Future hydrology and climate in the River Nile basin: a review

This article was downloaded by: [194.204.215.228]On: 17 January 2012, At: 02:41Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Hydrological Sciences JournalPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/thsj20

Future hydrology and climate in the River Nile basin: areviewGiuliano Di Baldassarre a , Mohamed Elshamy b , Ann van Griensven a , Eman Soliman c , MaxKigobe d , Preksedis Ndomba e , Joseph Mutemi f , Francis Mutua f , Semu Moges g , YunqingXuan a , Dimitri Solomatine a h & Stefan Uhlenbrook a ha UNESCO-IHE Institute for Water Education, NL-2601 DA Delft, The Netherlandsb Nile Forecast Center, Ministry of Water Resources and Irrigation, Korniche El-Nile, EmbabaGiza, 12666, Egyptc Nile Basin Initiative Water Resources Management Project, Cairo, Egyptd Makerere University, Faculty of Technology, Department of Civil Engineering, PO Box 7062,Kampala, Ugandae Water Resources Engineering Department, College of Engineering and Technology,University of Dar es Salaam, PO Box 35131, Dar es Salaam, Tanzaniaf Department of Meteorology, University of Nairobi, PO Box 30197-00100, Nairobi, Kenyag Department of Civil Engineering, Faculty of Technology, Addis Ababa University, AddisAbaba, Ethiopiah Section of Water Resources, Delft Universities of Technology (TU Delft), Delft, TheNetherlands

Available online: 28 Mar 2011

To cite this article: Giuliano Di Baldassarre, Mohamed Elshamy, Ann van Griensven, Eman Soliman, Max Kigobe, PreksedisNdomba, Joseph Mutemi, Francis Mutua, Semu Moges, Yunqing Xuan, Dimitri Solomatine & Stefan Uhlenbrook (2011): Futurehydrology and climate in the River Nile basin: a review, Hydrological Sciences Journal, 56:2, 199-211

To link to this article: http://dx.doi.org/10.1080/02626667.2011.557378

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form toanyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses shouldbe independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly inconnection with or arising out of the use of this material.

199Hydrological Sciences Journal – Journal des Sciences Hydrologiques, 56(2) 2011

Future hydrology and climate in the River Nile basin: a review

Giuliano Di Baldassarre1, Mohamed Elshamy2, Ann van Griensven1, Eman Soliman3, Max Kigobe4,Preksedis Ndomba5, Joseph Mutemi6, Francis Mutua6, Semu Moges7, Yunqing Xuan1,Dimitri Solomatine1,8 & Stefan Uhlenbrook1,8

1UNESCO-IHE Institute for Water Education, NL-2601 DA Delft, The [email protected]

2Nile Forecast Center, Ministry of Water Resources and Irrigation, Korniche El-Nile, Embaba Giza 12666, Egypt3Nile Basin Initiative Water Resources Management Project, Cairo, Egypt4Makerere University, Faculty of Technology, Department of Civil Engineering, PO Box 7062, Kampala, Uganda5Water Resources Engineering Department, College of Engineering and Technology, University of Dar es Salaam, PO Box 35131, Dar esSalaam, Tanzania6Department of Meteorology, University of Nairobi, PO Box 30197-00100, Nairobi, Kenya7Department of Civil Engineering, Faculty of Technology, Addis Ababa University, Addis Ababa, Ethiopia8Section of Water Resources, Delft Universities of Technology (TU Delft), Delft, The Netherlands

Received 30 June 2010; accepted 14 December 2010; open for discussion until 1 September 2011

Citation Di Baldassarre, G., Elshamy, M., van Griensven, A., Soliman, E., Kigobe, M., Ndomba, P., Mutemi, J., Mutua, F., Moges, S.,Xuan, J.-Q., Solomatine, D. & Uhlenbrook, S. (2011) Future hydrology and climate in the River Nile basin: a review. Hydrol. Sci. J.56(2), 199–211.

Abstract A critical discussion of recent studies that analysed the effects of climate change on the water resourcesof the River Nile Basin (RNB) is presented. First, current water-related issues on the RNB showing the particularvulnerability to environmental changes of this large territory are described. Second, observed trends in hydro-logical data (such as temperature, precipitation, river discharge) as described in the recent literature are presented.Third, recent modelling exercises to quantify the effects of climate changes on the RNB are critically analysed. Themany sources of uncertainty affecting the entire modelling chain, including climate modelling, spatial and tempo-ral downscaling, hydrological modelling and impact assessment are also discussed. In particular, two contrastingissues are discussed: the need to better recognize and characterize the uncertainty of climate change impacts on thehydrology of the RNB, and the necessity to effectively support decision-makers and propose suitable adaptationstrategies and measures. The principles of a code of good practice in climate change impact studies based on theexplicit handling of various sources of uncertainty are outlined.

Key words hydrology; climate change; water resources management; River Nile

Hydrologie et climat futurs dans le bassin du Nil: une revueRésumé Une discussion critique des études récentes qui ont analysé les effets du changement climatique surles ressources en eau du bassin du Nil (RNB) est présentée. Premièrement, les problèmes actuels liés à l’eaudans le RNB montrant la vulnérabilité particulière de ce vaste territoire aux changements environnementauxsont décrits. Deuxièmement, les tendances observées dans les données hydrologiques (comme la température,les précipitations, le débit des rivières) sont présentées, telles qu’elles sont décrites dans la littérature récente.Troisièmement, les exercices récents de modélisation quantitative des effets des changements climatiques dansle RNB sont analysés de manière critique. Les nombreuses sources d’incertitude qui affectent toute la chaînede modélisation, incluant la modélisation du climat, la descente d’échelles spatiale et temporelle, la modélisa-tion hydrologique, et l’évaluation des impacts sont également discutées. En particulier, deux questions contrastéessont discutées: la nécessité de mieux identifier et caractériser l’incertitude des impacts du changement clima-tique sur l’hydrologie du RNB, et la nécessité de soutenir efficacement les décideurs et de proposer des stratégiesd’adaptation et des mesures appropriées. Les principes d’un code de bonnes pratiques dans les études d’impact duchangement climatique sont décrits, qui reposent sur le traitement explicite des diverses sources d’incertitude.

Mots clefs hydrologie; changement climatique; gestion des ressources es eau; Fleuve Nil

ISSN 0262-6667 print/ISSN 2150-3435 online© 2011 IAHS Pressdoi: 10.1080/02626667.2011.557378http://www.informaworld.com

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

200 Giuliano Di Baldassarre et al.

1 INTRODUCTION

In recent years, a large part of the scientific commu-nity has made efforts analysing the impact of pro-jected climate change on water resources and propos-ing adaptation strategies (e.g. Loaiciga et al., 1996;Vicuna & Dracup, 2007; Hattermann et al., 2008;Wilby et al., 2008; Allamano et al., 2009; Gouldenet al., 2009). The usual framework of this type ofstudy can be summarized as follows (e.g. Elshamyet al., 2009b): (a) choice of one or more scenar-ios from the IPCC (International Panel on ClimateChange) special report on emission scenarios (Bateset al., 2008), which depend on the future economyand energy use policies; (b) choice of one or moreglobal circulation models (GCM); (c) downscaling ofthe GCM climate output such as rainfall to the spe-cific river basin scale; (d) use of the downscaled GCMoutputs as inputs for a hydrological model; and (e)analysis of hydrological model results by comparingthem to the corresponding results related to the cur-rent climate or different possible future climates. Thisapproach has become very popular as it potentiallyallows the quantification of changes in floods, flowduration curves, and whatever part of the hydrologicalcycle (Blöschl & Montanari, 2010).

