Assessment of surface water resources in Great Ruaha River

Tanzania Journal of Engineering and Technology (Tanz. J. Engrg. Technol.), Vol. 38 (No. 1), June 2019 27

Vol. 38(1), pp. 27-46, June 2019ISSN 1821-536X (print)ISSN 2619-8789 (electronic)

Tanzania Journal of Engineering and TechnologyCopyright © 2019 College of Engineering andTechnology, University of Dar es Salaam

Full Length Research Paper

Assessment of surface water resources in Great Ruaha River sub-basinTanzania: Application of WEAP model for water allocation and utilization

impacts analysis

Rose Gervas, Deogratias M.M. Mulungu* and Joel K. Nobert

Department of Water Resources Engineering, College of Engineering and Technology,University of Dar es Salaam, P.O. Box 35131, Dar es Salaam, Tanzania.

*Corresponding author: [email protected]; [email protected]

ABSTRACT

The Great Ruaha River (GRR) sub-basin is one of the most important waterways inTanzania as it supports important economic activities. The sub-basin is progressively facedwith an inevitable situation of increasing water demand among competing users while thequantity and quality of water is diminishing. The focus of this study was to assess allocationof existing (2012) and future (up to 2025) quantities of surface water in the GRR sub-basinwith consideration of priorities given in the Tanzania Water Resources Management Act,2009 in the order: domestic, environment, agriculture, livestock and non-domestic.Simulation of water allocation scenarios of irrigation expansion (IE) and implementation ofenvironmental flow requirements (EWD) and their impacts on met demands was done byusing WEAP model. Results showed that under current and future conditions, availablestreamflows are sufficient for the first two water use priorities. Implementation of EWD andIE scenarios will change demands in comparison with reference scenario by 80%, -38% and45% and shortages by 147%, 123% and 13% in Little Ruaha, Ndembera and Kisigocatchments respectively. To eliminate water shortages, construction of reservoirs, use ofalternative supply sources (especially in agriculture) and water demand managementmeasures are recommended.

Keywords: Great Ruaha River, Rufiji basin, scenarios, water allocation, water shortage,WEAP model.

INTRODUCTION

Background Information

Recently it is commonly recognized thatthe basic human needs such as food,shelter and well-being are being affectedby human activities on the environment.Climate change and increasing use of riverwater resources have resulted intosignificant changes in river flow regimes.As a result, aquatic life and ecology havemainly been affected. Water is an

important aspect of development,livelihood and ecosystems. However, thereare water management challenges due togrowing water demands (Kashaigili et al.,2005).

The Great Ruaha River (GRR) is a majorand important tributary in the Rufiji Basin,Tanzania. It contributes to water flow forhydropower production, irrigation andlivelihood to the rural areas. Hydropowerproduction in the river accounts for 50% ofthe Tanzania's installed capacity (Kadigi et

Assessment of surface water resources in Great Ruaha River sub-basin Tanzania: Application of WEAPmodel for water allocation and utilization impacts analysis

28 Tanzania Journal of Engineering and Technology (Tanz. J. Engrg. Technol.), Vol. 38 (No. 1), June 2019

al., 2004). The river is also important forecology of the Usangu wetlands and theRuaha National Park (RNP) (Kashaigili etal., 2005; Kashaigili et al., 2006).

There is great interest on the GRR foragricultural development, but previousexperience indicated severe water shortageat Mtera reservoir resulting into powershortages and rationing in the country.Also, there were experienced low to dryflows since 1993 in GRR, which is themajor source of water for the RNP that iscovering 30% of basin area (SMUWC,2001). As a result, wildlife and aquatic lifeis impaired leading to threats to tourismattractions, irrigation water supply andlivestock grazing areas. This dryness hashad implications on the water allocated tolivelihood of the people, the economy andsignificant biodiversity in the Ruahacatchment area (Kashaigili et al., 2005).Surface water is preferable thangroundwater for domestic and agriculturalutilization due to challenges in obtaininggroundwater potential data (Lankford,2002). The total irrigated area located inthe upland areas of the basin varies from20,000 to 40,000 ha as a function ofannual rainfall. This area includes large-scale, state-owned farms, traditionalsmallholder; improved smallholder, andsmallholder peripheral to the state farm.

