an indicator based assessment for water resources ... · an indicator based assessment for water...

Water Resour Manage (2010) 24:4359–4379DOI 10.1007/s11269-010-9663-3

An Indicator Based Assessment for Water ResourcesManagement in Gediz River Basin, Turkey

Baris Yilmaz · Nilgun B. Harmancioglu

Received: 12 November 2009 / Accepted: 27 April 2010 /Published online: 13 May 2010© Springer Science+Business Media B.V. 2010

Abstract In this study, a water resources management model that facilitatesindicator-based decisions with respect to environmental, social and economic dimen-sions is developed for the Gediz River Basin in Turkey. The basic input of theproposed model is the quantity of surface water that is greatly allocated to irrigationpurposes; therefore, supply and demand interrelations in agricultural water useconstitute the main focus of the study. The model has been applied under threedifferent hydro-meteorological scenarios that reflect baseline as well as better andworse conditions of water supply and demand, not only to reach an assessment ofwater budget, but also to evaluate the impacts of proposed management alternativesunder different conditions. The Water Evaluation and Planning (WEAP) softwareis used as a simulation and evaluation tool to assess the performance of possiblemanagement alternatives, which is measured by nine proposed indicators. The resultsof the study have indicated that the Gediz River Basin is quite sensitive to droughtconditions, and the agricultural sector is significantly affected by irrigation deficitsthat increase sharply in drought periods. Even if the optimistic scenario is assumedto occur, it is not possible to observe a significant improvement in the water budget;however, the negative impacts of climate change can possibly exacerbate the watercrisis. The indicators also verified that, efficient water management is crucial toensure the sustainable use of water resources with respect to environmental, socialand economic dimensions.

B. Yilmaz (B)Dept. of Construction Technology, Golmarmara Vocational High School,Celal Bayar University, 45410, Turgutlu, Manisa, Turkeye-mail: [email protected]

N. B. HarmanciogluWater Resources Management Research and Application Center (SUMER),Dokuz Eylul University, 35160, Buca, Izmir, Turkeye-mail: [email protected]

4360 B. Yilmaz, N.B. Harmancioglu

Keywords Water resources management · Hydro-meteorological scenario ·Indicator · WEAP · Gediz River Basin

1 Introduction

At the most basic level, two related global trends greatly exacerbate the watercrisis. These trends relate to the rapid increases in population growth and economicdevelopment, both of which strongly increase water demand as well as pollution.On the contrary, the quantities of water that any country can economically develop,unfortunately, continue to decrease or remain limited. For the above and a varietyof other reasons like global warming, improved living standards, urbanization, andindustrialization, water managers have been faced with more complex and difficultproblems in the early 21st century, and it is expected that coping with water prob-lems will be harder in the future. Therefore, the traditional water management thatconsiders water supply enhancement investments in regional dimensions via eco-nomic analyses has been replaced by sustainable water management which is a partof sustainable development defined by World Commission on Environment andDevelopment as “meeting the needs of the present without compromising the abilityof future generations to meet their own needs”. Water resources management playsa crucial role in this area since water is essential for socio-economic developmentespecially in arid and semi arid regions where agriculture is the major industry. Inthe light of this philosophy, many researchers have expressed that sustainabilityis to be measured by indicators that serve not only to understand the status andongoing trends in a river basin, but also to assess the results of particular manage-ment approaches and practices (Harmancioglu 2004; Guimarães and Magrini 2008;Walmsley 2002).

On the other hand, water resources problems are typically characterized by uncer-tainties with respect to hydrologic and meteorological events as well as to demandpatterns. Assigning inaccurate values to these events and patterns can invalidate theresults of the study and the accruing decisions. A scenario analysis that can modelmany water problems is an alternative approach in water resources management bydescribing a set of possible future outcomes. Through the development of scenariosand alternative strategies, future states of water supply and water demand canbe assessed under varying climatic and hydrological conditions (Varis et al. 2004;Jeong et al. 2005; Pallotino et al. 2005), water resources management plans (Kochand Grünewald 2009; Loukas et al. 2007; Al-Omari et al. 2009) and water demandmanagement practices (Chen et al. 2005; Lévite et al. 2003).

River basin simulation models based on a node-link network representation toderive basin water balance are the essential tools to run scenarios. They enable theuser to evaluate a variety of measures related to water resources and demandmanagement under various hydrological conditions. The indicators obtained directlyby models and/or from a post-process of model results serve to test the performanceof alternative management policies. Since there has been a pronounced need forcoping water crisis in recent years, these quantitative evaluations provide valuableinformation to water managers.

The Gediz River Basin like many river basins in Eastern and SouthernMediterranean countries also suffers from water scarcity due to rapid demographicand economic development, urbanization, industrialization and inefficient irrigation

Assessment for Water Resources Management in Gediz River Basin 4361

activities. The basin approaches a total population of 2.5 million. The basin demon-strates a broad range of water management problems, two major ones being wa-ter scarcity and pollution. Although the basin experiences recurrent droughts, waterscarcity is also explained by the competition for water among various uses (waterallocation problems), mainly irrigation versus the domestic and fast growing indus-trial demand. Current analyses on hydrological budget of the basin indicate that theoverall supply of water for various uses is approximately equal to the overall demand(Harmancioglu et al. 2005). This means that there is no reserve left for further waterallocation, thus, to maintain the sustainable development of the region, to providesufficient quantity and quality of water for all sectors and to assess the long-termimpacts of water policies, water demands and supply availability should be evaluatednot only in terms of existing trends and possible future tendencies in water use, butalso by possible hydro-meteorological and climate change scenarios downscaled tothe basin. Hence, a comprehensive assessment of water budget in the Gediz RiverBasin and an evaluation of management plans in terms of economic, social andenvironmental indicators have become imperative.

