annual flow volume in the Blue Nile River at

Khartoum is nearly 50.4x109 m

3/year that is around 8% higher than at Sennar with estimated

transmission losses of 0.85x109

m3/year mainly due to evaporation and seepage (Shahin, M., 1985).

However, Hurst et al. (1950), estimated that the average flow volume at Sennar remain equal or

larger to that of Khartoum.

2.4 Lakes and reservoirs and irrigation schemes

Lake Tana, a natural lake, and two manmade reservoirs are situated in the Blue Nile River basin. The

manmade reservoirs, located at Rosieres and Sennar, were completed in 1966 and 1925 respectively,

for irrigation and power generation purposes. The Gezira Scheme, Sudan’s oldest and largest gravity

irrigation scheme, was progressively expanded since its inception in 1925. It receives water from

Sennar reservoir with present coverage of about 880,000 hectares divided into 114,000 tenancies

(Coutsoukis, P., 2004). The increase of Sudan’s share from the Nile water, under the 1959 Nile

Waters Agreement with Egypt, led to the construction of the Managil extension of the Gezira

Scheme, situated between the White and the Blue Nile rivers upstream of Khartoum. Major existing

irrigation schemes, largely dependent on the Blue Nile River flows, have been listed in Table 2.

Table 2. List of Irrigation Schemes in the Blue Nile River basin (FAO Aquastat Programme, 2007)

Irrigation Scheme Year of Operation Water with-drawal location Area (hec.) Main Crops *

Gezira and Managil 1931 and 1963 Sennar 874,300 C,G,W,S

Rahad 1977 (1st supply)

1981 (80% area)

1983 (entire area)

Rosieres 126,000 C,G,S

Blue Nile Agri. Pro. Corp. 1970s Sennar 122,640 C

El Suki Early 1970s Sennar 36,500 C,G,S

Guneid Sugar Co. 1955 Sennar 16,250 Sr.C

Guneid Extension Mid 1970s Sennar 19,070 C,G

* C: Cotton; G: Groundnut; S: Sorghum; W:Wheat; Sr.C: Sugarcane

7

Sudan was expected to become the bread basket of the Arab World in the 1970s, and irrigation

schemes such as the Rahad Scheme were established with large investments from oil-rich Gulf

nations (Coutsoukis, P., 2004). Large-scale irrigated agriculture expanded in Sudan from 1.17

million hectare in 1956 to more than 1.73 million hectare by 1977.

3. Methodology

3.1 Input data

SWAT is a physically based semi-distributed hydrological model, and it needs spatial topographical

and meteorological data as main input. This input data, which was used to develop hydrological

model of the Blue Nile river basins, can be grouped into four major categories:

I. Spatial topographical information at 1 Km spatial resolution comprising of a Digital Elevation

Model (DEM), a Digital Stream Network (DSN) and landuse/lancover maps were obtained

from the Earth Resources Observation Systems Data Center (EROSDC) of USGS, South

Dakota - USA. The soil map was obtained from FAO-UNESCO, that provides soil

characteristics of about 5000 soil types at a spatial resolution of 10 km in two layers; a 1st layer

from 0-300 mm and a 2nd

layer from 300-1300 mm. Other soil properties concerning particle

size distribution, bulk density, organic carbon content, available water capacity and saturated

hydraulic conductivity, were obtained from Reynolds et al. (1999).

II. Hydro-meteorological information comprised of monthly discharge data series at four

measurement stations shown in Figure 1 and daily time series of rainfall, temperature, wind

speed, solar radiation and relative humidity. The monthly river discharge data was obtained

from the Global Runoff Data Centre, Koblenz - Germany. Daily time series of rainfall (mm);

and maximum and minimum temperature (oC) for the period 1971-1995, were obtained from

the Swiss Federal Institute of Aquatic Science and Technology – EAWAG (Schuol et al.,

2005). The data set contains generated information from dGen for 200 synthetic stations

distributed over the entire Nile basin, out of which 20 stations are located in the drainage area

of the Blue Nile river basin. Monthly statistics of wind speed, relative humidity and solar

radiation were derived using CropWat 8.0 freely available from Food and Agriculture

Organization (FAO), Rome, Italy.