Increasing economic activity, especially inirrigated agriculture (notably maize andrice) and in the production of energy, isplacing huge pressure on the waterresources of the catchment. Overallocation and heavy abstraction of waterhas resulted in dramatically reduced in-stream flows. There is increase inupstream diversions and consequentreduced inflows to the Usangu plains thatresulted into conflicts between upstreamand downstream communities. Mwakalila(2005) found out that, the main causes ofwater conflicts are: insufficient water(40%), unequal water distribution (30%)and water mismanagement (30%). Further

there is no proper mechanism formonitoring of water permits which havebeen issued so as to know if theabstractions are according to the givenpermits. There is also increase in illegalwater abstractions for different activitiesspecifically irrigation. Furthermore, thereare noticeable river flow reductions in thecentral Usangu Plains, Ruaha NationalPark, and to the Mtera and Kidatureservoirs due to increase in expansion ofirrigated land. Despite all these challenges,there is no any current scientificinformation on water availability andanalysis of water allocation for differentusers according to their demand and wateravailability for future expansions.

Chanzi (2017) performed water allocationusing WEAP in Kilombero sub-basin ofthe Rufiji basin and Munkyala (2017)analyzed water allocation scenarios usingthe Nile Basin Decision Support System inNdembera sub-basin of the GRR. WhileKilombero sub-basin is outside the GRRbasin and Ndembera just a portion of theGRR basin, there is a need to have adedicated study for the entire GRR sub-basin following its importance in the RufijiRiver basin. Due to this, decision makersneed to be informed on the availableresource, meeting the demands and toolsor mechanisms for water allocations in theGRR sub-basin. Variable and changingclimatic conditions, notably in the form ofincreasing temperatures, associated withincreases in evaporation and transpiration,are affecting both the supply of, anddemand for, water and energy. Also,population growth is driving an increasingdemand for agricultural land, clearance ofnatural vegetation and poor landmanagement practices, leading to soilerosion and the degradation of watersources, increased water abstraction andcompetition for water resources.

For that matter, the study entails to assessthe water resources allocation so as toimprove water resources use and

Rose Gervas, Deogratias M.M. Mulungu and Joel K. Nobert


management in the GRR sub-basin, anddevelop alternative plausible futurescenarios in order to assess impacts ofdifferent development options for waterallocation. This was implementedintegrally using computer based decisionsupport software called Water Evaluationand Planning System (WEAP). Ultimately,the study is anticipated to contribute tosustainable management, development,allocation and utilization of the waterresources in the GRR sub-basin byproviding information critical to decisionmaking.

METHODS AND MATERIALS

Description of the Study Area

The GRR sub-basin (Figure 1) is located inthe southwest of Tanzania approximatelylatitudes 604’ and 90410 South, andlongitudes 33040’ and 37041’ East(Kashaigili et al., 2005). The GRR sub-basin has three main river catchments:Great Ruaha, Kisigo and Little Ruahacatchments (Figures 1 & 3). The lower partof GRR sub-basin includes the Mtera andKidatu reservoirs. The Great Ruahacatchment and the lower part of GRR sub-basin are part of the main river whileKisigo and Little Ruaha catchments formtributaries to the GRR River which flowsto join Rufiji River downstream. The GRRis an important and main tributary of theRufiji River draining an area of about85,554 km2. It lies within the eastern armof the Rift Valley, marked by distinctescarpments in the southern and easternparts and forms the upper catchment of theGRR. The Rufiji River Basin (RRB) is thelargest drainage basin in Tanzania,

covering some 177,000 km2 or about 18%of the Tanzanian mainland. Located in theupper part of the Great Ruaha catchmentare the Usangu plains with a total area of20,811 km2, which is about 24% of theGRR sub-basin and 12% of the RufijiBasin. There are many rivers draining intothe plains while the major ones areMbarali, Kimani, Chimala and Ndembera.The small rivers are: Umrobo, Mkoji,Lunwa, Mlomboji, Ipatagwa, Mambi andMswiswi rivers. After these tributaries jointhe GRR, the river passes through a naturalweir at Nyaluhanga and it supplies waterto the Eastern wetland and the Ihefuswamp. The GRR is the major source ofwater as it flows through RNP to the Mterareservoir. After confluence of GRR andLittle Ruaha River, the river flows intoMtera reservoir. At Mtera, the power plantwith 80 MW installed capacity generatespower for the national grid. The outflowfrom Mtera reservoir goes into Kidatu afterconfluence with Lukosi and Yovi Riverssupplying water for the 204 MW installedcapacity Kidatu power plant. The GRRthen flows into Kilombero plains afterconfluence with Kilombero River and itforms a major tributary of Rufiji River(Kashaigili et al., 2005).

The GRR which in the past was perennial,since early 1990s has consistently ceasedflowing in the dry season largely due tohuman diversions of water upstream of theUsangu wetlands (SMUWC, 2001;Lankford et al., 2004). Furthermore, since1993, outflow from the Ihefu, the perennialswamp has ceased for prolonged periodsduring every dry season (Kashaigili et al.,2005).