The OPTIMA (Optimisation for Sustainable Water Resources Management,co-funded by the European Commission FP6 Programme) project where the overallaim was to develop, implement, test, critically evaluate, and exploit an innovativeapproach to water resources management intended to increase efficiencies and toreconcile conflicting demands, can be cited as a detailed study that focuses on watermanagement in the Gediz River Basin. In that case, a simulation based water re-sources planning and optimization system was developed and applied to addresswater issues (Harmancioglu et al. 2008; Cetinkaya et al. 2008). However, the evalua-tions are based on 1-year-simulation of predefined years demonstrating wet, nor-mal and dry conditions. Neglecting long term effects of management alternatives isanother weakness.

Based on the above considerations, this study aims to develop an indicator-basedassessment for water resources management in the Gediz River Basin. Within thisframework, the objectives of the study are to: (1) state the prevailing water problemsin terms of water supply/use patterns and water management practices; (2) developlong term hydro-meteorological scenarios that reflect baseline as well as better andworse conditions of water supply and demand; (3) describe the proposed manage-ment alternatives and sustainability indicators; (4) reach a comprehensive assessmentof water resources management alternatives in terms of developed indicators. Thewater availability and demand are simulated under three different climate scenariosto reflect baseline as well as better and worse conditions. The simulations are estab-lished for the period between 2003 and 2030, 2003 being the last of published hydro-meteorological data. In the simulation process, assumptions are made, based on longterm hydrologic and operational data as well as on relevant project reports. TheWEAP (Water Evaluation and Planning System) model is used for the simulations.

2 Gediz River Basin

The Gediz River, with a length of 275 km, drains an area of some 18,000 km2

and flows from east to west into the Aegean Sea just north of Izmir, Turkey.The Gediz River Basin (GRB) is located geographically at the interval of 38◦01′–39◦ 13′ northern latitude and 26◦ 42′–29◦ 45′ eastern longitude. It has a typical


Mediterranean climate with hot, dry summers and cool winters. In the basin, meanannual temperature and mean annual precipitation are 15.6◦C and 635 mm, respec-tively. January and February are the wet, and July and August are the driest months.75% of the total annual precipitation is observed between November and March.

The basin covers an area of about 110,000 ha which are subject to extensiveagricultural practices with large irrigation schemes. The main crops cultivated arecotton, maize, grape, vegetables and cereals. Due to climatic conditions, irrigationis most important requirement of agriculture which is the main economic activity inthe basin. As in many other ‘agriculture dominant’ basins, a great portion of sur-face water resources, i.e., 75%, is allocated to irrigation. Harmancioglu et al. (2005)mention, irrigation uses a large share of the surface water resources in the GRB witha total about 660 106 m3.

The population of the GRB was about 1.7 million in 2000, with an annualgrowth rate of 1.5%. However, the internal migration from rural to urban areas(especially to Izmir) and the rapid urbanization in the major cities exert pressureon domestic water demand which increases at an annual rate of 2%. The growth ofthe industrial demand is even more drastic with an annual rate of 10%, due to therapid industrialization (SMART 2005).

The seaward fringe of the Gediz delta is an important nature reserve and hasbeen designated as a Ramsar site to protect rare bird spices. The area receives excesswater from the Gediz River for much of the year, but since 1990, with restrictionson irrigation releases, it suffers from water shortages. The summer months are thecritical times for providing water to the nature reserve due to the intensive irrigationpractices. The water demand of Birds Paradise is 14.2 106 m3 and 7.9 106 m3 in a dryand a wet year, respectively (De Voogt et al. 2000).

Due to the antiquity of water conveyance systems (open channel) which lead tohigh water losses, lack of maintenance of irrigation systems and farmers’ lack ofknowledge about appropriate irrigation practices, it is certain that the current useof water for irrigation purposes is inefficient. The canal loses are approximately32%, and the irrigation efficiency is 60% (Silay and Gunduz 2007). Therefore, themodernization of irrigation techniques should be encouraged, and more productiveuse of water should be a fundamental objective, not only for agriculture but also forother water demanding sectors.

In this study, although a water resources management model is intended for allGediz Basin, six large scale irrigation districts and the Birds Paradise are consideredas water demand sites due to lack of adequate and reliable data on domestic andindustrial water uses that greatly consume groundwater resources. In addition,Alasehir catchment is excluded from the analyses due to the same restrictions.

3 Modeling Gediz River Basin

3.1 Water Evaluation and Planning System (WEAP)

WEAP, developed by the Stockholm Environment Institute, is a practical tool forwater resources planning, which incorporates both the water supply and the waterdemand issues in addition to water quality and ecosystem preservation, as requiredby an integrated approach to basin management. WEAP, which is free for academic


use, is also user friendly, easy-use software, and its applications generally involve thefollowing steps: (1) System definition including time frame, spatial boundary, systemcomponents and configuration (2) Constitution of ‘current accounts’ which providesa snapshot of actual water demand, resources and supplies for the system (3) Buildingscenarios based on future trends on hydrology, management strategies, technologi-cal developments and/or other factors that affect demand, supply and hydrology(4) Evaluating the scenarios with regard to criteria such as adequacy of waterresources, costs, benefits, and environmental impacts.

WEAP operates on a monthly time step, starting from the first month of thecurrent accounts year and continuing up to the last month of the last scenario year; itcomputes water mass balance for every node and link in the system for the simulationperiod. The detailed features of WEAP can be found in Yates et al. (2005) and userguide of the model (SEI 2007).

3.2 Analysis Setup

The Gediz River network with primary tributaries, meteorological stations, streamgauging stations (SGS) and reservoirs can be seen in Fig. 1. The monthly runoff dataof the stations obtained from the streamflow discharge annals (seven under ElectricalWorks Authority’s (EIE) operation and five under State Hydraulic Works’ (DSI)operation) are used to represent the river headflows in the analysis.