III. Annual irrigated area and water withdrawals were taken from FAO Aquastat Programme

(2007), while lake/reservoirs characteristics and monthly irrigated areas were taken from a

study conducted by van der Krogt (2005).

IV. Crop and soil characteristics in SWAT database files were updated for local conditions in the

study area. Concerned parameterization of the landuse classes e.g. leaf area index, maximum

stomata conductance, maximum root depth, optimal and minimum temperature for plant

growth were based on literature and the available SWAT landuse classes. While, the

characteristics/conditions of the urban areas and fertilizers assumed to have been similar to

those in United States given in SWAT database files.

3.2 Quality control

A quality control was performed for checking the homogeneity of the available hydro-meteorological

data at 95% significance level using tool RAINBOW. This tool performs frequency analysis and test

the homogeneity of datasets (Raes et al., 2006). Average annual and monthly values of precipitation,

maximum and minimum temperature, and discharge series were tested at four locations: Kessie,

Rosieres, Sennar and Khartoum. Generally, the hydrological data and precipitation records meet the

data homogeneity, however, a number of temperature stations failed homogeneity test for average

annual values.

8

3.3 Future climate scenarios

IPCC and researchers suggest that General Circulation Models (GCMs) are reliable scientific

platforms to simulate the impacts of increased greenhouse gases on climatic variables (Conway and

Hulme, 1996; IPCC, 1999; Dibike and Coulibaly, 2005; Wilby and Harris, 2006). Outcomes of these

GCMs are commonly forced as input to the hydrologic models to project changes in water resources

(Nash and Gleick, 1993; Arnell and Reynard, 1996; Conway and Hulme, 1996; Yates and Strzepek,

1998; Kim et al., 2004; Wilby and Harris, 2006). Adopting this methodology, the hydrological

models are initially calibrated for the current climatic data, followed by defining future climatic data

either by generating future climatic variables or perturbing the current climatic data and finally

simulating the scenarios to compare the simulated results with those under present climate. Several

researchers investigated the impacts of climate change on the Nile River basin using similar

methodology (Conway and Hulme, 1996; Yates and Strzepek, 1998; NMSA, 2001). However, coarse

gridded outcomes of these GCMs remain a major limitation in applying them to small study areas.

IPCC proposed several methods to apply the GCM outcomes to regional level studies (IPCC, 1999).

The simplest application is the direct use of the original grid information nearest to the considered

measurement station, which has been adopted in this study.

According to IPCC (2007), no driver other than GHGs provides a scientifically sound explanation of

most of the warming and changes in precipitation patterns observed both globally and regionally

over the past few decades. IPCC has identified five major story lines of GHGs emission scenarios,

out of which A2 and A1B were selected for this study. IPCC establish these emission scenarios as

follows:

A2 storyline and scenario family describes a very heterogeneous world. The underlying

theme is self-reliance and preservation of local identities. Fertility patterns across regions

converge very slowly, which results in high population growth. Economic development is

primarily regionally oriented and per capita economic growth and technological changes are

more fragmented and slower than in other storylines.

A1B scenario describes a future world of very rapid economic growth, low population

growth, and the rapid introduction of new and more efficient technologies. Major underlying

themes are convergence among regions, capacity building, and increased cultural and social

interactions, with a substantial reduction in regional differences in per capita income. This

scenario family develops into four groups that describe alternative directions of technological

change in the energy system.

Main reasons for selecting A2 scenario are prevailing and expected high population growth rates and

fight for survival of regional identities in the study area, which fit perfectly in the framework of this

emission scenario. A1B scenario was selected to pose an optimistic approach for a prospering and

conflict free region, aiming at low population growth rate. GCMs outcomes can be obtained either as

the estimates of average annual and monthly precipitation and temperature for the future climate, or

as relative change in climate variable relative to baseline period 1961-1990. Change in precipitation

and temperature fields imposed on the baseline time series is one of the approaches to study climate

change impact assessment on the water resources (IPCC, 2001). Based on selected emission

scenarios and spatial resolution of GCMs, outputs of two models MIROC3.2 and INMCM3.0 were

selected. These models have been developed by the research centers CCSR/NIES/FRCGC (Japan)

and Institute of Numerical Mathematics (Russian Academy of Science, Russia) respectively.