In this context, a number of studies analysed theeffects of climate change on the hydrology of theRiver Nile Basin (RNB), the world’s longest river.In fact, the RNB could be vulnerable to water stressunder climate change because of the limited wateravailability and the increasing demand for water fromdifferent sectors (e.g. Bates et al., 2008). In addition,there is a serious concern about the fact that sea-level rise could adversely impact on people living inthe Nile Delta and other coastal areas. Nevertheless,Conway (2005) found that there is no clear indicationof how River Nile flow would be affected by cli-mate change, because of the uncertainty in projectedrainfall patterns in various part of the basin and theinfluence of complex water management (and watergovernance structures). More recently, Githui et al.(2008) used a technique of adjustment (the so-calleddelta change method; e.g. Hay et al., 2007) of histori-cal time series to project GCM impacts on flood risksin the Nzoia River, one of the major river systemsdraining into Lake Victoria. In addition, Elshamyet al. (2009a) and Nawaz & Bellerby (2007) analysedclimate change effects on the main Nile at Dongolaand the Blue Nile at Diem, using a spatio-temporalstatistical downscaling technique for various GCMsand showed varying trends depending on the GCM

Fig. 1 Simulated decadal mean flows at Dongola on themain Nile from six GCM experiments. The values repre-sent averages of 10 realizations of statistically downscaledscenarios for each experiment while the base refers to thebaseline period 1992–2001.

used (Fig. 1). Furthermore, Soliman et al. (2008)investigated climate change effects on the Blue Nilecatchment using the regional climate model RegCM3to downscale the results of the ECHAM5 generalcirculation model (Max Planck Institute, Hamburg,Germany). These studies demonstrate the large diver-sity in the use of IPCC scenarios, climate modelsand downscaling techniques (time series adjustments,statistical and physically-based methods). These dif-ferent techniques may lead to opposing trends andcontradicting recommendations for policy makers.

Climate models are seen as the most usefuland powerful tools for providing detailed informa-tion to evaluate how climate variability and emergingsignals of climate change are likely to impact thewater resources of the RNB. Climate models haveprogressed rapidly, and nowadays sophisticated cli-mate models enable simultaneous handling of theatmosphere–land–ocean components of the climatesystem with certain temporal and spatial resolutions.Indeed, treatment of hydrological cycle and surfaceprocesses including albedo, roughness, vegetationtype and ground characteristics are crucial modulesof a comprehensive climate model. The fundamentalsof climate modelling can be found in literature (e.g.Trenberth, 1992; Hamilton & Ohfuchi, 2008).

Climate models have demonstrated the ability toreproduce the observed characteristics of present andpast climate over many regions of the globe (e.g.Randall et al., 2007). In particular, climate modelshave been reported to be the only tools that pro-vide some insights on how the future climate mightevolve over projection time scales ranging from afew years to decadal and multi-decadal periods inthe future (Meehl et al., 2007). Nevertheless, climate

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

Future hydrology and climate in the River Nile basin: a review 201

models still have a poor capacity to foresee precip-itation (Randall et al., 2007) and, despite the strongefforts that have been made over the last decades bymany researchers, the uncertainties in projections offuture climate change have not significantly decreased(Roe & Baker, 2007; Anagnostopoulos et al., 2010;Kundzewicz & Stakhiv, 2010). Besides, the changesindicated by climate models may be too small incomparison to the natural variability of hydrolog-ical processes and uncertainty of runoff measure-ments (Beven, 2006; Koutsoyiannis et al., 2009; DiBaldassarre & Montanari, 2009); although part of thereason may be the huge difference in scales of bothsides.

It has been customary for water communities touse climate model outputs as quantitative informa-tion for assessing climate change impacts on waterresources management, flood risk management, andrain-fed agriculture, to name a few. However, cau-tion is always needed in considering certain modellingaspects, such as: (i) the choice of the particular modelor set of global models to use; (ii) domain configura-tions for regional climate models; and (iii) choosingappropriate model physics, especially those handlingmoist convective processes related to reproducingobservational climatology and inter-annual featuresof regional and local precipitation. With respect to thecoupling procedure with other models, it is sometimesnecessary to choose plausible methods of statisticaldownscaling as well as finding a way of generatingrobust precipitation estimates from various models tobe able to drive hydrological models at various spa-tial and temporal scales, e.g. reproducing decadal andmulti-decadal features as well as climate change pro-jections at specific areas of the RNB. Recent studiesthat can help in addressing some of these modellingconcerns include: Kang & Hong (2008), Schoof et al.(2009), Mutemi et al. (2007), Wilby et al. (2009) andWilby (2010). These issues constitute the subject ofthis review as a basis for formulating climate mod-elling and downscaling application for the RNB waterresources, flooding, and livelihood activities such asagriculture under the variability of current climateand uncertainty of emerging climate change signals.

The purpose of this paper is to: (1) review thefindings of recent climate change impact studies onthe River Nile Basin (RNB); (2) classify and char-acterize the sources of uncertainty that need to beconsidered in climate change impact studies; and (3)present the elements of a code of best practices foruse in climate change impact studies. More specif-ically, the first part describes the RNB (River Nile

Basin) and the current water-related issues and showsthe particular vulnerability of this large and impor-tant basin. Then, this paper reports on observationalrecords of hydrological data, such as temperature, pre-cipitation and river discharge, and critically examinesthe trends reported in the scientific literature. In addi-tion, recent modelling studies that were performed toanalyse the effects of climate change on the RNB arereviewed. Finally, in the last sections, we discuss, clas-sify and characterize the sources of uncertainty to beconsidered in the climate change studies, by consid-ering the entire modelling chain, and provide someinitial insights for a code of good practice for studieson climate change impacts.

2 THE RIVER NILE BASIN

The hydrological behaviour of the River Nile led toone of the first scientific questions. Thales of Miletus(640–546 BC) tried to understand the hydrologicalparadox of the Nile where flooding always occurs insummer when rainfall in Egypt is almost nonexis-tent (Koutsoyiannis et al., 2010). Also, the extremelylong time series of Nile water level are unique, asthey extend for several centuries (e.g. Hassan, 2004).Figure 2 shows the annual minimum water level,and the 25-year average, of the River Nile for theyears 622–1284 measured at the Roda Nilometer nearCairo (Beran, 1994). Nilometer data led to the dis-covery of the so-called Hurst phenomenon (Hurst,1951; Montanari et al., 1997; Koutsoyiannis, 2002)and clearly highlights the huge climatic variability atlarge time scales (Fig. 2; Koutsoyiannis, 2003).

The River Nile (Fig. 3) is the world’s longest river(6670 km), and its catchment extends over 10 EastAfrican countries with several water uses, such aswater supply for agricultural, industrial and domesticuse, power generation, and environmental manage-ment (Georgakakos, 2007). The drainage basin of the

Fig. 2 Roda Nilometer near Cairo: time series of theannual minimum water level of the Nile River for the years622–1284 AD.

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

202 Giuliano Di Baldassarre et al.

Fig. 3 The River Nile Basin and location of gauging stations.

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

Future hydrology and climate in the River Nile basin: a review 203

Nile covers 3 255 000 km2, about 10% of the areaof Africa. The Nile flow is controlled by a num-ber of manmade structures between Lake Victoria(covering parts of Kenya, Tanzania and Uganda) andthe Ethiopian Highlands through Sudan to the HighAswan Dam in Egypt (Koutsoyiannis et al., 2008),and several planned facilities in the middle reaches(Ethiopia and Sudan).