Demirkopru and Gol Marmara are the reservoirs that supply water for down-stream irrigation demands. Demirkopru Dam supply water for all the six irrigationdistricts while Gol Marmara is operated to fulfill the water deficit in summer season.Since there are no sufficient and reliable long term streamflow data for the rivers thatfeed Afsar and Buldan dams, these dams are not taken into account in this study. InWEAP, the monthly operation rules of the reservoirs are introduced with the initialstorage and the buffer coefficient which is the fraction of water in the reservoir avail-

Fig. 1 The Gediz River Basin


able each month for release. In accordance with the monthly operation reports ofthe reservoirs which are taken from the annual operation reports of DSI II.RegionalDirectorate, the buffer coefficients are determined through the calibration process.The leakage losses are assumed equal to zero in the computations.

In the analyses, the Adala, Ahmetli and Menemen irrigation districts (IDs) aretaken into account as demand sites. Each of these districts is disaggregated into twosub-irrigation districts, since the namesake regulators divert water to right bank (RB)and left bank (LB) irrigation schemes. This segregation is a must to compute thewater demand reasonably, due to the different crop patterns as well as the variationin growing seasons of crops in the two schemes. The relevant data are obtained fromthe DSI II.Regional Directorate in Izmir. The priority of each demand site is equallyset to 1 to reflect the highest priority. The WEAP schematic where the Gediz RiverBasin is demonstrated within a node-link network is given in Fig. 2.

Physical and contractual constraints of regulators and canals are also incorporatedto analyses. Since the water distribution scheme in tertiary canals is out of the scopeof the study, only the conveyance losses (including the evaporation and leakagelosses) in the main and secondary canals are accounted with a general loss rate.

The crop pattern is also an important factor to determine irrigation water demand.Although the main crops in the GRB are cotton, maize, grape and vegetables, theircultivation areas differ within the IDs. In addition, the increasing cultivation of maizeand the decreasing cultivation of cotton is a noticeable response against water scar-city after recurrent droughts. Therefore, the crop patterns of each ID in recent yearsare obtained from DSI II.Regional Directorate, not only to compute the waterdemand, but also to build the demand management alternative. It should also benoted that, although a large number of crop types is cultivated in GRB, the presentstudy focuses on four main crops while other crops are aggregated into in accordancewith the cultivation percentages of main crops. The crop coefficient (Kc) is another

Fig. 2 The schematic used in WEAP


data requirement for the analyses. Kc is crop specific and varies by the length of thecrop development stages as well as climate conditions. The CROPWAT softwaredeveloped by FAO is an essential tool to develop Kc values not only with its pre-defined crop types, but also with included meteorologic data base. Although the Kc

values are determined within this program, the Kc values determined by DSI areused in the analyses. This is considered as a reasonable way to simulate the realwater allocation paradigm and also to calibrate the model since the water allocationis managed by DSI.

The climate conditions are demonstrated by the long term monthly average tem-perature and precipitation data of the Menemen, Manisa and Salihli meteorologicalstations, respectively. To identify irrigation water demand, potential evapotranspi-ration (ETo) in each hydrological unit is computed by CROPWAT which alsoenables the computation of effective precipitation (Pe). The Birds Paradise is anotherdemand site where water requirement is calculated by using evapotranspiration(ETo) and precipitation (P) data of Menemen meteorological station. As mentionedby De Voogt et al. (2000), the additional water requirements occur from March toOctober.

The ‘current accounts’ in WEAP terminology, represents the basic definition ofthe water system in its present state, and is assumed to be the starting year for allscenarios. It includes the specifications of supply and demand data for the first yearof the study on a monthly basis. Since the current accounts is inferred to as ‘the bestavailable estimate of the system’, the long term monthly averages of runoff as well asthe monthly averages of temperature, precipitation and evapotranspiration are usedto develop the current accounts, and set to the year 2003. The crop pattern in currentaccounts year is assumed as the descriptive pattern of the demand sites in 2003.With respect to basic economic analysis, the cost, price and maximum yield data aretaken from Turkish Chamber of Agricultural Engineers. It should be noted that, theprice of water is incorporated within the cultivation cost. The yield response factor(ky) which refers to the relationship between relative yield decrease and relativeevapotranspiration, are obtained from FAO (1979). The operation and maintenancecosts of supply nodes are obtained from OPTIMA project report (OPTIMA 2006).The long term reservoir storages in January are assumed as the initial reservoirstorages which are 350 106 m3 and 90 106 m3 for Demirkopru Dam and Gol Marmara,respectively.

3.3 Model Calibration

The SGS 518 and the storage volumes in Demirkopru Dam are used to calibrate themodel. Since the operation rules of the dams are irregular and are arranged accordingto the yearly water demands, the calibration is executed individually with the relevantdata for the years from 1995 to 2003. The calibration graphs those refer to 2001(dry year), 1996 (normal year) and 1999 (wet year) are depicted in Fig. 3. The Nash–Sutcliffe Efficiency (NSE) and Pearson’s correlation coefficient (r) are representedthe model performance as ‘very good’ (Moriasi et al. 2007).