Figure 1: Location of GRR sub-basin in Tanzania

Data and Methods

To assess water allocation in GRR, Gervas(2013) estimated current and future waterdemands and water availability fordevelopment activities based on masterplans in the Ministry of Agriculture, FoodSecurity and Cooperatives (MAFC), andthe Ministry of Water. Also, in order toperform water allocations for differentuses, the WEAP model was calibrated andused for simulations of water use scenariosin the GRR sub-basin. The input data waschecked for outliers and preprocessed to asuitable format before use for WEAPmodeling.

Assessment of water allocation scenariosin the sub-basin

Kashaigili et al. (2005) found out that theneed to achieve efficient allocation ofwater resources in Tanzania is becomingimperative, as water scarcity increases.The Tanzania Water ResourcesManagement Act, 2009 gives waterallocation priorities in the order: domestic,environment, agriculture, livestock andnon-domestic. This was adhered in thisstudy. Due to the fact that water resourcesmanagement at the basin level involvesdifferent complexities, computer modelswere used to help allocation of availablewater resources among the different users.There are a number of water resourcesallocation models (e.g. WEAP, StockholmEnvironmental Institute (SEI), Sieber et



al., 2005); Mike Basin (DHI, 2006);REsource ALlocation Model (REALM),(Perera et al., 2005) and Aquarius), and thedecision on which one to use depends onamong other things, on the level of naturalsystems representation complexity desired,data availability, cost and details desiredby the analyst. Water allocation andsimulation of future water demand andsupply using scenario in GRR sub-basinwas done by WEAP model, which uses awater balance principle and runs onmonthly basis (Figure 2). "The userrepresents the system in terms of itsvarious sources of supply (e.g. rivers,groundwater, and reservoirs), withdrawals,water demands, and ecosystemrequirements" (Sieber et al., 2002). WEAPmodel in this case was chosen as it is

comprehensive, straightforward and easy-to-use, and attempts to assist rather thansubstitute for the skilled planner. Datarequirements and availability was alsoconsidered during selection of the model.WEAP can perform many functions atonce since it uses a database that providesa system for maintaining and manipulationof water demand and supply information.One of the first water allocation studies inTanzania was done by Mulungu and Taipe(2012) who successfully set up the WEAPmodel and performed water allocation inWami River sub-basin. Other studiesinclude that of Chanzi (2017) inKilombero sub-basin of the RRB andKishiwa et al. (2018) in Upper PanganiRiver sub-basin of the Pangani RiverBasin.

Figure 2: The conceptual structure of WEAP model (Source: Sieber et al., 2005)

WEAP application involves having arecent “baseline” year where wateravailability and demands are well known.Then the model is used in scenario mode

which expresses or simulates the "what if"conditions of the different waterdevelopment alternatives or options(Arranz and McCartney, 2007). When



water is limited, the algorithm isformulated to progressively restrict waterallocation to those demand sites given thelowest priority (Sieber et al., 2005).

Setting-up of the WEAP model in GRRsub-basin

The study area is setup using the physicalelements consisting of water demand andsupply system at their spatial locations,timeframe for analysis, units of variablesand hydrologic regimes. Accordingly, thestudy area was set and defined bygeographical boundaries and specific water

supply system (GRR sub-basin). The GRRsub-basin with an approximate area of85,554 km2 was achieved by setting theboundary function of the model, selectingTanzania from the global map in WEAPand adding an already delineatedcatchment with three catchments (Figure3) into the program. The point of focuswas the demand sites and the watersupplies in GRR sub-basin. The study areain WEAP is represented by clearlyoutlined database for water supply anddemand, which allows for storage,management and analysis of data.

Figure 3: GRR sub-basin with the sub-catchments in WEAP model

Time settings

The time setting includes defining thecurrent account year and the last year ofscenarios and the desired analysis. Then

WEAP runs the simulation on monthlybasis for the current to the last year ofaccounts. The current account is usuallythe most recent year where reliable andcomplete data are available, and it forms



the basis for future demands and scenariosprojection. The current account and lastyear of scenarios chosen for this study was2012 and 2025 respectively. The period ofanalysis started with the current accountyear (2012) and went to the last year ofscenarios (2025). A continuous monthlyanalysis was performed on each yearspecified for the period of analysis. Thestudy period was chosen because most ofthe Tanzania development plans areaiming at 2025.