The flow rate of the Albert Nile at Mongalla,the most reliable gauge downstream of the outlet ofLake Albert, is almost constant throughout the yearand averages about 1000 m3/s. After Mongalla, theNile is known as the Bahr El Jebel, which entersthe enormous swamps of the Sudd region in south-ern Sudan. About half of the Nile water, in additionto local rainfall, is lost in this swamp to evapora-tion and transpiration. The average flow rate in theBahr El Jebel at the downstream end of the swampsis about 510 m3/s. To the west of Bahr El-Jebel liesthe Bahr El-Ghazal basin, which is the largest Nilesub-catchment (520 000 km2), but contributes a rel-atively small amount of water, about 2 m3/s, due toevaporation in Bahr el-Ghazal swamps with possiblespillage to the Sudd swamp. Bahr el-Jebel meets Bahrel-Ghazal at Lake No and then flows to the east tomeet the Sobat River, forming the White Nile. TheSobat River drains about half as much land as theBahr El-Ghazal, 225 000 km2, but contributes around400 m3/s to the Nile (Shahin, 2002). During floods,the Sobat carries a large amount of sediment, addinggreatly to the White Nile’s colour. The average flowof the White Nile at Malakal, just below the SobatRiver junction, is about 900 m3/s, while the peak flow,occurring in October, is approximately 1200 m3/sand the minimum flow, in April, is about 600 m3/s.From here, the White Nile flows to Khartoum, whereit merges with the Blue Nile to form the River Nile.Further downstream, the Atbara River, the last sig-nificant Nile tributary, originating from the EthiopianHighlands, merges with the Nile. During the dry sea-son in the Ethiopian Highlands, from November toApril, the White Nile contributes between 70 and 90%of the total discharge from the Nile.

The Blue Nile contributes approximately60–70% of the total annual River Nile discharge.The flow of the Blue Nile varies considerably overits annual cycle and is the main contributor to thelarge natural variation of the Nile flow. During thewet season, during late August, the peak flow ofthe Blue Nile often exceeds 5700 m3/s. Before theconstruction of dams on the main river, the annualdischarge varied by a factor of 15 at Aswan: peakflows of over 8200 m3/s occurred during the later

Table 1 Gross domestic product (GDP) per capita andaverage population growth (%) in the period 2005–2010 (United Nations list 2005–2010 and InternationalMonetary Fund, 2009).

Country GDP(US$)

World rank(out of 180)

Populationgrowth (%)

World rank(out of 230)

Burundi 174 179 3.90 2Congo∗ 171 180 3.22 9Egypt 2450 112 1.76 71Eritrea 328 175 3.24 7Ethiopia 418 170 2.51 29Kenya 842 146 2.65 23Rwanda 512 160 2.76 21Sudan 1388 126 2.22 43Tanzania 547 157 2.47 30Uganda 472 161 3.24 8World

average10433

∗Democratic Republic of Congo.

parts of August and early September, while minimumflows of about 550 m3/s were recorded during lateApril and early May.

Water resources in the RNB are very vulnerableto environmental changes because of the precariousequilibrium of the system and in view of the fact thatseveral of the RNB countries are among the poorestin the world. The GDP per capita and correspondingworld rank, according to the International MonetaryFund (2008) website (www.imf.org), for the 10 coun-tries of the RNB are reported in Table 1. Futureclimate change is one of the driving factors that mightimpose further stress on the already vulnerable waterresources of the RNB.

It is important to note that other driving fac-tors, such as population growth and consequentland-use changes and urbanization, might affectwater resources in the RNB more than climatevariability/change (Wilby et al., 2008; Koutsoyianniset al., 2009). Table 1 also shows, as an example, theaverage population growth rate (%) during the period2005–2010 (listed by country) and the correspondingworld rank according to the United Nations website(http://esa.un.org/unpp) for some of the Nile coun-tries. The rates are generally high and range between1.76 and 3.90% per year.

3 PAST CHANGES (TEMPERATURE,PRECIPITATION AND DISCHARGE DATA)ON THE RNB

3.1 Temperature

The mean annual temperature over the basin variesfrom a minimum of 10◦C over Ethiopian andEquatorial highland areas to about 30◦C in the

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

204 Giuliano Di Baldassarre et al.

central Sudan. The mean annual temperatures in theEquatorial Lakes region have little variations through-out the year and range between 16 and 27◦C depend-ing on locality and altitude. In the Sudan plains,the minimum temperature occurs in January and themaximum in May or June, when it rises to a dailyaverage of 41◦C in Khartoum. From the IPCC FourthAssessment Report (IPCC, 2007) and the countryreport of Ethiopia (NMA, 2007), there is a generalconsensus on the rise of temperature. In particular,higher water temperatures have been reported in lakesin response to warmer conditions (Bates et al., 2008).The minimum temperatures over Ethiopia show anincrease of about 0.37◦C per decade, which indicatesthe signal of warming over the period of the analysis1951–2005 (NMA, 2007). Within the Nzoia catch-ment that drains into Lake Victoria, located in Kenya,there is an increasing trend of 0.79◦C per decadein the lowlands and 0.21◦C per decade in the high-lands (Githui, 2008). However, temperature data arevery sparse and have a lower availability compared toprecipitation data for the RNB.

3.2 Precipitation

Rainfall in the Nile Basin shows various modes ofseasonality within the annual cycle and inter-annualvariability depending on the location of the specificsub-catchment with respect to the Equator and moistadvecting wind regimes. The annual cycle of rainfallof the larger White Nile area (consisting of Kenya,Uganda and Northern Tanzania) is a bimodal patternwith the two main wet seasons occurring March–Mayand October–December; while the Blue Nile area(consisting of Ethiopian Highlands and Sudan) showsa unimodal distribution, with the main wet seasonspanning the period June–September. According toSutcliffe & Parks (1999), the annual rainfall amountvaries from less than 50 mm/year (over the loweror main Nile) in northern Sudan and southern Egyptto more than 1200 mm/year over the Ethiopian partof the Nile. Further south it varies from around900 mm/year over the Sudd swamp to more than1100 mm/year over the Lake Victoria sub-basin, andreaches as high as 1600 mm/year over the lake itself.

Studies over the Nile basin provide conflictingevidence regarding the existence of any long termtrend in the rainfall (Wing et al., 2008; Conway et al.,2004). While there is generally no significant changedetected in the annual rainfall in most of the Nilesub-basins, there appears to be decreasing season-ality in some key watersheds of the upper Nile in

Ethiopia such as the southern Blue Nile and Baro-Akobo (Wing et al., 2008). Conway & Hulme (1993)also supported the idea that, except for Lake Victoria,all sub-basins of Nile experienced slightly-to-stronglydecreasing trends in precipitation. The three catch-ments of Bahr el Ghazal, Sobat and Central Sudanrecorded significant drops in annual precipitation,whereas the observed changes in many other catch-ments were not significant. Referring to the longerterm, e.g. the period 1905–1984, Sayed et al. (2004)provided evidence that the RNB has shown a slightlyincreasing trend in rainfall over the observation periodof 1905–1965, followed by a prolonged decline reach-ing its minimum in 1984, and then recovering signifi-cantly during the 1990s. This is a generally consistentexplanation of precipitation trend presented in IPCCTechnical Paper VI (Bates et al., 2008). However, thescientific literature does not provide clear indicationson the trend in the occurrence of extreme rainfallevents.