Through the model calibration, transmission link loss rate, irrigation efficiencyand the irrigation return flow rate are determined as 32%, 60% and 16%, respec-tively. Since the irrigators prefer to fulfill the irrigation demands in July and August,the buffer coefficients are set to 1 (no restriction) for these months. However, if the


0

100

200

1 2 3 4 5 6 7 8 9 10 11 12

106m

3

Months

Simulated

Observed

0

200

400

600

1 2 3 4 5 6 7 8 9 10 11 12

106m

3

Months

Simulated

Observed

(a) Runoff volume at SGS 518 for 2001 (b) Storage volume in Demirkopru for 2001

0

60

120

1 2 3 4 5 6 7 8 9 10 11 12

106m

3

Months

Simulated

Observed

0

200

400

600

800

1 2 3 4 5 6 7 8 9 10 11 12

106m

3

Months

Simulated

Observed

(c) Runoff volume at SGS 518 for 1996 (d) Storage volume in Demirkopru for 1996

0

150

300

450

1 2 3 4 5 6 7 8 9 10 11 12

106m

3

Months

Simulated

Observed

500

800

1100

1 2 3 4 5 6 7 8 9 10 11 12

106m

3

Months

Simulated

Observed

(e) Runoff volume at SGS 518 for 1999 (f) Storage volume in Demirkopru for 1999

NSE = 0.959 r = 0.998

NSE = 0.979 r = 0.995

NSE = 0.890 r = 0.955

NSE = 0.915 r = 0.990

NSE = 0.923 r = 0.994

NSE = 0.956 r = 0.987

Fig. 3 The calibration graphs

storage volume of the Demirkopru Dam is available (e.g. higher than 650 106 m3

in June and higher than 300 106 m3 in September), water releases are allowed inearly and late irrigation season with the buffer coefficients 0.1 and 1.0, respectively.These results are in accordance with the rates mentioned by DSI engineers (Silayand Gunduz 2007) as well as the buffer coefficients which are similar with waterallocation principles of DSI.

4 Management Alternatives and Scenario Analysis

4.1 Alternatives Evaluated

The management alternatives are explained below along with their main assump-tions, schedules and their relevant areas of application (Table 1). The proposedalternatives are compatible with basin features, and the main objective is to increasethe water supply as well as irrigation efficiency or to decrease the water demand.

The canal maintenance alternative (A1) addresses the gradual reduction of lossesfrom 32% to 15% in 6 years with an even cost distribution throughout this period.This alternative assumes that the investments will be implemented evenly in alltransmission links. This is a reasonable way to ensure the same investment priority


Table 1 Managementalternatives evaluated

Alternatives Description

Do nothing (A0) No additional measures to thecurrent system

Canal maintenance (A1) Maintenance of irrigation networksin order to reduce the water losses

Crop pattern change (A2) Substitution of existing crops byother crops that have less waterdemand

Drip irrigation (A3) Changes in irrigation system(in favor of drip irrigation) toincrease efficiency

Pressured systems (A4) Substitution of existing waterdistribution system by apressured system

A2 + A1 = (A5) Demonstrates the strategyconstituted by crop changes& canal maintenance

A2 + A3 = (A6) Demonstrates the strategyconstituted by crop changes& drip irrigation systems

A2 + A4 = (A7) Demonstrates the strategyconstituted by crop changes& pressured systems

between irrigation districts. The recent bids for similar investments as well as thepress releases of irrigation association managers are used to estimate the cost whichis about 400 e/ha, where the operation and maintenance (O&M) costs are fixed.

After the severe droughts, cultivation of maize instead of cotton had been the onlyresponse of farmers to the water scarcity problem. On the basis of variations withinthe recent years, this response is assumed to continue as such. The A2 alternative isdesigned to follow this trend, and is applied to the irrigation districts according totheir own trends. The main focus of this alternative is increasing the cultivation areaof maize while decreasing the cotton cultivation. In addition, the slow increase ofgrape cultivation is also added to the analyses.

Since agriculture is the major economic activity in GRB, it is vital for the localeconomy. However, all crops need to be irrigated in summer season due to theprevailing climate conditions. Although the farmers have been informed on howto improve irrigation efficiency, e.g., via drip irrigation systems, the existing waterdistribution system of open canals and the economic capacity of the farmers are thelimitations in this regard. In recent years, the farmers are offered some significantsubsidies to construct water saver irrigation technologies. Accordingly, the alterna-tive A3 is developed to promote the irrigation efficiency to 90% from 60%. Thisalternative includes two parts. The first part is the replacement of the current waterdistribution network by a pressured (piped) system which is scheduled into 6 years(2004–2010). The cost is incorporated into the analysis as 2,500e/ha, where the O&Mcosts are fixed. The water loss in the piped system is assumed to be 2% of waterpassing through the link. The second part refers to a transition from furrow irrigationmethods to drip irrigation. The alternative A3 is introduced to the model by assumingthat the share of drip irrigation will be in order of 80% of irrigated area in 2030. Theinitial implementation will begin after 2010 (starting year of the pressured system).


The cost of the drip irrigation method is estimated to be 5,000 e/ha, and the annualO&M cost is evaluated as 10% of the capital cost.

The alternative A4 evaluates only the first part of alternative A3 which considersthe use of drip irrigation. That is, it focuses only on the replacement of the currentwater distribution network by a piped one. With so doing, it is possible to evaluatehow the performance indicators change if high conveyance losses decrease to anegligible size. It is also a reasonable way to evaluate an alternative that costs lessrelative to A3.

The other alternatives (A5, A6 and A7) are also developed to evaluate thecumulative effects of previous alternatives. Here, the crop pattern change alternative(A2) is considered together with A1, A3 and A4 to develop A5, A6 and A7,respectively.

4.2 Reference Scenarios

The reference scenarios are based on hydro-meteorological changes, and simulationsrun to identify the possible impacts on water budget in terms of water supply anddemand. Accordingly, three main reference scenarios are developed with combina-tions of water availability and demand scenarios. Here, it should be noted that thedemand scenarios are based only on temperature increases that leads to incrementsin crop irrigation water requirement due to higher potential evapotranspiration.Since the Gordes Dam will be in operation within the simulation period (2003–2030),the planned water withdrawal to Izmir (8.5 106 m3/month) is also incorporated intothe reference scenarios.

The Business as Usual (BAU) scenario foresees the preservation of long termaverages with respect to water availability and water demand. In order to formulatethe BAU scenario, the monthly stream flow data which were monitored between1977 and 2003 are repeated for the simulation period (2003–2030). The water demandcomputations which use monthly averages of temperature and precipitation arecarried out by considering constant irrigation areas as well as the same crop patternsfor all irrigation districts.