The GCMs outcome of monthly relative change in the baseline precipitation and temperature fields

predicted for the period 2010-2039 were obtained from IPCC Data Distribution Centre. These mean

9

monthly relative changes were applied to baseline climatic records on pro-rata basis for generating

averaged future climate by the year 2025. It helped to preserve temporal and spatial correlation of

climatic events in the baseline period for the future climate. The impacts of these outputs on annual

precipitation, and maximum and minimum temperature were averaged over various reaches of the

study area and are shown in Figure 5. Generally, both the GCMs show a considerable increase in

precipitation by the year 2025 in the basin except MIROC-A2 for Sennar and Khartoum reaches,

where a decrease in average annual precipitation is projected. Maximum and minimum daily

temperature predictions by both the GCMs are consistent, and an increasing trend can be observed.

However, model INMCM predicts a remarkable increase in minimum temperature for both the

emission scenarios.

Figure 5: GCMs predictions for average annual change in precipitation and maximum and minimum temperature

by the year 2025 in the Blue Nile River Basin

Average monthly relative change in climatic variables are shown for both Rosieres and Khartoum in

Figure 6, that explains better understanding of climate impacts on the hydrological processes of the

basin.

(a) Average monthly precipitation change in Rosieres reach (b) Average monthly precipitation change in Khartoum reach

10

(c) Average monthly maximum temperature change in

Rosieres reach

(d) Average monthly maximum temperature change in

Khartoum reach

(e) Average monthly minimum temperature change in

Rosieres reach

(f) Average monthly minimum temperature change in

Khartoum reach

Figure 6: GCMs prediction for relative monthly changes in average precipitation, minimum and maximum temperature

averaged over Rosieres and Khartoum reaches

Most of the precipitation that falls in the Ethiopian highlands from July to October, brings about 80%

flows in the Blue Nile annually, therefore, small percentage change shown during this period means

significant impact on the river flows. An increasing trend in precipitation during the wet season in

the drainage area of Rosieres is predicted, except for MIROC-A2, which shows a decrease in

precipitation in the later months of wet season. However, relatively consistent outputs are predicted

for the wet season in the drainage area of Khartoum by both the models. During the dry period,

INMCM gives relatively higher precipitation increase in Rosieres and Khartoum areas than MIROC.

Maximum and minimum temperature in Rosieres and Khartoum drainage areas are generally

expected to increase by both the models, however, relatively smaller increase is predicted than

INMCM.

3.4 Model development

3.4.1 Model build-up

SWAT model for the Blue Nile River basin was developed using AVSWATX - SWAT2005. The

model buildup primarily consists of necessary steps of automatic watershed delineation,

implementation of reservoirs, irrigation water withdrawals and weather data input. Considering the

data availability and application of the model for climate change impact assessment, the existing

11

reservoirs were implemented using specified water release rates, while, the irrigation water

withdrawals from Rosieres and Sennar were implemented as consumptive water uses. SWAT

calculates evaporation separately from upland and reservoir processes. Therefore, it is accounted

twice in case of reservoir implementation and needs to be ignored either of the two. Amongst three

existing reservoirs implemented in the model, Rosieres and Sennar being smaller in the surface area,

were ignored during the HRUs definition process by setting threshold values of 4% and 7% for land

use and soil classes respectively. Lake Tana has large surface area and needed special pre-processing

of topographical data to enclose the entire lake in a sub-basin for ignoring evaporation losses from

the upland processes. For this purpose, a threshold area of 303,947 hactares was chosen to form an

isolated sub-basin preserving the original inlet and outlets of the lake.

Stream outlets were selected based on three factors: i) Inlets and outlet of Lake Tana retained as per

reality, ii) Delineation of sub-basins to contain atleast one precipitation and temperature gauge in

each subbasin iii) Saving simulation results at discharge measurement stations of Kessie, Rosieres,

Sennar and Khartoum. The model was simulated from 1971 to 1982, with initial two years set to

warm-up period. The calibration and validation periods were set: 1973 to 1978 and 1979 to 1982,

respectively. Muskingum method, Penmen Monteith equation and CN method were used to simulate

channel routing, evapotranspiration and surface runoff, respectively.