Schematic setup

The model setup was done by defining thesupply and demand features of theresource system and the system wasconfigured. The demand sites that weremodelled by WEAP were indicated bynumbers showing water allocationpriorities (Figure 4).

Figure 4: WEAP schematic for GRR sub-basin with demand sites' priorities

Linking demand and supply

In WEAP, demands and supply are linkedby using transmission links. Transmissionlinks were used to convey water from riversupplies to demand sites with an allowancefor losses, physical capacity and otherconstraints on water conveyance. Thisincluded water supply for irrigation incatchments with irrigation. The links canalso be used in case of wastewater andtreatment plants. The assumption was thatrivers or their tributaries as the only watersource in the catchment used to satisfy the

demand sites determined in Gervas (2013).No any other supply source wasconsidered in this study due to limited timeand resources.

Water allocation priorities

The water allocation in WEAP was doneon the basis of the water demand prioritiesdefined and set for each of the demandsite. For each site in GRR sub-basin thepriority for withdrawal was established(Table 1) based on the Water Policy, 2002and Water Resources Management Act of



Tanzania, 2009. The sites with the highestpriorities start with number 1 in order andget water first depending on wateravailable. In this case during watershortage, the high priority site demands are

met first before other demands. Thisimplies that when there is plenty of waterto satisfy everyone, demand priorities areunnecessary.

Table 1: Water allocation priorities in GRR sub-basin

Sub catchment Urban Rural EFR Agriculture Livestock

Little Ruaha 1 2 3 4 5

Great Ruaha 1 2 3 4 5

Kisigo NA 1 2 NA 3

Note: NA means there was no demand in the catchment, hence no priority was set

Developed water allocation scenarios inthe GRR sub-basin

Water allocation assessment requires ascenario approach that considers the futurewith fundamentally different developmentand environmental assumptions andpolicies. Different scenarios that were builtin the WEAP model for GRR sub-basinare discussed hereunder. The model wasbuilt using observed data and it wasassumed that the implementation of theproposed scenarios will take place fromthe year 2013.

(a) Reference Scenario (2012-2025)

The reference scenario was a scenario thatuses the actual data to represents thechanges that are likely to occur in thefuture, in the absence of any new policymeasure, sometimes called a "business asusual" scenario. This was done to helppeople learn what likely could occur ifcurrent trend continues.

(b) Irrigation expansion scenario (2013-2025)

Irrigation expansion scenario was run toassess the changes to demand and supplypattern that will take place when theplanned expansion of irrigation agriculturetakes place. The area under irrigation is to

be increased by 4.7% (based on theTanzania Irrigation Master Plan, 2025).

(c) EFR implementation scenario (2013-2025)

Environmental flow requirement scenariowas going to answer the question of whatwill happen if minimum flow requirementis implemented in the rivers of the sub-basin (based on the Water ResourcesManagement Act, 2009). Environmentalflows are the flows of water in rivers andstreams that are necessary to maintainhealthy aquatic ecosystems. It is aninstream water demand designed to mimicthe natural condition in rivers.Accordingly, a low flow characteristic(Q95) estimated from Flow Duration Curve(FDC) was used as the environmentalwater requirement. The FDC is a plot ofdischarge against percent of time that aparticular discharge was equaled orexceeded. It is the most informative meansof displaying the complete range of riverdischarges from low flows to flood events(Pyrce, 2004). The area under the flowduration curve gives the average monthlyflow, the median monthly flow is the 50%value and the minimum flow to protect theriver is the 95% value (i.e. Q95). Meanmonthly values were used to estimate theQ95 using plotting position method. Thesame values of EFR from historical datawere used for the future.



Calibration of the WEAP model

The WEAP model was calibrated so as torepresent the natural water system in thesub-basin. McCartney et al. (2005)indicated that calibration is done byestimating historical water demandpatterns and simulating resultant flow.However, WEAP model does not have aninbuilt automatic calibration routine. Thisimplies that calibration was done manuallyby modifying the parameters and runningthe simulation until an optimum outcomeis arrived at. The model calibrationinvolves altering demand priorities inTable 1 and changing assumptions abouthistoric demand patterns (McCartney etal., 2005; Arranz and McCartney, 2007;Mulungu and Taipe, 2012) so as toimprove or maximize the fit between thesimulated and observed flows. Threemodel performance criteria were used toevaluate model performance: Nash-Sutcliffe efficiency (NSE), coefficient ofdetermination (R2) and correlationcoefficient (R). Moreover, it was assumedthat the hydrology of the catchment is notexpected to be significantly affected byclimate change within the planninghorizon for the study (i.e. 12 years up to

2025). Therefore, the future stream flowvariation was mirrored to the 2025 period.