The Pessimistic scenario (PES) focuses on low water availability and increasingdemand. The study dealing with the climate change effects in Gediz River Basin(Ozkul 2009) estimates the decreases in stream flows and precipitation as well asthe increases in the average monthly temperatures. The mentioned study determinesthe expected future variations in those hydro-meteorological parameters by usingdifferent climate change scenarios for the years 2030, 2050 and 2100. The results for2030 in B2-MES scenario which emphasize a world of moderate population growthand intermediate level of economic development and technological change are usedto formulate the pessimistic scenario. Since the decrement in runoff is estimated to beabout 23%, the monthly runoff time series used in the BAU scenario are decreasedwith this ratio to obtain pessimistic water availability conditions. The changes inprecipitation and temperature with respect to the B2-MES scenario are used to setup the demand side of the water system (Table 2). Since the estimations are given for2030, a value in any given month within the simulation period is computed by linearinterpolation.

The Optimistic scenario (OPT) foresees high water availability and stable waterdemand. In this scenario, the stream flow data used in BAU are increased by 23%.


Table 2 Rainfall and temperature changes used in PES

Month Jan Feb Mar Apr May Jun

Rainfall (%) −3.3 −0.7 −0.2 −5.9 −12.4 −24.9Temperature (◦C) 0.9 0.9 0.8 1.1 1.4 1.6

Month Jul Aug Sep Oct Nov Dec

Rainfall (%) −35.2 −13.5 −9.9 −17.1 −6.2 −4.4Temperature (◦C) 1.6 1.7 1.5 1.4 1.1 1.0

With so doing, the runoff series are considered not only as wet year averages, but alsoas the reversed conditions of the pessimistic scenario. Temperature, precipitation andirrigation area are assumed constant in defining the stable water demand.

5 Evaluation Indicators

In order to evaluate the achievements of management alternatives, nine indicatorsrelevant with environmental, social and economic sustainability are developed(Table 3). It should be noted that since a simulation model is used, the focus is givento the numeric outputs of the model; in other words, the developed indicators arebased on the quantitative assessments of alternatives.

Table 3 The evaluation indicators

Indicator Description

Agricultural Sustainability Index (ASI) The temporal aggregation of supply/demand ratiotime series (only for irrigation) according to theperformance measures where the satisfactoryrange is considered between 0.8 and 1.0

Environmental Sustainability Index (ESI) The temporal aggregation of supply/demand ratiotime series (only for environmental needs)according to the performance measures wherethe satisfaction value is 1 (full coverage)

Water Exploitation Rate (WER) The percentage of surface water potential that isallocated for irrigation; (annual average is usedin the evaluations)

Yield Reliability (YR) Average yield reliability of main cultivated crops(the satisfactory range is considered between 0.75and 1.00 for all crops)

Irrigation Water Deficit (IWD) Represents annual unmet demand for irrigation(annual average is used in the evaluations);106 m3

Domestic Supply Reliability (DSR) The supply reliability of transmission link to Izmirfrom Gordes dam

Benefit/Cost Ratio (B/C)∑

Benefits/∑

Costs of management alternatives forthe simulation period

Irrigation Water Use Efficiency (IWUE) Production value (monetary) of agricultural practicesper allocated water for irrigation (annual average isused in the evaluations); e/m3

Total Production Value (TPV) Annual total production value of agricultural practices(annual average is used in the evaluations);million e


The methodology for developing indicators is based on two approaches. The firstone is the use of average values for indicator time series that is obtained annuallyduring the simulation period. The WER, IWD, IWUE and TPV indicators are com-puted using this approach. Here, a differentiation is made for the B/C which isobtained by dividing the total benefits (overall cultivation revenue) to the totalcosts (the sum of capital and operational costs). The second approach that hasbeen recommended by the American Society of Civil Engineers and InternationalHydrological Programme (IHP-working group of UNESCO) is the temporal aggre-gation of indicator time series using performance measures of reliability, resilienceand vulnerability (ASCE 1998). This procedure can be illustrated by considering anyselected indicator C, whose time series of values is denoted as Ct, where the simulatedtime period, t, extends to some future time, T (Fig. 4). To define performancemeasures, the lower limit (LL) and the upper limit (UL) of satisfactory range shouldbe identified. These limits may change within a year and over multiple years, and arebased on the decision makers’ judgements.

Reliability (RE) is defined as the probability that any particular Ct value will bewithin the range of values considered satisfactory (Eq. 1).

RE of (C) = Number of satisfactory Ct valuesTotal number of simulated periods

(1)

Resilience (RS) is an indicator describing the speed of recovery from an unsatis-factory condition. It is the probability that a satisfactory value Ct+1 will follow anunsatisfactory Ct value (Eq. 2).

RS of (C)= Number of times a satisfactory Ct+1 value follows an unsatisfactory Ct valueTotal number of unsatisfactory values

(2)

Vulnerability (VU) is a statistical measure of the extent (magnitude) or the dura-tion of failures in a time series. The extent (magnitude) of a failure is the amount thata value Ct exceeds the upper limit, UL(Ct), of the satisfactory values or the amountthat value falls below the lower limit, LL(Ct), of the satisfactory values. In this study,vulnerability is defined as expected extent-vulnerability (Eq. 3), and the durations offailures are excluded.

VU of (C) =∑

individual extents of Ct failuresTotal number of individual extents of Ct failures

(3)

Fig. 4 Indicator time seriesand range of satisfactoryvalues

Time

Indi

cato

r (C

)

Satisfactory range

Upper limit

Lower limit


Sustainability Index that ranges from 0, for its lowest and the worst possible value,to 1, as its highest and the best possible value, is computed by multiplying thereliability, resilience and (1-vulnerability) values since reliability and resilience arethe maximizing while vulnerability is the minimizing criteria for sustainability. Thus,the agricultural sustainability index (ASI) and environmental sustainability index(ESI) are calculated with Eqs. 4 and 5, respectively. The indicator used for ASI isthe supply/demand ratio (S/D) of irrigation districts as well as of the environmentalneeds for ESI.