3.4.2 Parameter sensitivity, calibration and validation

The climate and hydrological regime is highly variable along the Blue Nile river from its origin in

the Ethiopian plateau to its outlet in the Sudanese plains. Therefore, it was important to perform a

sensitivity analysis and an optimization process at four locations namely, Kessie, Rosieres, Sennar

and Khartoum shown in Figure 1. Monthly observed discharge series at respective locations were

used to perform a Latin Hypercube – One At Time (LH-OAT) sensitivity analysis (v. Griensven,

2005), using the sum of the squared errors (SSQ) between the observed and simulated values as an

objective function. Optimization of the 10 most sensitive parameters was performed using the

Shuffled Complex Evolution algorithm (SCE-UA), included in SWAT as the Parameter Solution

(ParaSol) tool (v. Griensven et al., 2002; Vandenberghe et al., 2001).

Multi-site calibration (1972-1978) process was started from the most upstream measurement station

of Kessie, followed by subsequent downstream stations and finalizing the optimization of the entire

drainage area at Khartoum, outlet of the Blue Nile River basin. During the optimization process, the

parameters were changed in a distributed way for 91 HRUs, which were selected for being dominant

in area coverage in a subbasin. The calibrated model was validated for an independent period of

1979-1982 and evaluated by comparing annual water yield, graphical comparison and goodness of fit

statistics.

4. Results

4.1 Model performance evaluation

The quality of the model was evaluated by comparing the values of simulated and actual values for

the annual water yield and graphical evaluation of discharge series, which was further complemented

with statistical indicators of coefficient of determination (R2), the Nash-Sutcliffe Efficiency (NSE)

and uncertainty quantification for the calibration and validation periods. The goodness of fit statistics

for the stream flow during calibration and validation periods have been tabulated in Table 3.

Maximum 27% difference is observed in the mean annual stream flow at Kessie during the

validation period for which only one year data was available for comparison. Standard deviation of

simulated annual stream flow during the calibration period is lower than observed, which is a result

of lower simulated peak flows due to combined effect of synthetic weather data input and

12

assumptions made for implementing the reservoirs. High NSE and coefficient of determination (R2)

values illustrate the model performance to simulate reality in close approximation. The values of

NSE and the coefficient of determination remain higher than 0.62 and 0.71 in all the cases. Graphical

comparison results are not shown here, but, it’s important to mention that generally the simulated

peaks were underestimated at all the stations that could be explained by assumptions made for low

outflow release rates from the reservoirs. However, fair simulation of low flows were observed at all

the stations except Sennar, where some pronounced differences are visible.

Table 3. Goodness of fit statistics of the model results

Station

Name

Annual mean flow (m3/sec) Standard deviation (m

3/sec) NSE R

2

Actual SWAT Actual SWAT

Kessie Calibration 498 452 61 59 0.91 0.92

Validation 350 255 46 32 0.62 0.83

Rosieres Calibration 1434 1510 270 236 0.82 0.83

Validation 1196 1082 133 142 0.83 0.86

Sennar Calibration 1188 1275 300 232 0.79 0.79

Validation 891 836 144 160 0.83 0.85

Khartoum Calibration 1329 1595 177 219 0.80 0.83

Validation 1000 1134 185 234 0.70 0.71

The predictive non-uniqueness of the model from inherent uncertainties (Beven et al., 2001) in

synthetic weather data, measurement errors and optimization methods was quantified for annual

water yield and monthly flows at basin scale on 95% significance level, after applying Box-Cox

transformation, results shown in Table 4. Though wide predictive ranges of simulated flows were

observed, the overall assessment of the model results supported its application for climate change

scenario. Based on the general guidelines over the statistical indices in the field of hydrological

modeling, the tabulated results demonstrate model‘s ability to predict the realistic approximation of

river flows.