3.3 River discharge

Observational studies over the main Nile or thetwo main Nile systems (the Blue Nile and WhiteNile) agree that the seasonal and inter-annual vari-ability is more significant than any long-term trend(Awulachew et al., 2008; Bowden et al., 2009;Conway & Hulme, 1993). This is consistent with thevariability of precipitation over the region. However,the variability of flow in the two Nile systems appearsto behave oppositely in their temporal patterns offluctuation (Bowden et al., 2009; Conway & Hulme,1993). For instance, the Blue Nile had shown markeddecreasing tendency of flows from 1960s to 1980swhile the White Nile flow was higher than aver-age during the same period. Analysis of the mainNile flow from 1871/72 to 2000/01 (Awulachewet al., 2008) showed that the flow had been fluctuatingbetween 63 to 122 × 109 m3 with a lowest record ofabout 40 × 109 m3 (1916/17) and a highest of about152 × 109 m3 in 1881/82.

As the bulk of inflow (more than 80%) to theLake Victoria is from the rainfall over the Lake itself(Tate et al., 2004), the variability in rainfall plays asignificant role in modulating the Lake level. The riseof lake levels in the early 1960s (Sutcliffe & Parks,1999) to a record high since the measurement startedis another example of climate variability and possibleconsequences of one extreme event in the Nile Basin.Based on the review of the publications on the cli-mate variability of the RNB, Conway (2005) showed

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

Future hydrology and climate in the River Nile basin: a review 205

that climate variability of the Lake Victoria basin andthe Ethiopian Highlands is primarily due to rainfallfluctuations.

4 POSSIBLE FUTURE CHANGES IN THERNB (MODEL RESULTS)

4.1 Downscaling techniques

A necessary step in utilizing climate model outputsis to find a way to bridge the scale gap between cli-mate model outputs and those being driven by them,for instance hydrological models. Downscaling refersto the fact the forcing data from a climate model(e.g. GCM at large scale) is reproduced at smallerscale by means of intermediate models. Dependingon the methods employed, the downscaling can begenerally categorized into two groups: (1) dynami-cal downscaling, which makes use of (limited area)regional climate models embedded into the coarsebackground field supplied by the (global) climatemodels; and (2) statistical downscaling, which mapsthe coarse grid value of GCM outputs to the (gauge)locations or to finer grids by deriving and applyinga statistical relationship between the GCM variablesand observations. A detailed review of these meth-ods can be found in Fowler et al. (2007). An exampleof a machine learning approach to downscaling (thatis potentially more accurate than the statistical one,and deserves testing in the RNB context) is pre-sented by Dibike & Coulibaly (2006). The followingsub-sections give some example applications of bothcategories over the RNB.

4.2 Dynamical downscaling

Following the first attempts by Mohamed et al.(2005), a regional climate model (RegCM3) hasbeen tested and applied in a series of studies in theregion by Soliman et al. (2008, 2009). RegCM3 wasdeveloped at the Abdus Salam International Centerfor Theoretical Physics (ICTP). The model is freelyavailable at http://www.ictp.it/∼pubregcm. Thesoil–vegetation–atmosphere interactions are param-eterized through the BATS scheme (Biosphere–Atmosphere Transfer Scheme). The radiative trans-fer scheme of NCAR CCM3 (Community ClimateModel 3; Kiehl et al., 1996) is used in RegCM3(Giorgi & Mearns, 1999), which includes the forc-ing effects of different greenhouse gases, cloud water,cloud ice and atmospheric conditions.

Soliman et al. (2008) configured and validatedRegCM3 over a domain covering two importantstreamflow-generating regions of the Nile Basin: theSobat and Blue Nile sub-basins. Using the output ofRegCM3, the NFS (Nile Forecast System) was usedto convert precipitation predictions into flow predic-tions for the Blue Nile at Diem (Fig. 4) with suffi-cient accuracy. Soliman et al. (2008) concluded that:(i) the spatial pattern of precipitation is generally wellcaptured by the model both in the rainy and dry sea-sons; (ii) with regards to temperature, the model isbiased towards warmer conditions (2–6◦C) over thewhole studied domain in all seasons although it cap-tured the spatial and temporal patterns sufficiently;and (iii) the multi-year flow simulation using NFS uti-lizing RegCM3 output showed good performance incapturing the seasonality of flows for the Blue Nile

12000

10000

8000

6000

Avg

.Mo

nth

ly F

low

MC

M/m

on

th

4000

2000

0Jan Feb Mar Apr May Jun Jul

Month

% C

hang

e

Aug Sep Oct Nov Dec

30

20

10

0

–10

–20

–30

–40

–50

Reference % ChangeFuture

Fig. 4 Changes in mean monthly streamflow of the Blue Nile at Diem during the year; between the current condition(reference) and the changed conditions in the year 2055 (future).

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

206 Giuliano Di Baldassarre et al.

(RMSE = 0.80), but the performance for the Sobatwas poor (RMSE = 0.30) due to erroneously pre-dicted extreme flows by the NFS due to probablyimproper calibration of the NFS over the Sobat inaddition to some high spots of rainfall predicted byRegCM3.

Demonstration of nesting the RegCM3 within aGCM is presented in Soliman et al. (2009) wherethe ECHAM5 A1B scenario was downscaled usingRegCM3 for the same domain described above. Themodel estimated future increases in Blue Nile flow atDiem of about 1.5% annually (∼740 × 106 m3/year).However, the estimated increase of flow was largerduring the beginning of the flood season (+10%),whereas the flow was predicted to decrease towardsthe end of the rainy season in October and November,as well as in the dry season.

4.3 Statistical downscaling

Elshamy et al. (2009a) applied a bias correctiondownscaling approach to downscale the output of17 general circulation models (GCMs) included inIPCC (2007) using the A1B emission scenario.Downscaled precipitation and potential (referencecrop) evapotranspiration (PET) scenarios for the2081–2098 period were constructed for the upperBlue Nile basin. The method is based on a distri-bution mapping approach to correct the intensity ofdaily precipitation outputs of GCMs (Ines & Hansen,2006). The idea is to fit a probability distributionto the daily rainfall of observed data as well as toGCM data for a control period (1961–1990). Then,in order to obtain the bias corrected scenario, correc-tion factors were estimated for the whole distributionand applied to the future period. The method con-serves the relative change (scenario/control) for thewhole rainfall distribution (i.e. mean, variability andextremes) such that no artificial relative changes areadded to the scenario after the bias correction. Sincethe observed dataset used (the merged satellite-gaugeestimates obtained from the NFS database) has a fineresolution of 20 × 20 km2, several pixels fall into onelarge grid cell of any GCM. By adjusting the fitteddistributions for each pixel within the GCM grid celland repeating the procedure for each GCM grid cell,a high resolution (spatially downscaled) rainfall fieldis obtained.

To bias-correct the PET future scenarios,monthly gridded correction factors have been cal-culated as ratios of the observed climatology andGCM climatology for the baseline period. These

factors were then applied to the future monthly PETclimatology. As the factors were calculated using theCRU data (Mitchell & Jones, 2005) of 0.5◦ resolu-tion, the correction implies downscaling to the CRUresolution in the same way as done for precipitation.