The WEAP model calibration was doneusing observed flow data from twogauging stations located on the GRRtributaries (Little Ruaha River at 1KA31and Ndembera River at 1KA15). The datafor these stations were processed and itinvolved filling of the missing values andidentification of outliers. However, outlierswere not identified. Filling of the missingvalues was done by using correlationmethod. This method was selected amongother methods because the method issimple to use compared to rainfall-runoffmodels and it performs better at monthlytime scale. From the data records 1956-2009 for the flow stations, based on dataquality checks (i.e. no or very little datagaps) and water demand data availability,model calibration was done from 2002 to2007 and from 1992 to 1996 for Ndembera(1KA15) and Little Ruaha (1KA 31),respectively. The seasonal flow variationfor these two stations is presented inFigure 5, which showed higher flowcontribution for 1KA31 than 1KA15 in theGRR with the peak flows in April andMarch, respectively.

Figure 5: Seasonal flow variation at 1KA31 and 1KA15



After WEAP calibration, assessment ofwater allocation involved scenario analysisof demands against supply so as tounderstand the magnitude of watershortage. The above developed scenarioswere implemented in WEAP and analyzedin each sub-catchment of the GRR sub-basin.

RESULTS AND DISCUSSION

Assessment of Water Allocation

WEAP model calibration

During model calibration, it was revealedthat changing the demand priorities forurban and rural water uses had no effect on

the NSE and R2 if there is only one supplysource. This may be due to the fact that thehigh demand sectors such as agriculturewere not altered. Accordingly, the modelcalibration was done by estimatinghistorical water demand patterns withoutchanging assumptions about historicdemand patterns. Figure 6 and 7 shows thesimulated streamflows related to theobserved streamflows for the gaugingstations. The statistical model efficiencycriteria resulted into NSE, R2 and R of 0.8,0.89 and 0.94 in Ndembera (Figure 6) andNSE, R2 and R of 0.98, 0.84 and 0.9 inLittle Ruaha (Figure 7). These resultsindicated good model fit and then themodel was used for water allocation andscenarios evaluation.

Figure 6: Observed and simulated flow for 1KA15 (NSE=0.8, R2=0.89)



Figure 7: Observed and simulated flow for 1KA31 (NSE=0.98, R2=0.84)

Water Allocation in GRR Sub-basin

Little Ruaha Catchment

The below are the results for Little RuahaRiver catchment (Figure 1 & 3):

(a) Current unmet demand

The analysis using WEAP model showedthat from June-November water supply isinsufficient to meet the demands for waterin the GRR sub-basin. The high unmetdemands are in August-October, whichwas mostly affected by water demand for

agriculture and environment. The reasonfor high unmet demands is that agriculturewas assigned a priority of 5 afterenvironment, livestock, rural and urbanwater users. The shortages in theagriculture are due to high amount ofwater required for environment (4th

priority) which had to be satisfied beforeagriculture (5th priority). The total unmetdemand in the year 2012 is 104 Mm3 ofwhich 94.43 Mm3 (88%) is experienced onagriculture, 12.83 Mm3 (12%) onenvironment and 0.1 Mm3 (0.09%) onlivestock (Figure 8).



Figure 8: Unmet demand for the year 2012(Note: Agr - agriculture, Env - environment, Liv - livestock)

(b) Irrigation expansion scenario

Irrigation expansion (IE) scenario wasdone to determine the impacts ofincreasing the irrigable land from 6,000 to11,562 ha on water demand and supply.Annual average water demand for thisscenario increased from 760 Mm3 inreference scenario to 771.5 Mm3 (1.5%increase) in irrigation expansion scenario.The annual average unmet water demandincreased from 179.5 Mm3 in referencescenario to 193.5 Mm3 in irrigationexpansion scenario (Figure 9). Monthlyaverage unmet water demand increasedfrom 15 to 16.1 Mm3 (Figure 9). Theresults identified that water shortageincreased by 7.8 % (annual mean) and7.35% on monthly, which impactnegatively other water users in the sub-catchment which can lead to low economicdevelopment.

(c) EFR implementation scenario

Implementation of EFR increasesshortages in the catchment as the amountof water available for agriculture, livestock

and non-domestic abstractions is reduced.A water allocation priority of 3 was givento EFR after domestic (urban and rural).The annual average water demand for thisscenario increased from 760 Mm3 inreference scenario to 766 Mm3 in EFRscenario. The annual average unmet waterdemand increased from 30.1 Mm3 inreference scenario to 33.2 Mm3 in EFRscenario (Figure 10). The results indicatedthat water shortage (annual mean) willincrease by 10.3%, which will impactnegatively on other water users in the sub-catchment and can lead to low economicdevelopment.