Table 4. Predictive range of annual water yield at 95% confidence interval of the Blue Nile River basin

Station Observed

(mm)

Calibrated

(mm)

Minimum

prediction (mm)

Maximum

prediction (mm)

%age range

Kessie 678 615 418 916 -32 to 49

Rosieres 244 257 154 355 -40 to 38

Sennar 186 199 113 281 -43 to 41

Khartoum 136 165 73 238 -56 to 44

4.2 Model Application for the climate change

Output of GCMs MIROC and INMCM for both the emission scenarios A2 and A1B were applied to

the calibrated and validated model to predict the future climate change impacts by the year 2025. The

impacts were studied on the hydrology of the system at annual and seasonal scale to assess variation

in water yield and flows at outlet of the basin. The impact assessment assumes no substantial change

in water withdrawals and man-made infrastructure in the Blue Nile river basin by the year 2025.

4.2.1 Annual water yield

Amongst the major concerns coupled with climate changes in shared river basins, scaling impacts on

the annual water yield at basin scale is of vital importance. Table 5 presents comparative insight into

13

expected changes in annual water yield and potential evapo-transpiration in the Blue Nile River basin

for future climate changes by the year 2025.

Table 5. Average annual changes in climate variables and water yield by 2025

Emission Scenario GCM ΔP (%) ΔTmax (oC) ΔTmin (

oC) ΔPET (%) ΔWYLD (%)

A2 MIROC 0 0.7 1.0 2 -11

INMCM 5 0.8 2.2 1 2

A1B MIROC 9 0.4 0.8 3 18

INMCM 4 0.9 2.3 2 25

MIROC presents relatively cooler climate for both the emission scenarios than that of INMCM

which shows higher increase in temperature in both the cases. Generally, the results show increase in

annual water yield at basin scale up to +25% except for MIROC-A2, which predicts water deficit in

the Blue Nile River basin. The A1B scenario representing a future world of very rapid economic

growth, low population growth, and rapid introduction of new and more efficient technologies,

shows 18% to 25% increase of the Blue Nile water yield. However, contrasting results are seen for

A2 emission scenario, which represents prevailing population growth rates and fight for sustaining

cultural and regional identities. For A2, MIROC predicts 11% decrease of the water yield, while

INMCM shows an increase of only 2%. The tabulated changes in climate variables present the

averaged impact over the entire study area, and therefore, these values do not demonstrate correlation

of climate variables in individual sub-catchments with similar changes in the water yield.

Furthermore, any change in precipitation does not result in immediate conversion to water yield, and

correlated analysis of the hydrological cycle may yield better insight. Except MIRCO-12, results

show improved water prospects for the riparian countries.

4.2.2 Low and high flow periods

The Blue Nile River is a flooding river, with 80% of the total yearly flow appearing in a period of

only four months. These flows are important for managing the reservoirs and droughts in the region.

For this purpose, predicted water yields during low flow periods (January to June) and high flow

periods (July to December) have been presented in Table 6. Results show a consistent increase in the

water yield at basin scale during the low flow periods, with a maximum of around 80% predicted by

the INMCM-A1B scenario. A1B also predicted a maximum of 20% increase in water yield during

high flow periods. However, the prevailing situation in the region closely approximated by A2,

shows a considerable decrease in expected water yield during high flow period for both the models,

with a maximum decrease of 15% predicted for MIROC. This situation can seriously hamper the

filling of the reservoirs during the high flow periods, resulting in moisture constrained agricultural

crops in later months and enhanced drought threats.

Table 6. Average water yield during low and high flow period at basin scale by 2025

Description Baseline MIROC INMCM

A2 A1B A2 A1B

Water yield during low flow period (mm) 9 11 12 13 16

Water yield during high flow period (mm) 112 96 131 110 135

14

4.2.3 Seasonal hydrological response

In addition to the availability of the flow in the river, its timing is of great concern to regulate and

manage the reservoir operations and the cropping pattern in Sudan and Egypt. Figure 7 presents the

model predictions of average monthly flows at Rosieres (left) and Khartoum (right) by the year 2025.

Figure 7. Figure 7: Comparison of average monthly flows at (left) Rosieres (right) Khartoum by 2025

Currently, the wet seasons spans from July to October with maximum flows occurring in August at

both stations. Generally, increased outflows from Rosieres reservoir are observed throughout the

year, except in April. This increase in reservoir outflows is understood to be a combined effect of

climate change and the loss of reservoir storage due to the sedimentation. The impact of increased

outflow at Rosieres, start appearing at downstream located Khartoum from September and continues

until April with consistent prediction of flow increase during low flow periods. However, the

increase in magnitude at Khartoum does not correspond with that of Rosieres, which is explained by

evapotranspiration losses due to high temperature in the 700 km long reach from Rosieres to

Khartoum.