ASI = RE(Si/Di) ∗ RS(Si/Di) ∗ (1 − VU(Si/Di)) (4)

ESI = RE(Se/De) ∗ RS(Se/De) ∗ (1 − VU(Se/De)) (5)

In this study, the satisfactory range of Fig. 4 is selected between 0.80 and 1.00 forirrigation purposes, which are the LL(Si/Di) and UL(Si/Di), respectively. However,for environmental needs, it is desired to meet all the water demand; thus, theLL(Se/De) and UL(Se/De) are both fixed to 1.

The reliability (Eq. 1) is solely used to determine the domestic supply reliability(DSR), which indicates the percent of time the requirement (the expected amount ofwater from a transmission link to domestic uses, Dt) is fully met with water allocatedto the transmission link, St, (Eq. 6). Here again, the LL(St/Dt) and UL(St/Dt) areboth fixed to 1 while developing the indicator for this study.

DSR = Number of periods the requirements fully met (St/Dt = 1)Total number of periods the transmission link operated

(6)

The yield reliability (YR) is another core social indicator that represents theprobability of achieving at least α % of maximum yield. Here, α is a subjective con-stant and indicates the satisfactory level. The YR is formulated in Eq. 7, where thecrop type and the number of crops are i and n, respectively.

YR = 1

n

n∑

i=1

Number of satisfactory yield values for crop iTotal number of simulated periods for crop i

(7)

Here, it should be noted that, since the crop yield, or more generally agriculturalproductivity, is largely influenced by the irrigation water deficit, the response of yieldto water deficit is quantified through the yield response factor (ky) which relatesrelative yield decrease to relative evapotranspiration deficit (Eq. 8):

1 − Ya

Ym= ky

(

1 − ETa

ETc

)

(8)

where, Ya, Ym, ETa, ETc and ky represent actual yield, maximum yield, actualevapotranspiration, crop (potential) evapotranspiration and yield response factor,respectively. ky values differ according to the crops as well as to the ETa that differsaccording to the irrigation system. Therefore, the YR indicator is a valuable indicatorto address the performance of management alternatives with socio-economic aspects,and it is incorporated to the analyses with a reasonable satisfaction level (α = 0.75).


6 Results and Discussion

6.1 Water Budget Evaluation

In Gediz River Basin, the surface water potential is 1,120 106 m3. However, thesurface water that can be stored in the reservoirs (700 106 m3) is quite less than theoverall surface water potential. Since the reservoir storages are the main sources ofirrigation water supply, it is reasonable to consider the stored water as the supplyavailability in assessing water budget. The water supply and water demand simulatedwith the reference scenarios are depicted in Figs. 5 and 6, respectively. When thetwo graphs are compared, it is obvious that the basin will suffer from water shortageespecially in dry periods, in other words irrigation demand coverage is quite sensitiveto droughts. The supply/demand ratios as simulation period averages regarding OPT,BAU and PES scenarios are 0.77, 0.71 and 0.58, respectively. That means the basinis already under water stress, and worsening conditions would accelerate the waterdeficits. When the basin experiences a drought period (for example 5 sequential dryyears), it is possible to encounter water deficits which range from 80 106 m3 to morethan 500 106 m3 (Fig. 7).

6.2 Indicator Based Evaluation

For the Gediz case, the performance matrix (PM) is set up with nine indicators versuseight alternatives (including A0, the do-nothing alternative). Since the entries ofPMs differ with hydro-meteorological conditions, three PMs are obtained for threereference scenarios. The PMs not only permit indicator-based assessments but alsoconstitute a stepping stone for the eventual decision making process.

In Table 4, where the best values are highlighted, performance evaluation underthe BAU scenario is presented. A0 and A2 alternatives are not seen as feasible

0

500

1000

1500

2000

2500

2003 2006 2009 2012 2015 2018 2021 2024 2027 2030

106 m

3

BAU OPT PES

Fig. 5 Total runoff into the reservoirs within the simulation period


500

525

550

575

600

625

2003 2006 2009 2012 2015 2018 2021 2024 2027 2030

106 m

3

BAU & OPT PES

Fig. 6 Total water demands in the reference scenarios

alternatives since they are dominated by the others, excluding A3 and A6 for B/Cindicator. Reasonably, A1, A4 and A7, which focus on the reduction of waterlosses in the conveyance system, as well as A3 and A6, which improve irrigationefficiency, are considered as the alternatives which are worth analyzing in depth.Since increased water availability is foreseen in the optimistic scenario, similar resultswith higher performance values are observed (Table 4). In the pessimistic scenario,the performance indicators are worse than those in BAU and OPT, as expected(Table 4). However, it is useful to evaluate the achievements of the alternatives under

50

200

350

500

2003 2006 2009 2012 2015 2018 2021 2024 2027 2030

106 m

3

BAU OPT PES

Fig. 7 Total unmet water demands in the reference scenarios


Tab

le4

Per

form

ance

eval

uati

onun

der

the

refe

renc

esc

enar

ios

Alt

erna

tive

ASI

ESI

WE

Ra

YR

IWD

aD

SRB

/CIW

UE

TP

V

Bus

ines

s-as

-usu

alA

00.

170.

100.

710.

5512

1.27

0.69

1.29

0.31

147.

39A

10.

220.

200.

660.

6780

.16

0.73

1.38

0.36

159.

98A

20.

170.

110.

720.

5411

5.26

0.70

1.28

0.30

148.

57A

30.

410.

490.

590.

8729

.76

0.75

1.21

0.43

174.

57A

40.

320.