(d) Irrigation expansion and EFRimplementation scenarios

The mean annual water demand for thecombined irrigation expansion and EFRimplementation scenario increases from760 Mm3 in reference to 771.5 Mm3

(Figure 11). The annual average shortageswill increase to 227.6 Mm3 compared to179.5 Mm3 for reference, 193.5 Mm3 forirrigation expansion and 33.2 Mm3 forEFR implementation. The increase is 27%



above reference scenario. This indicatedthat when combined irrigation expansionand EFR implementation are considered,more water shortage will occur and thus

increased impact on other water users inthe sub-catchment leading to more loweconomic development.

Figure 9: Annually and seasonally unmet water demand for reference and IE scenarios

Figure 10: Annually and seasonally unmet water demand for EFR scenario



Figure 11: Annual average unmet water demand for all the scenarios

Ndembera catchment

For Ndembera River catchment, results ofthe three scenarios are shown below(Figures 1, 3 and 12):

(a) Irrigation expansion scenario

Irrigation expansion scenario was done todetermine the impacts of increasing theirrigable land from 5,004 to 8,758 ha onwater demand. Annual average waterdemand for this scenario increased from374.1 Mm3 in reference scenario to 380.1

Mm3 (1.6% increase) in irrigationexpansion scenario. The annual averageunmet water demand (Figure 12) increasedfrom 238.3 Mm3 in reference scenario to243.9 Mm3 in irrigation expansionscenario (2.4% increases). Monthlyaverage unmet water demand increasedfrom 15 to 16.1 Mm3 (Figure 13). Theseincreased water shortages of 2.4%annually and 7.3% monthly will impactnegatively on the other water users in thesub-catchment which can lead to loweconomic development in Ndembera.



Figure 12: Location of Ndembera and Kisigo catchments in GRR sub-basin

Figure 13: Annually and seasonally unmet water demand for reference and IE scenarios

(b) EFR implementation scenario

The annual average water demand for thisscenario increased from 374.1 Mm3 inreference scenario to 376.9 Mm3 in EFRscenario. The annual average unmet water

demand will increase from 77.4 Mm3 inreference scenario to 79.9 Mm3 in EFRscenario (Figure 14). The 3.2% increaseannually could impact the downstreamuses and demands with low priority.



(c) Irrigation expansion and EFRimplementation scenarios

The mean annual water demand for thecombined irrigation expansion and EFRimplementation scenario increases from374.1 Mm3 in reference to 380.1 Mm3.

The annual average shortages will increaseto 322.2 Mm3 compared to 316.6 Mm3 forreference, 243.9 Mm3 for irrigationexpansion and 79.9 Mm3 for EFRimplementation. This presents an increaseof 1.77% above reference scenario.

Figure 14: Annually and seasonally unmet water demand for reference and EFRscenario for Ndembera catchment

Kisigo catchment

The below are the results for Kisigo Rivercatchment (Figure 1, 3 and 12):

(a) EFR implementation scenarioKisigo sub-catchment is largely coveredwith the Ruaha National Park andcurrently there is no agricultural activitiestaking place. Kisigo River is seasonal butit has very high runoff during the wetseason, which contributes significantly tothe flow of the Great Ruaha River. The

annual average water demand for thisscenario increased from 64.42 Mm3 inreference scenario to 66.01 Mm3 in EFRscenario. The annual average unmet waterdemand increased from 42.79 Mm3 inreference scenario to 44.12 Mm3 in EFRscenario (Figure 15). This is an increase of3.1% of unmet demand annually. Theresults show that the unmet water demandwill decrease or increase in some yearsfrom the reference to EFR implementationscenario.



Figure 15: Annually and seasonally unmet water demand for EFR implementationscenario for Kisigo catchment

DISCUSSION

Comparing the above three catchments, inrelation to annual irrigation waterdemands, the Little Ruaha catchment isexpected to have the highest increase(10.3%) in unmet demand unlikeNdembera with 2.4% increase by 2025.The large increase for Little Ruahacatchment is tied to its large catchmentarea of about 6,210 km2 with the currentand potential total irrigable area of about25,300 ha despite the likely more riverflows compared to Ndembera catchmentwith only about 3,190 km2 catchment areaand 12,627 ha of current and futureirrigation area. The entire Usangu plainhas the largest potential irrigationagricultural area of about 208,000 ha in theGRR basin. The same applies to EFRwhere there is increase of 10.3% against3.2% and 3.1% for Little Ruaha,Ndembera and Kisigo catchments,respectively. Despite its catchment sizeand hence more water supply for the LittleRuaha River, irrigation needs more waterthus leading to unmet water demand. It isworth to note that the surface flowcontributions of the Great Ruaha, Little