A shift of the maximum river flow from August to September at Rosieres and subsequently at

Khartoum in case of MIROC-A1B is pronounced, which is justified by the precipitation increase in

respective months predicted in Rosieres reach produced in Figure 6. Similarly, A2 predicts a

decrease in river flows at Khartoum during July to September and an increase during October to

December. On the other hand, A1B shows a decrease in river flows at Khartoum for July and August

only. This is indeed a great concern for filling reservoirs in Egypt, which may enforce late filling of

reservoirs in Egypt but at the same time, low evaporation losses from the reservoirs in cooler months

may be expected. However, this late increase in river flows during low flow period at Khartoum,

needs to be checked by incorporating ongoing heightening of Rosieres dam with likely completion

by the year 2015.

5. Conclusions

This study is an effort to assess climate change impacts on the water resources of the Blue Nile River

basin by the year 2025. The hydrological modeling tool Soil and Water Assessment Tool (SWAT)

was employed to set up the model, using spatial maps with reasonably fine resolution, synthetic

weather data and limited information on the spatial distribution of irrigation schemes and reservoir

management. This study has contributed greatly for acquiring comprehensive information about

operation and management of reservoirs and irrigation water withdrawals with special emphasis on

checking the data quality and identifying the problematic data sets. Considering the topographical,

15

climatic and hydrological regime variability in the basin, the model was optimized at multiple sites

along the river route from the Ethiopian highlands to Sudanese plains, namely Kessie, Rosieres,

Sennar and Khartoum. The model performance was judged by comparing the simulated and observed

annual water yield, using graphical plots and statistical indices for calibration and validation periods.

Both NSE and coefficient of determination remain higher than 0.62 in all the cases, which

correspond to values of a good hydrological model in the cited literature.

Uncertainty in the outputs of GCMs is generally high, because of lack of scientific understanding of

various processes and monitoring facilities, therefore, climate change impacts were assessed by using

outputs from more than one GCM. The scenarios application was implemented using the outputs of

two GCMs namely MIROC and INMCM for the emission scenarios A2 and A1B, to the calibrated

and validated SWAT model of the study area. The GCMs output, given as a relative change of the

baseline records of the climatic variables, were implemented on the daily time series of precipitation

and maximum and minimum temperature. The study suggests the following major findings:

i) Increase in annual water yield at the basin scale is predicted by both models for emission

scenario A1B, while contrasting results in the water yield is observed in case of emission

scenario A2.

ii) Generally, an increase in precipitation magnitude during low flow months over the entire

basin is predicted by both the models.

iii) Consistent increase in low flows and decrease in high flows during July and August at outlet

of the basin is predicted for both the GCMs.

iv) A substantial increase in reservoir outflows at Rosieres during high flow period is observed,

which is seen as a combined result of increase in precipitation in the reach and loss of

reservoir storage due to the sedimentation.

Overall, the results suggest that the water resources of the Blue Nile river basin may not be adversely

affected by climate change, but it is dependent on future actions taken. Expeced increase in water

resources may help to meet future water needs in the region. Continuation of the study to investigate

the effects on environmental and ecological processes may be of great interest towards an integrated

basin solution.

16

References

Arnell NW, Reynard NS (1996) The effects of climate change due to global warming on river

flows in Great Britain. J Hydrol (Amst) 183:397–424. doi:

Beven, K. and Freer. Equifinality, data assimilation and uncertainity in the mechanistic modeling

of complex environmental systems using the GLUE methodology. Journal Hydrology 249

(2001), pp 11-29. 2001.

Conway, D., and Hulme, M. The impacts of climate variability and future climate change in the

Nile basin on water resources in Egypt. Water Resources Development 12 (3): 277-296. 1996.

Conway, D. From headwater tributaries to international river: Observing and adapting to climate

variability and change in the Nile Basin. Global Environmental Change 15: 99-114. 2005.

Dibike, Y., B. and Coulibaly P. Hydrologic Impact of Climate Change in the Saguenay

Watershed: Comparison of Downscaling Methods and Hydrologic Models, Journal of

Hydrology, 307(1-4):145-163. 2005.