The downscaled rainfall and PET were used todrive the NFS hydrological model to assess theirimpacts on the flows of the upper Blue Nile at Diem.The study found disagreements among the GCMs onthe direction of precipitation change. Changes in totalannual precipitation ranged between −15% to +14%but more models reported reductions (10) than thosereporting increases (7). Several models (6) reportedsmall changes within 5%. The ensemble mean ofall models showed almost no change in the annualtotal rainfall. All models predicted the temperatureto increase between 2◦C and 5◦C and consequentlyPET to increase by 2–14%. Elshamy et al. (2009a)assessed the changes to the water balance using theBudyko framework. The basin is shown to originallybelong to a moisture constrained regime. However,during the wet season the basin is largely energy con-strained. For no change in rainfall, increasing PETthus leads to a reduced wet season runoff coefficient.The ensemble mean runoff coefficient (about 20% forbaseline simulations) is reduced by about 3.5% inthe future. Assuming no change or moderate changesin rainfall, the simulations indicated that the waterbalance of the upper Blue Nile basin may becomemore moisture constrained in the future. The pre-dicted ensemble mean annual flow at Diem is reducedby 15% compared to the baseline (Fig. 5) within arange of –60% to 45%.

Another example of statistical downscaling waspresented by Kigobe (2009) who applied generalizedlinear models (GLMs) as a stochastic downscalingtechnique to provide projections at scales finer thanthe GCM model-grid resolutions for the equatorialclimate of East and Central Africa. This involveddeveloping a weather generator model to simulatedaily sequences of rainfall. Several GLMs were fit-ted to observed daily rainfall and the correspondingatmospheric reanalysis data and subsequently appliedto GCM outputs by estimating the conditional jointdistribution for the entire time series using Bayesianapproaches. The results of this study showed thatGLMs performed well when simulating the histori-cal period with inter-annual variability and monthlystatistics successfully reproduced albeit some bias forsome regions. The GLM framework was extendedto predicting local scale precipitation under scenar-ios of global climate change. For the Kyoga basin in

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

Future hydrology and climate in the River Nile basin: a review 207

Fig. 5 Predicted changes in (a) rainfall for the Upper Blue Nile, and (b) flow at Diem under bias corrected downscaledclimatic scenarios.

Table 2 Example of statistical downscaling results: pro-jected increase in precipitation (%) for different seasons(Kigobe, 2009).

December–February

March–May

June–August

September–October

2020s 14% 19% 24% 3%2050s 21% 42% −12% 45%2080s 68% −3% 72% 46%

Uganda, averaged monthly changes in precipitationshowed significant seasonal variability. In particular,the results further show that spatial patterns varysignificantly due to the spatial heterogeneity and dif-ferences in hydroclimatic variations associated withthe general circulation teleconnections in the equa-torial climate of East and Central Africa. Table 2shows the results of statistical downscaling (in termsof precipitation increase) obtained using six GCMprojections under the A2 emission scenario, for thedifferent seasons. The projected changes are mixed,but generally, based on spatial averaging, precipita-tion is projected to decrease during the period June–August for the 2050s (Table 2). The findings alsoshowed that the total rainfall received in the first rain-fall season (March–May) is still higher than the totalreceived in the second rainfall season (September–November). Based on the six GCM projections underthe A2 emissions scenario, the projected changes inannual precipitation are 12% for the 2020s, 26% forthe 2050s, and 36% for the 2080s.

5 DISCUSSION

We have summarized a number of studies that werecarried out to understand the possible impacts of

climate change on the Nile system. However, as men-tioned above, other driving factors such as populationgrowth, and consequent land-use changes and urban-ization (e.g. Uhlenbrook, 2009) might play a morerelevant role in influencing the water resources ofthe RNB than climate change. Several models havebeen used to estimate the effects of climate changeon the river basin hydrology, water management,hydropower, urban drainage, water quality, the aquaticecology, etc. Nevertheless, there is still a real needto gain insights into how the RNB is going to beinfluenced by the ever-changing climate.

As already mentioned, analysis of climate changeis complicated by the associated uncertainties, whichare very significant (Koutsoyiannis et al., 2007). Infact, the entire modelling chain (climate modelling,spatial and temporal downscaling, hydrological mod-elling and impact assessment) is affected by relevantuncertainty, which is important to take into account indecision making processes (e.g. Pappenberger et al.,2005).

Concerning the climate modelling, predictions ofprecipitation have been shown to be highly uncertain(Roe & Baker, 2007; Anagnostopoulos et al., 2010;Kundzewicz & Stakhiv, 2010), with increasing uncer-tainty as one goes down in scale and moves to moreextreme events (Blöschl et al., 2007). In contrast,changes in temperatures predicted by climate modelsare usually considered more reliable. However, IPCCmodels predicted an increase in global air temper-atures over the past decade of about 0.2◦C (Meehlet al., 2007), whereas observations have shown differ-ent figures (Kerr, 2009). Kay et al. (2006) showed thatthe largest source of uncertainty is related to the struc-ture of GCM models, as also pointed out by Elshamy

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

208 Giuliano Di Baldassarre et al.

et al. (2009a), followed by emission scenarios andhydrological modelling. Yet, it is worth noting thatbias corrections, applicable for the known past, mightfail for the unknown future as the behaviour (in termsof the required corrections) might be relatively differ-ent from the past. In addition, none of the discussedstudies included the uncertainty of the hydrologicalmodels. Nevertheless, one should not underestimatethe uncertainty of the models (e.g. Beven, 2006;Ndomba et al. 2008; Ndomba & Birhanu, 2008; DiBaldassarre et al., 2010a) used to simulate the futurehydrological cycle. In fact, it is well known thathydrological models are difficult to calibrate and vali-date (e.g. Beven & Binley, 1992; Di Baldassarre et al.,2009), and their prediction capability in a possiblywarmer planet is impossible to evaluate, because ofthe lack of temporal and spatial data applicable todifferent climate conditions (Loaiciga et al., 1996).

6 BEST PRACTICE

This critical review has shown that a large number ofstudies have been carried out in the RNB using cli-matic model output as the input of hydrological mod-els in order to project future hydrological regimes.Much less systematic work has been done to estimateuncertainty (Koutsoyiannis et al., 2007). However,acknowledgement of the uncertainties is fundamen-tal to the decision-making process. Climate changepredictions based on a single simulation ignore thelarge uncertainties, and will provide misleading infor-mation to the public and decision makers, and/ormay lead to wrong decisions. However, ensemblepredictions that reflect uncertainties from scenarios,climate models, downscaling techniques and hydro-logical models tend to predict nearly every potentialchange and may hamper fact-based decision support(Fig. 1).

This points out the need for best practice in cli-mate change impact studies, which go beyond theIPCC Technical Guidelines for Assessing ClimateChange Impacts and Adaptations (IPCC, 1994). Thispractice should include the following requirements:

(a) results should not be presented in a simplifiedway assuming a one-way cause–effect relation-ship;

(b) ensembles of several climate model predictionsshould be used to reflect their large variability;

(c) the performance of the models applied to histor-ical data should be provided;

(d) appropriate downscaling techniques should beused and the underlying assumptions should bereported; and

(e) appropriate uncertainty analysis techniquesshould be applied to the modelling exercise.

The latter should reflect the likelihood of the predic-tions, and go beyond traditional analysis of modelparameter uncertainty towards predictive models ofuncertainty (e.g. Solomatine & Shrestha, 2009).

The most practical means of quantifying uncer-tainty in a complex transient system like the changingclimate and consequent impacts on many socio-economic sectors including hydrology is a proba-bilistic approach (Collins et al., 2006). Applicationsof climate modelling to assess uncertain futureimpacts on hydrology and other socio-economic sec-tors should as far as possible use ensemble approachbecause ensemble predictions have clear probabilisticadvantage (Hagedorn et al., 2005).