Ruaha and Kisigo rivers to the total inflowto the Mtera reservoir are estimated as56%, 18% and 26%, respectively(Kashaigili et al., 2005). Also, Mbunguand Kashaigili (2017) revealed reductionand changes of streamflows with time andtherefore more water shortages areexpected on the planned waterdevelopment and related ecosystems in theLittle Ruaha catchment. Moreover,Munkyala (2017) indicated that irrigationis the big water user in the Ndemberacatchment, which accounts for about 77%of the all water demands in the catchment.Several scenarios were proposed andanalyzed using the Nile Basin DecisionSupport System to address the issue ofwater shortages in the catchment. Thisimplies that although there are plans fordeveloping irrigation in the GRR basin,this has to be done cautiously in order tobenefit from it and not bringing morechallenges or loss of investment. Detailedprior analysis of the crops and theirirrigation water requirement, and inconsideration of other demands in thebasin and available water sources need tobe done. In this case, the WEAP model canfurther be used in consideration of the



varied irrigation demands and watersupplies including groundwater.

CONCLUSIONS

The current surface water demands andsupply were linked using WEAP model todetermine the future water allocation basedon the set scenarios. In this study waterwas allocated based on priorities set in theWater Resources Management Act, 2009with priorities of 1, 2, 3, 4 and 5 indomestic (urban and rural), environment,agriculture and livestock respectively. TheWEAP modelling of the GRR along withthe input data has several limitations and anumber of assumptions were made. TheGRR sub-basin was sub-divided into threecatchments and the hydrological and waterdemand data were lumped accordingly.The individual water demands belongingto the same sector were combined, giventhe same water allocation priority and thesurface water generated in the catchmentand upstream was supplied for them.Groundwater was assumed to have noeffect on water demand and streamflow.All these assumptions and limitations mustbe taken into consideration wheninterpreting the results. The current totaldemands before implementation of thescenarios are 623.6, 388.17 and 110.23Mm3 for Little Ruaha, Ndembera andKisigo of which irrigation, environment,domestic, non-domestic and livestockconstitutes 49%, 27%, 15%, 5% and 3%,respectively.

The future water demands increased for allthe simulated scenarios producing greatershortages. With comparison to referencescenario, the planned irrigated agricultureexpansion will increase shortages in theLittle Ruaha and Ndembera catchments by149% and 126 % respectively. There is nocurrent and future plan for agriculture inKisigo as the largest part of it falls in theRuaha National Park. The EFRimplementation scenario will also increaseshortages in the Little Ruaha, Ndembera

and Kisigo catchments by 147%, 121%and 13%, respectively. This implies that inview of the EFR considerations, which arenow a pre-requisite all over the world,practicing these regulations implies thatthere would be water shortages in thedemand sectors compared to the currentpractice which does not put EFR underconsiderations.

RECOMMENDATIONS

The study recommends improving thesurface water allocation results in GRRsub-basin and indeed taking advantage ofother powerful capabilities of WEAPmodel as follows:

Construction of reservoirs to store andrelease water, supply managementtechniques for instance, consideringother sources of water (especially inagriculture) and water demandmanagement measures (e.g. use ofirrigation efficiency methods such asdrip or sprinkler) can eliminate theobtained water shortages in thecatchments.

Incorporation of other features andsupply sources like groundwater,water quality and cost benefit analysismodules in order to get a completepicture of the integrated waterresources management of the GRRsub-basin

Impact of climate change on futurestreamflow may also be used as one ofthe scenarios in assessing future waterallocation

Acknowledgement

The authors acknowledge financial supportextended to the first author by iWASHGlows Tanzania for the study scholarshipfor the Master of Integrated WaterResources Management at the Universityof Dar es Salaam, upon which this paper isbased.



REFERENCES

Arranz R. and McCartney M. (2007).Application of the Water Evaluationand Planning (WEAP) model to assessfuture water demands and resources inthe Olifants catchment, South Africa,Colombo, Sri Lanka. IWMI workingpaper 116: International WaterManagement Institute.

Chanzi G. (2017). Assessment of WaterAvailability and Uses at KilomberoBasin Using SWAT and WEAPModels. MSc Dissertation in WaterResources Engineering, Department ofWater Resources Engineering,University of Dar es Salaam, Tanzania.