320.

660.

7956

.28

0.75

1.35

0.37

167.

46A

50.

240.

220.

670.

7174

.64

0.73

1.38

0.35

161.

25A

60.

470.

460.

600.

8836

.70

0.75

1.19

0.43

172.

00A

70.

380.

310.

650.

7957

.34

0.75

1.34

0.37

166.

61O

ptim

isti

cA

00.

250.

170.

620.

6497

.03

0.80

1.35

0.30

154.

75A

10.

330.

280.

570.

7857

.65

0.83

1.44

0.35

167.

03A

20.

250.

180.

610.

6297

.44

0.80

1.35

0.30

153.

71A

30.

690.

690.

490.

9417

.94

0.85

1.25

0.44

178.

90A

40.

510.

510.

550.

8840

.32

0.85

1.40

0.37

172.

36A

50.

360.

330.

560.

7857

.38

0.83

1.44

0.35

166.

49A

60.

720.

660.

500.

9425

.08

0.85

1.24

0.43

176.

32A

70.

540.

460.

540.

8841

.93

0.85

1.40

0.37

171.

51P

essi

mis

tic

A0

0.07

0.04

0.78

0.33

180.

720.

531.

130.

3512

9.58

A1

0.13

0.08

0.76

0.50

132.

980.

601.

240.

3714

3.71

A2

0.08

0.04

0.78

0.33

176.

710.

541.

130.

3412

8.95

A3

0.22

0.22

0.71

0.65

74.4

50.

621.

100.

4415

7.67

A4

0.17

0.13

0.76

0.55

107.

540.

621.

230.

3915

1.18

A5

0.13

0.08

0.75

0.50

130.

310.

611.

240.

3714

3.22

A6

0.24

0.23

0.71

0.65

73.9

80.

621.

100.

4415

7.00

A7

0.18

0.14

0.75

0.56

106.

100.

621.

230.

3915

0.50

a Indi

cato

rto

bem

inim

ized

;A

0:do

noth

ing;

A1:

cana

lm

aint

enan

ce;

A2:

crop

chan

ge;

A3:

drip

irri

gati

on;

A4:

pres

sure

dsy

stem

s;A

5:cr

opch

ange

+ca

nal

mai

nten

ance

;A6:

crop

chan

ge+

drip

irri

gati

on;A

7:cr

opch

ange

+pr

essu

red

syst

ems


reference scenarios. In this regard, indicator based assessments are carried out on thebasis of ‘percent improvement relative to the A0 alternative’.

The agricultural sustainability index, (ASI), obtained through the reliability,resilience and vulnerability of irrigation demand coverage (supply/demand ratio),change between 0 and 1. It is not reasonable to strictly characterize the status ofirrigation water budget; however, ASI values close to 1 refer to a ‘good performance’.This means the irrigation water budget is already under stress in the GRB andsignificant recovery in ASI can be obtained by the alternative policies. ReferringTable 4, ASI can be increased by 243% through the cumulative effects of dripirrigation methods and crop pattern change. The other alternatives also enhanceASI by different percentages. It should be noted that, the percent improvementsin ASI under pessimistic conditions are larger than the ones in business-as-usual andoptimistic conditions for all alternatives.

Similar results are also valid for environmental sustainability index (ESI), whereA3 and A6 alternatives show significant improvements. A4 and A7 alternatives alsoincrease ESI. Since the water requirement of the Birds Paradise in the summerseason (12 106 m3) is quite less than the seasonal irrigation demand (550 106 m3), itis possible to fulfill the water demand of the bird sanctuary with water conservativealternatives that serve to increase ESI.

The water exploitation rates (WER) are 0.62, 0.71 and 0.78 in OPT, BAU andPES, respectively. However, when A3 alternative is implemented, they sharplydecrease to 0.49, 0.59 and 0.71 in the respective reference scenarios. Consideringthe average water quantity stored in the reservoirs is almost 700 106 m3, a 10%decrease in WER leads to water savings in the order of 70 106 m3. The alternativesA3 and A6 produce 9% decrease in WER in the PES scenario and better performin BAU and OPT. Noticeably, the water exploitation rate obtained via alternativesA3 and A6 under the PES scenario is same with the one which A0 perform inBAU scenario. Generally, all management alternatives reduce the exploitation rateat different percentages; however, the reduction is more noticeable in BAU and OPTscenarios due to the higher water availability conditions.

The yield reliability (YR) representing probability of achieving at least 75% ofthe maximum yield can be improved by all alternatives, excluding the crop patternchange alternative (A2). This can be explained by the yield response factors (YRF)of crops. The YRF of maize (1.25) is higher than YRF of cotton (0.85). Therefore,the yield decrease in maize is expected to be higher than cotton, if evapotranspirationdeficits occur. In addition, the water releases in June when maize requires more waterthan cotton, are generally restricted to fully meet the irrigation demand in July andAugust. Accordingly, alternative A2 leads to a decrease in YR. On the other hand, anincrease of 97% in YR in PES scenario is achieved through the favorite alternativesA3 and A6. The other alternatives also increase this indicator value at significantrates, and effective responses to yield reliability under pessimistic conditions arenoticeable.

Irrigation water deficit (IWD), or the average amount of water to meet the ETshortfall (ETc − ETactual), is a valuable indicator for the analyses. Although the totalET shortfall during the irrigation season is determined as 72.8 106 m3, the indicatorincreases to 121 106 m3 in the BAU scenario due to low irrigation efficiency (0.60).IWD is equal to 97 106 m3 and 180 106 m3 in the OPT and PES scenarios, respectively.Considering the average annual crop water requirement in the system (300 106 m3),


which indicates an irrigation water demand in the order of 500 106 m3, it is obviousthat the deficit amounts presented above are enough to decrease the yield as well asthe revenues of farmers. However, significant improvements of IWD can be achievedwithin the proposed alternatives. Accordingly, IWD can be improved by 59%; inother words, almost 100 106 m3 less deficit is possible with the A3 or A6 alternativesunder pessimistic conditions. If we assume BAU or OPT scenarios, the alternativesproduce more decreases in irrigation deficit.