Very recently, Blöschl & Montanari (2010) pro-vided an inspiring idea to be part of such a code ofgood practice: impact studies should not only presentthe assumptions, results and interpretation, but alsoprovide a clear explanation of why certain changes arepredicted by the applied models. The idea is that weshould not trust that the results are valid unless weunderstand why an impact study predicts changes ina given hydrological variable (Blöschl & Montanari,2010).

Furthermore, to plan appropriate adaptation andmitigation measures, it is important to consider alsonon-climatic factors, such as population growth, andchanges in per capita and agricultural water demand(Conway, 1996; Vörösmarty et al., 2000). In fact,economically- and demographically-driven growth indemand generally leads to large changes in per capitawater availability and often outweighs climatically-induced changes, especially on the short and mediumterms. For instance, fluctuations of the level of LakeVictoria in the Nile basin, such as the large declinein levels between 2005 and 2007, have impactedupon lake-shore communities in Kenya, Tanzania andUganda (Conway, 2005; Pearce, 2006). Both climatevariability and management of the lake outflow inUganda for hydroelectric power are likely to havebeen responsible for the recent decline in lake levels(Pearce, 2006; Sutcliffe & Petersen, 2007). Moreover,Di Baldassarre et al. (2010b) showed that the dra-matic increase of flood fatalities in the African conti-nent is mostly due to the growth of urban populations

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

Future hydrology and climate in the River Nile basin: a review 209

and, in particular, human settlements in flood-proneareas.

Thus, given the large uncertainties in climateprediction, and the fact that climate is often onlyone of the factors influencing adaptation decisions,an approach that avoids heavy reliance on climateprediction and assesses the robustness of adaptationdecisions to a range of plausible futures is prefer-able (Dessai & Hulme, 2007; Goulden et al., 2009).Stakhiv (1998) recommends that a “no-regret” strat-egy could be provided by the use of the adaptive man-agement principle for water resource management.

7 CONCLUSIONS

Several studies on the impacts of climate change onthe hydrology of the River Nile, have been reportedand critically discussed herein by considering bothdata-based investigations (trend analysis) and model-based studies. The discussion focused on two con-trasting issues. On the one hand, there is a need tobetter recognize and characterize the uncertainty ofclimate change impacts on the hydrology of the RNB;on the other hand, there is the necessity to effec-tively support decision makers and propose adap-tation strategies and measures. For the latter, it iscrucial that these studies consider the uncertaintiesand assumptions and are based on an understand-ing of the full DPISR chain (Drivers – Pressures –Impacts – Status – Responses), including other globalchanges. It is important to follow a code of good prac-tice, the main principles of which have been outlinedabove. Moreover, given that (a) climate projectionsare uncertain and (b) very often demographically- andeconomically-induced growth in demand for water isexpected to outweigh climate-driven changes, adapta-tion in the water sector should focus on building adap-tive capacity and no-regret type activities in responseto multiple factors (Goulden et al., 2009). The com-bination of uncertainty and the need to consider non-climate factors is leading to a greater emphasis onflexibility, adaptive management and responses thatare robust to uncertainty (Dessai & Hulme, 2007). Inour future studies of the RNB we will follow explicitlyand further develop these considerations.

Acknowledgements The activities related to thispaper have been funded by the UNESCO-IHEPartnership Research Fund (UPaRF) within theACCION project (Adaptation to Climate ChangeImpact On the Nile river basin). The co-editor,Demetris Koutsoyiannis, and the two referees, Harry

Lins and Alberto Montanari, are also acknowledgedfor their useful and constructive comments.

REFERENCES

Allamano, P., Laio, F. & Claps, P. (2009) Global warming increasesflood risk in mountainous areas. Geophys. Res. Lett. 36(24),L24404.

Anagnostopoulos, G. G., Koutsoyiannis, D., Christofides, A.,Efstratiadis, A. & Mamassis, N. (2010) A comparison of localand aggregated climate model outputs with observed data.Hydrol. Sci. J . 55(7), 1094–1110.

Awulachew, S. B., McCartney, M., Steenhuis, T. S. & Ahmed, A. A.(2008) A review of hydrology, sediment and water resourceuse in the Blue Nile Basin. Colombo, Sri Lanka: InternationalWater Management Institute, IWMI Working Paper131.

Bates, B. C., Kundzewicz, Z. W., Wu, S. & Palutikof, J. P. (eds) (2008)Climate change and water. Geneva: Intergovernmental Panel onClimate Change Secretariat, Technical Paper.

Beran, J. (1994) Statistics for Long-Memory Processes. New York:Chapman & Hall, Monographs on Statistics and AppliedProbability, vol. 61.

Beven, K. (2006) A manifesto for the equifinality thesis. J. Hydrol.320, 18–36.

Beven, K. J. & Binley, A. M. (1992) The future of distributed models:model calibration and uncertainty prediction. Hydrol. Processes6, 279–298.

Blöschl, G., Ardoin-Bardin, S., Bonell, M., Dorninger, M.,Goodrich, D., Gutknecht, D., Matamoros, D., Merz, B., Shand,P. & Szolgay, J. (2007) At what scales do climate variability andland cover change impact on flooding and low flows? Hydrol.Processes 21, 1241–1247.

Blöschl, G. & Montanari, A. (2010) Climate change impacts—throwing the dice? Hydrol. Processes 24, 374–381.

Bowden, J., Semazzi, F. H. M., Anyah, R., Onol, B. & Tawfik, A.(2009) Decadal temperature, rainfall and hydrological trendsover the Greater Horn of Africa. Raleigh: North Carolina StateUniversity, http://climlab02.meas.ncsu.edu/Pubs/journals/AMS_climatechange.pdf

Collins, M., Booth, B. B. B., Harris, G. R., Murphy, J. M. & Sexton,D. M. H. (2006) Towards quantifying uncertainty in transientclimate change. Climate Dynamics 27, 127–147.

Conway, D. (1996) The impacts of climate variability and future cli-mate change in the Nile Basin on water resources in Egypt. Int.J. Water Resour. Devel. 12, 277.

Conway, D. (2005) From headwater tributaries to international riverbasin: adaptation to climate variability and change in the RiverNile basin. Global Environ. Change 15, 99–114.

Conway, D. & Hulme, M. (1993) Recent fluctuations in precipitationand runoff over the Nile sub-basins and their impact on mainNile discharge. Climatic Change 25, 127–151.

Conway, D., Mould, C. & Bewket, W. (2004) Over one century ofrainfall and temperature observations in Addis Ababa, Ethiopia.Int. J. Climatol. 24, 77–91.

Dessai, S. & Hulme, M. (2007) Assessing the robustness of adap-tation decisions to climate change uncertainties: a case studyon water resources management in the east of England. GlobalEnviron. Change 17(1), 59–72.

Di Baldassarre, G. & Montanari, A. (2009) Uncertainty in riverdischarge observations: a quantitative analysis. Hydrol. EarthSystem Sci. 13, 913–921.

Di Baldassarre, G., Montanari, A., Lins, H., Koutsoyiannis, D.,Brandimarte, L. & Blöschl, G. (2010a) Flood fatalities inAfrica: from diagnosis to mitigation. Geophys. Res. Lett.doi:10.1029/2010GL045467.

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

210 Giuliano Di Baldassarre et al.

Di Baldassarre, G., Schumann, G. & Bates, P. D. (2009) A tech-nique for the calibration of hydraulic models using uncer-tain satellite observations of flood extent. J. Hydrol. 367,276–282.

Di Baldassarre, G., Schumann, G., Bates, P., Freer, J. & Beven, K.(2010b) Floodplain mapping: a critical discussion on deter-ministic and probabilistic approaches. Hydrol. Sci. J . 55(3),364–376.