DHI Water and Environment (2006).MIKE Basin Simulation Model: AVersatile Decision Support Tool forIntegrated Water ResourcesManagement and Planning Horshelm,Denmark.http://www.dhisoftware.com/mikebasin/index.html. Retrieved on 29th July2013.

Gervas R. (2013). Analysis of waterresources allocation in the Great RuahaRiver sub-basin, Tanzania. Master ofIntegrated Water ResourcesManagement, Department of WaterResources Engineering, University ofDar es Salaam, Tanzania.

Kadigi R.M.J., Kashaigili J.J. and MdoeN.S. (2004). The economics of irrigatedpaddy in Usangu Basin in Tanzania:Water utilization, productivity, incomeand livelihood implications. Physicsand Chemistry of the Earth, 29(15-18):1091-1100.

Kashaigili J.J., Kadigi R.M.J., LankfordB.A., Mahoo H.F. and Mashauri D.A.(2005). Environmental flows allocationin river basins, Exploring allocationchallenges and options in the GRRcatchment in Tanzania, Physics andChemistry of the Earth, Parts A/B/C,30(11-16): 689-697.

Kashaigili J.J., Mc Cartney M.P., MahooH.F, Lankford B.A., Mbilinyi B.P.,

Yawson D.K. and Tumbo S.D. (2006).Use of a hydrological model forenvironmental management of theUsangu Wetlands, Tanzania. ResearchReport 104, Colombo Sri Lanka:International Water ManagementInstitute, 48 p. Doi:http://dx.doi.org/10.3910/2009.104.

Kishiwa P., Nobert J., Kongo V. andNdomba P. (2018). Assessment ofimpacts of climate change on surfacewater availability using coupled SWATand WEAP models: Case of UpperPangani River Basin, Tanzania.Proceedings of the InternationalAssociation of Hydrological Sciences378: 23-27.

Lankford B. (2002). Irrigation, livelihoodsand river basins, LADDER WorkingPaper No. 14. ODG, Norwich, UK.

Lankford B.A., van Koppen B., Franks T.and Mahoo H.F. (2004). Entrenchedviews or insufficient science? Contestedcauses and solutions of waterallocation; insights from the GreatRuaha River Basin, Tanzania.Agricultural Water Management 69(2):135-153.

McCartney M.P., Yawson D.K. andHuber-Lee A. (2005). Simulating waterresources development in the Olifantscatchment, South Africa. IWMI Theme1, Working Paper (Draft). InternationalWater Management Institute (IWMI).

Mbungu W.B. and Kashaigili J.J. (2017).Assessing the hydrology of a data-scarce tropical watershed using the Soiland Water Assessment Tool: Case ofthe Little Ruaha River Watershed inIringa, Tanzania. Open Journal ofModern Hydrology, 7: 65-89.

Mulungu D.M.M. and Taipe C.L.R.(2012). Chapter 2: Water evaluationand planning in the Wami River Basin:Application of the WEAP model inBioenergy and Food Security, the BEFSAnalysis for Tanzania. Edited by E.Beall, FAO 54, ISSN 2226 – 6062.

Munkyala D. (2017). Assessment ofsurface water resources potential and its



allocation in Ndembera catchment,Rufiji basin. Master of Integrated WaterResources Management, Department ofWater Resources Engineering,University of Dar es Salaam, Tanzania.

Mwakalila S. (2005). Water ResourcesManagement Guidelines in RuahaBasin, Tanzania, Conference paperh041160, International WaterManagement Institute, 18p..

Perera B.J.C, James B. and KularathnaM.D.U. (2005). Computer software toolREALM for sustainable waterallocation and management, Journal ofEnvironmental Management, 77: 291-300.

Pyrce R. (2004). Hydrological low flowindices and their uses. WSC Report No.04-2004, Watershed Science Centre

(WSC), Trent University, Peterborough,Canada, 50p.

Sieber J., Huber-Lee A. and Raskin P.(2002). WEAP: Water Evaluation andPlanning System User Guide (forWEAP 21), Stockholm EnvironmentalInstitute-Boston, and Tellus Institute.

Sieber J., Yates D., Huber Lee A. andPurkey D. (2005). WEAP: a demandpriority and preference driven waterplanning model: Part 1, modelcharacteristics. Water International,30(4): 487–500.

SMUWC (2001). Sustainable Managementof the Usangu Wetland and itsCatchment. Supporting Report 12:Groundwater. A report to the Ministryof Water and Livestock.

Documents

Assessment of surface water resources in Great Ruaha River