Elshamy, M. E., Seierstad, I. A., and Sorteberg, A. Impacts of climate change on Blue Nile flows

using bias-corrected GCM scenarios, Hydrol. Earth Syst. Sci., 13, 551-565. 2009.

Githui, F., Gita, W., Mutua, F. and Bauwens, W. Climate change impact on SWAT simulated

streamflow in western Kenya. International Journal of Climatology. 2008.

FAO Aquastat Progamme. Dams and Agriculture in Africa. Water Development and

Management Unit (NRLW). Food and Agriculture Organization of the United Nations (FAO),

Viale delle Terme di Caracalla, 00100 Rome, Italy. 2007.

Fontaine, T.A., Klassen, J.F., Cruickö, T.S., and R.H. Hotchkiss. Hydrological response to

climate change in the Black Hills of South Dakota, USA. Hydrologic Sciences Journal––Journal

des Sciences Hydrologiques 46, pp. 27–40. 2001.

Frederick, K., D. Introduction: Water Resources and Climate Change, Frederick, K. D. (ed.)

Northampton MA: Edward Elgar Publishing. 514 pp. 2002.

Gleick, P. H. The vulnerability of runoff in the Nile basin to climatic changes. Environmental

Professional 13: 66-73. 1991.

Hurst, H.E., Black, R.P., and Simaika, Y.M. The Nile Basin. Volume-IX. The Hydrology of

Sobat and White Nile and the topography of the Blue Nile and Atbara. Ministry of Public Works,

Physical Department, Cairo, Egypt. 1959.

IPCC (Intergovernmental Panel on Climate Change). Guidelines on the use of scenario data for

climate impact and adaptation assessment. version 1, prepared by Carter, T. R.; Hulme, M.; Lal,

M. Task Group on Scenarios for Climate Impact Assessment. Geneva, Switzerland: IPCC. 69 pp.

1999.

IPCC (Intergovernmental Panel on Climate Change). Climate change 2007: Impacts, Adaptation

and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change. Cambridge, UK. Cambridge University Press. 976

pp. 2007.

Kim, U.; Lee, D.; Yoo, C. Effects of climate change on the streamflow for the Daechung dam

watershed. Journal of Korea Water Resources Association 37(4): 305-314. 2004.

Kim, U., Kaluarachchi, J. J., and Smakhtin, V. U. Climate change impacts on hydrology and

water resources of the Upper Blue Nile River Basin, Ethiopia. Colombo, Sri Lanka: International

Water Management Institute. 27p (IWMI Research Report 126). 2008.

17

MOWR (Ministry of Water Resources). Abbay River Basin Integrated Development Mater Plan

Project: Phase 2, Vol. VI, Water Resources Development, Part 2, Large Irrigation and

Hydropower Dams. Report, Addis Ababa, Ethiopia. pp. 22-24. 1998.

Neitsch, S.L., Arnold, J.G., Kiniry and Williams J.R. Soil and Water Assessment Theoretical

Documentation. Version 2005. Grassland, Soil and Water Research Laboratory, Agricultural

Research Service, Texas, United States of America. 2005.

NMSA (National Meteorological Service Agency). Initial national communication of Ethiopia to

the United Nations Framework Convention on Climate Change (UNFCCC). Report, Addis

Ababa, Ethiopia, 113 pp. 2001.

Coutsoukis, P. Sudan Irrigated Agriculture [Online]. The Library of Congress Country Studies;

CIA World Factbook. Available at

http://www.photius.com/countries/sudan/economy/sudan_economy_irrigated_agricultur~1784.ht

ml [Assessed November 2010]. 2004.

Raes, D., Willems, P. and GBaguidi, F. RAINBOW – a software package for analyzing data and

testing the homogeneity of historical data sets. Proceedings of the 4th

International Workshop on

‘Sustainable management of marginal drylands’. Islamabad, Pakistan, 27-31. 2006.

Reynolds, C. A., Jackson, T. J., and Rawls, W. J. 1999. Estimated Available Water Content from

the FAO Soil Map of the World, Global Soil Profile Databases, and Pedo-transfer Functions.