Dibike, Y. B. & Coulibaly, P. (2006) Temporal neural networks fordownscaling climate variability and extremes. Neural NetworksSpecial Issue on Earth Sciences and Environmental Applica-tions of Computational Intelligence, Cherkassky, Krasnopolsky,Solomatine, Valdes, eds) 19, 135–144.

Elshamy, M. E., Sayed, M. A.-A. & Badawy, B. (2009a) Impactsof climate change on Nile flows at Dongola using statisticallydownscaled GCM scenarios. Nile Water Science & EngineeringMagazine, Special issue on Water and Climate 2, 1–14.

Elshamy, M. E., Seierstad, I. A. & Sorteberg, A. (2009b) Impacts ofclimate change on Blue Nile flows using bias-corrected GCMscenarios. Hydrol. Earth System Sci. 13(5), 551–565.

Elshamy, M. E. & Wheater, H. S. (2009) Performance assessment ofa GCM land surface scheme using a fine-scale calibrated hydro-logical model: an evaluation of MOSES for the Nile Basin.Hydrol. Processes doi:10.1002/hyp.7298.

Fowler, H. J., Blenkinsop, S. & Tebaldi, C. (2007) Linking climatechange modelling to impacts studies: recent advances in down-scaling techniques for hydrological modeling. Int. J. Climatol.27, 1547–1578.

Georgakakos, A. (2007) Decision support systems for integratedwater resources management with an application to the Nilebasin. Ch. 5 in: Topics on System Analysis and Integrated WaterResources Management (A. Castelletti & R. Soncini-Sessa,eds). Amsterdam: Elsevier.

Giorgi, F. & Mearns, L. O. (1999) Introduction to special section:regional climate modeling revisited. J. Geophys. Res. 104,6335–6352.

Githui, G., Gitau, W., Mutua, F. & Bauwens, W. (2008) Climatechange impact on SWAT simulated streamflow inwesternKenya. Int. J. Climatol. doi:10.1002/joc.1828.

Goulden, M., Conway, D. & Persechino, A. (2009) Adaptation to cli-mate change in international river basins in Africa: a review.Hydrol. Sci. J . 54(5), 805–828.

Hagedorn, R., Doblas-Reyes, F. & Palmer, T. N. (2005) The ratio-nale behind the success of multi-model ensembles in seasonalforecasting. I. Basic concept. Tellus 57A, 219–233.

Hamilton, K. & Ohfuchi. W. (2008) High Resolution NumericalModelling of the Atmosphere and Ocean. Berlin: Springer-Verlag.

Hassan, F. A. (2004) Historical Nile floods and their implications forclimate change. Science 212, 1142–1145.

Hattermann, F., Krysanova, V., Post, J., Dworak, Th., Wrobel, M.,Kadner, S. & Leipprand, A. (2008) Understanding conse-quences of climate change for water resources and water-relatedsectors in Europe. In: The Adaptiveness of IWRM, AnalysingEuropean IWRM Research (J. G. Timmerman, C. Pahl-Wostl &J. Moltgen, eds), 89–112. London: IWA Publishing.

Hay, L. E., Wilby, R. L. & Leavesley, G. H. (2007) A comparison ofdelta change and downscaled GCM scenarios for three moun-tainous basins in the United States. J. Am. Water Resour. Assoc.36(2), 387–397.

Hurst, H. E. (1951) Long term storage capacity of reservoirs. Trans.ASCE 116, 770–779.

Ines, A. V. M. & Hansen, J. W. (2006) Bias correction of dailyGCM rainfall for crop simulation studies: Agric. For. Met. 138,44–53.

IPCC (Intergovernmental Panel on Climate Change) (1994) Technicalguidelines for assessing climate change impacts and adapta-tions. WMO, UNEP. ISBN 0-904-813–118.

IPCC (Intergovernmental Panel on Climate Change) (2007) IPCCFourth Assessment Report: Climate Change 2007 (AR4). CoreWriting Team (R. K. Pachauri & A. Reisinger, eds). Geneva:IPCC.

Kang, H.-S. & Hong, S.-Y. (2008) Sensitivity of the simulatedEast Asian summer monsoon climatology to four convec-tive parameterization schemes. J. Geophys. Res. 113, D15119.doi:10.1029/2007JD009692.

Kay, A. L., Jones, R. G. & Reynard, N. S. (2006) RCM rainfall for UKflood frequency estimation. I. Method and validation. J. Hydrol.318, 151–162.

Kerr, R. A. (2009) What happened to global warming? Scientists sayjust wait a bit. Science 326, 28–29.

Kiehl, J. T., Hack, J. J., Bonan, G. B., Boville, B. A., Breigleb, B. P.,Williamson, D. & Rasch, P. (1996) Description of the NCARcommunity climate model (CCM3).Boulder, CO: NationalCenter for Atmospheric Research, Tech. Rep. NCAR/TN-420+STR.

Kigobe, M. (2009) Modelling the effects of land use change andclimate variability on the hydrology of the Upper Nile. PhDThesis, Department of Civil Engineering, University of London,London, UK.

Kigobe, M., McIntyre, N. & Wheater, H. S. (2007) Climatechange and anthropogenic impacts on the Upper Nile.HYDRA Postgraduate Symposium (London, UK). Available at:http://www3.ic.ac.uk/pls/portallive/docs/1/44939698.pdf

Koutsoyiannis, D. (2002) The Hurst phenomenon and fractionalGaussian noise made easy. Hydrol. Sci. J . 47(4), 573–595.

Koutsoyiannis, D. (2003) Climate change, the Hurst phenomenon, andhydrological statistics. Hydrol. Sci. J . 48(1), 3–24.

Koutsoyiannis, D., Efstratiadis, A. & Georgakakos, K. (2007)Uncertainty assessment of future hydroclimatic predictions: acomparison of probabilistic and scenario-based approaches. J.Hydromet. 8(3), 261–281.

Koutsoyiannis, D., Huaming, Y., Georgakakos, A. (2008) Medium-range flow prediction for the Nile: a comparison of stochasticand deterministic methods. Hydrol. Sci. J . 53(1), 142–164.

Koutsoyiannis, D., Kundzewicz, Z. W. K., Watkins, F. & Gardner, C.(2010) Something old, something new, something red, some-thing blue—editorial. Hydrol. Sci. J . 55(1), 1–3.

Koutsoyiannis, D., Montanari, A., Lins, H. F. & Cohn, T. A.(2009) Climate, hydrology and freshwater: towards an inter-active incorporation of hydrological experience into climateresearch—DISCUSSION of “The implications of projected cli-mate change for freshwater resources and their management.”Hydrol. Sci. J . 54(2), 394–405.

Kundzewicz, Z. W. & Stakhiv, E. Z. (2010) Are climate models“ready for prime time” in water resources management appli-cations, or is more research needed? Editorial. Hydrol. Sci. J .55(7), 1085–1089.

Loaiciga, H., Valdes, J. B., Vogel, R. & Garvey, J. (1996) Globalwarming and the hydrologic cycle. J. Hydrol. 174, 83–127.

Meehl, G. A. et al. (2007) Global climate projections. In: ClimateChange 2007: The Physical Science Basis. Contribution ofWorking Group I to the Fourth Assessment Report of theIntergovernmental Panel on Climate Change (S. Solomon, D.Qin, M. Manning, Z. Chen, M. Marquis, K. B. Averyt, M.Tignor & H. L. Miller, eds). Cambridge: Cambridge UniversityPress.

Mitchell, T. D. & Jones, P. D. (2005) An improved method ofconstructing a database of monthly climate observations andassociated high-resolution grids. Int. J. Climatol. 25, 693–712.

Mohamed, Y. A., van den Hurk, B., Savenije, H. H. G. &Bastiaanssen, W. G. M. (2005) Impact of the Sudd wetlandon the Nile hydroclimatology. Water Resour. Res. 41(8),W08420.

Montanari, A., Rosso, R. & Taqqu, M. S. (1997) Fractionallydifferenced ARIMA models applied to hydrologic time series:

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012

Future hydrology and climate in the River Nile basin: a review 211

identification, estimation and simulation. Water Resour. Res. 33,1035–1044.

Mukubwa, A. H. (2008) Modelling of sediment yield and deposi-tion in the planned Rusumo hydropower reservoir using SWATmodel. MSc Thesis, UNESCO-IHE, Delft, The Netherlands.

Mutemi, J. N., Ogallo, L. A., Krishnamurti, T. N., Mishra, A. K. &Vijaya Kumar, T. S. V. (2007) Multimodel based superensembleforecasts for short and medium range NWP over various regionsof Africa. Met. Atmos. Phys. 95, 87–113.

Nawaz, R. & Bellerby, T. (2007) Blue Nile flow sensitivity toprojected climatic change until 2100. In: Changes in WaterResources Systems: Methodologies to Maintain Water Securityand Ensure Integrated Management (Proc. Symp. HS3006 atIUGG2007, Perugia, July 2007). (N. van de Giesen, Xia Jun, D.Rosbjerg & Y. Fukushima, eds), 284–294. Wallingford: IAHSPress, IAHS Publ. 315.

Ndomba, P. M. & Birhanu, B. Z. (2008) Problems and prospectsof SWAT model applications in nilotic country catchments: areview. Nile Basin Water Engineering Scientific Magazine 1,41–52.

Ndomba, P. M., Mtalo, F. W. & Killingtveit, A. (2008) SWAT modelapplication in a data scarce tropical complex catchment inTanzania. Phys. Chem. Earth 33, 626–632.

NMA (National Meteorological Agency of Ethiopia) (2007) Climatechange national adaptation program of action (NAPA) ofEthiopia. Addis Ababa: NMA.

Pappenberger, F., Beven, K. J., Hunter, N., Gouweleeuw, B., Bates, P.,de Roo, A. & Thielen, J. (2005) Cascading model uncertaintyfrom medium range weather forecasts (10 days) through arainfall–runoff model to flood inundation predictions withinthe European Flood Forecasting System (EFFS). Hydrol. EarthSystem Sci. 9, 381–393.

Pearce, F. (2006) Uganda pulls plug on Lake Victoria. New Scientist2538, 12.

Randall, D. A., Wood, R. A. & Bony, S. (2007) Climate modelsand their evaluation. In: Climate Change 2007: The PhysicalScience Basis. Contribution of Working Group I to the FourthAssessment Report of the Intergovernmental Panel on ClimateChange (S. Solomon, D. Qin, M. Manning, Z. Chen, M.Marquis, K. B. Averyt, M. Tignor & H. L. Miller, eds),589–662. Cambridge: Cambridge University Press.

Roe, G. H. & Baker, M. B. (2007) Why is climate sensitivity sounpredictable? Science 318(5850), 629–632.

Sayed, M. A.-A. & Nour El-Din, M. & Nasr, F. (2004) Impacts ofglobal warming on precipitation patterns on the Nile basin.Second Regional Conference on Action Plans for IntegratedDevelopment (Cairo, Egypt, 12–15 April 2004).

Schoof, J. T., Shin, D. W., Cocke, S., LaRow, T. E., Limb, Y.-K. &O’Brien, J. J. (2009) Dynamically and statistically downscaledseasonal temperature and precipitation hindcast ensembles forthe Southeastern USA. Int. J. Climatol. 29, 243–257.

Shahin, M. M. (2002) Hydrology and Water Resources of Africa, 276,287–288. Berlin: Springer-Verlag. ISBN 1402-00866–X.

Soliman, E. S. A., Sayed, M. A.-A. & Jeuland, M. (2009)Impact assessment of future climate change for the BlueNile basin using a RCM nested in a GCM. Nile WaterScience & Engineering Special Issue on Water and Climate 2,31–38.

Soliman, E. S. A., Sayed, M. A.-A., Nour El-Din, M. & Samy, G.(2008) Integration of NFS with regional climate model tosimulate the Nile basin hydro-climatology. Nile Basin WaterEngineering Scientific Magazine 1, 75–85.

Solomatine, D. P. & Shrestha, D. L. (2009) A novel method to estimatemodel uncertainty using machine learning techniques. WaterResour. Res. 45, W00B11. doi:10.1029/2008WR006839.

Stakhiv, E. Z. (1998) Policy implications of climate change impactson water resources management. Water Policy 1(2), 159–175.

Sutcliffe, J. V. & Parks, Y. P. (1999) The Hydrology of the Nile.Wallingford: IAHS Press, Special Publ. 5.

Sutcliffe, J. V. & Petersen, G. (2007) Lake Victoria: derivation of acorrected natural water level series. Hydrol. Sci. J . 52(6), 1316–1321.

Tadele K. & Forch G. (2007) Impact of land use/land coverchange on streamflow: a case study of Hare watershed,Ethiopia, 80–84. FWU (Research Institute for Water Resourcesand Environment). FWU Water Resources Publications no.06/2007. ISSN 1613–1045.

Tate, E., Sutcliffe, J., Conway, D. & Farquharson, F. (2004) Waterbalance of Lake Victoria: update to 2000 and climate changemodeling to 2100. Hydrol. Sci. J . 49(4), 563–574.

Trenberth, K. (1992) Climate System Modeling. Cambridge:Cambridge University Press.

Uhlenbrook, S. (2009) Climate and man-made changes and theirimpacts on catchments. In: Water Policy 2009, Water as aVulnerable and Exhaustible Resource (P. Kovar, P. Maca & J.Redinova, eds) (Proc. Joint Conf. of APLU and ICA, 23–26June 2009, Prague, Czech Republic), 81–87. ISBN 978-80-213-1944–8.

Vicuna, S. & Dracup, J. A. (2007) The evolution of climate changeimpact studies on hydrology and water resources in California.Climatic Change. doi:10.1007/s10584-006-9207–2.

Vörösmarty, C. J., Green, P., Salisbury, J. & Lammers, R. B. (2000)Global water resources: vulnerability from climate change andpopulation growth. Science 289(5477), 284–288.

Wilby, R. L. (2010) Evaluating climate model outputs for hydrologi-cal applications—Opinion. Hydrol. Sci. J . 55(7), 1090–1093.

Wilby, R. L., Beven, K. J. & Reynard, N. S. (2008) Climate changeand fluvial flood risk in the UK: more of the same? Hydrol.Processes 22, 2511–2523.

Wilby, R. L., Troni, J., Biot, Y., Tedd, L., Hewitson, B. C., Smith,D. M. & Sutton, R. T. (2009) A review of climate risk informa-tion for adaptation and development planning. Int. J. Climatol.29, 1193–1215.

Wing, H., Cheung, A., Gabriel, B. S. & Singh, A. (2008). Trends andspatial distribution of annual and seasonal rainfall in Ethiopia.Int. J. Climatol. 28(13), 1723–1734.

Dow

nloa

ded

by [

194.

204.

215.

228]

at 0

2:41

17

Janu

ary

2012