Proceedings of the AGU 1999 Spring Conference, Boston, MA. May31-June 4, 1999.

Schuol, J., and Abbaspour, K.C. A Daily Weather Generator for Predicting Rainfall and

Maximum-Minimum Temprature using Monthly Statistics Based on a Half-Degree Climate Grid.

Swiss Federal Institute of Aquatic Science and Technology (Eawag), Überlandstrasse 133, P.O.

Box 611, CH-8600 Dübendorf. 2005.

Sene, K., J.; Tate, E., L.; Farquharson, F., A., K. Sensitivity studies of the impacts of climate

change on White Nile flows. Climatic Change 50: 177-208. 2001.

Shahin, M. Hydrology of the Nile Basin. International Institute for Hydraulic and Environmental

Engineering Oude Delft 95, 2601 DA Delft, The Netherlands. 1985.

Stonefelt, M.D., Fontaine, T.A. and Hotchkiss, R.H. Impacts of climate change on water yield in

the upper Wind River basin. Journal of the American Water Resources Association 36(2): 321-

336. 2000.

Strzepek, K. M., and Yates, D. N. Economic and social adaptation to climate change impacts on

water resources: a case study of Egypt. Water Resources Development 12: 229-244. 1996.

Sutcliffe, J., V. and Parks, Y., P. The Hydrology of the Nile. IAHS Special Publication No. 5,

International Association of Hydrological Sciences, Wallingford, England. 179 pp. 1999.

UNESCO (United Nations Educational, Scientific and Cultural Organization). National Water

Development Report for Ethiopia. UN-WATER/WWAP/2006/7, World Water Assessment

Program, Report, MOWR, Addis Ababa, Ethiopia. 273 pp. 2004.

USBR. Land and Water Resource of the Blue Nile Basin, Ethiopia. Appendix IV Land

classification. United States Department of Interior Bureau of Reclaimation, Washington, DC,

USA. 1964b.

Van der Krogt, W. N. M. 2005. Annex C-Simulation of Lake Nasser inflow (RIBASIM).

Prepared for LNFDC/ICC Report Q3181.00 (Unpubl.), WL/Delft Hydraulics, the

Netherlands.

18

van Griensven, A. Sensitivity, auto calibration, uncertainty and model evaluation in SWAT2005

(DRAFT). UNESCO-IHE, The Netherlands. 2005.

van Griensven, A., Francos A. and Bauwens W. Sensitivity analysis and auto-calibration of an

integral dynamic model for river water quality. Water Science and Technology, 45(5), 321-328.

2002.

Vandenberghe, V., van Griensven A. and Bauwens W. Sensitivity analysis and calibration of the

parameters of ESWAT: Application to the river Dender. Water Science and Technol., 43(7), 295-

301. 2001.

Van Liew, M. W., Garbrecht, J. Sensivity of hydrologic response of an experimental watershed

to changes in annual precipitation amounts. American Society of Agriculture and Biological

Engineers, Paper 012001. 2001.

Varanou, E, E. Gkouvatsou, E. Baltas, and M. Mimikou. Quantity and quality integrated

catchment modelling under climatic change with use of Soil and Water Assessment Tool

model. J. Hydrol. Engr. 7(3): 228-244. 2002.

Waterbury, J. Hydropolitics of the Nile Valley. Syracuse University Press, New York, 301 pp.

1979.

Whittington, D., Wu, X., and Sadoff. C. Water resources management in the Nile basin: The

economic value of cooperation. Water Policy 7: 227-252. 2005.

Wilby, R. L.; Harris, I. A framework for assessing uncertainties in climate change impacts: Low-

flow scenarios for the River Thames, UK. Water Resources Research 42: W02419. 2006.

Wurbs, R. A., Muttiah, R. S. and Felden, F. Incorporation of climate change in water availability

modeling. Jounral of Hydrologic Engineering, 10 (5): 375-385. 2005.

Yates, D., N. and Strzepek, K., M. An assessment of integrated climate change impacts on the

agricultural economy of Egypt. Climatic Change 38: 261–287. 1998a.

Yates, D., N. and Strzepek, K., M. Modeling the Nile Basin under climatic change. Journal of

Hydrologic Engineering 3(2): 98-108. 1998b.