Izmir is the third largest city in Turkey and consumes groundwater resources ingeneral. Due to the population increase resulting domestic water demand, it has beenplanned to deliver additional water (60 106 m3/year) to Izmir via piped transmissionlink from Gordes Dam. The domestic supply reliability (DSR) refers to satisfactionof this demand which is determined as 5 106 m3/month, considering equal share ofannual demand. The DSR is equal to 0.80, 0.69 and 0.53 in OPT, BAU and PES,respectively. The indicator value under optimistic conditions (0.80) means that themonthly demand will not be fully met approximately for 3 months in the summerseason. Moreover, the risk of supply failure increases if we consider the pessimisticconditions. Again, significant recovery is possible with A3 and A6 alternatives.

With respect to economic indicators, the alternatives reflect different results.Under PES scenario, the alternatives A1 and A5 are determined as the most bene-ficial ones, where the B/C ratios are the same and equal to 1.24. This is higher than 1.1indicated by A3 and A6. Moreover, the alternatives A3 and A6 are under performedwith respect to A0 in all reference scenarios. In contrast to the B/C indicator, thealternatives A3 and A6 are better score than the others regarding irrigation wateruse efficiency (IWUE) which indicates the agricultural benefit per unit of irrigationwater allocated (e/m3). In the A3 and A6 alternatives, the percent improvementof IWUE in PES is determined as 26%, while A1 and A5 result in an increase ofonly 6%. This indicates improvements in the order of 22% and 11% in the totalproduction value (TPV) if we consider the alternatives A3 (or A6) and A1 (or A5),respectively. The improvement of TPV in alternative A4 is 17% in the PES scenarioand is also remarkable in other reference scenarios. It is observed that, although dripirrigation methods increase benefits and decrease water consumption by efficientwater use technologies, A3 and A6 do not appear to be beneficial in comparisonwith alternatives which have lower cost, such as canal maintenance (A1). It is alsoconcluded that the alternatives including drip irrigation methods probably performbetter if we extend the simulation period or implement the investments more rapidlythan that specified in this study.

6.3 Multi Criteria Decision Making

MCDM has been widely used in water resources literature as a major component ofdecision making. In this process, the Simple Additive Weighting (SAW) method isused with a normalization procedure (Eq. 9) to transform the performance valuesonto a commensurable scale between 0 and 1, where 1 represents the best perfor-mance. Because lower values are better for the minimizing indicators, the reciprocalsof aij, (1/aij), are used in normalization procedure. The equal weights representing


Table 5 Alternative rankingfor reference scenarios

Rank BAU OPT PES

1 A3 A3 A62 A6 A6 A33 A7 A4 A74 A4 A7 A45 A5 A5 A56 A1 A1 A17 A2 A2 A08 A0 A0 A2

the idea of objectivity are assigned to each indicator, being sum of 1. The ranking ofalternatives (Table 5) is identified on the basis of ui (Eq. 10).

rij = aij

max(aij), i = 1, 2, . . . . ., n, j = 1, 2, . . . . ., m (9)

ui =m∑

j=1

rij wj, i = 1, 2, . . . . ., n, (10)

For BAU and OPT scenarios, the management alternative (A3) that combinesthe replacement of the current water distribution network by piped systems todecrease water losses and the use of ‘water saver’ technologies such as drip irrigationto improve irrigation efficiency is determined as ‘best’ alternative with respect toenvironmental, social and economic dimensions through the use of a multi criteriaapproach. The maintenance of existing water conveyance network (A1) as well ascrop pattern change applications (A2) are not considered as adequate measuresfor coping with water scarcity. However, in pessimistic scenario, the rank orders ofalternatives are different, and A6 alternative is seen as the best alternative. Thatmeans, A3 alternative should be supported further by additional measures, such ascrop change applications, even if the conditions are worse than expected, in otherwords, even when the pessimistic scenario occurs. Interestingly, A2 alternative isdetermined as the ‘worst’ alternative in pessimistic scenario. The cumulative effectsof yield response factor and crop price are considering as the main reasons for thatresult. Regarding rank of A5 and A1 alternatives, one can say that they are fairaverage alternatives for every hydro-meteorological scenario.

7 Conclusion

Following from the above discussions, the major results derived from the scenarioanalyses of possible hydro-meteorological variations in the Gediz River Basin andthose evaluated on the basis of the proposed management alternatives can besummarized as the following:

1. The Basin is already under water stress and is also quite sensitive to droughtconditions. If the pessimistic conditions which lead to decreased water supplyand increased water demand occur, the resulting successive water deficits will


significantly affect the agricultural sector. Moreover, even when the optimisticscenario is assumed to occur, it is not possible to observe a significant improve-ment in the water budget. Accordingly, efficient water management policies arecrucial to solve water problems and to ensure sustainable development in theGediz River Basin.

2. Considering environmental, social and economic indicators, replacement of thewater conveyance system by pressured lines coupled with the application ofwater saver technologies such as drip irrigation methods, is determined as themost efficient management strategy for the Basin. With this strategy, it is notonly possible to minimize the negative impacts of droughts, but also to stabilizeor improve the current performance indicators. It should be noted that, theproposed alternative should be supported by additional measures, such as cropchange applications even if pessimistic conditions occur. On the other hand,the proposed alternatives should be the basic and long term policy for socio-economic development in the Gediz River Basin.

3. Since water transfer from Gordes Dam to Izmir is inevitable, the proposedalternative should be implemented as early as possible. This will ensure morebenefits in agriculture and will lead to economic achievements.

4. Although they are easy and/or cheap, the traditional measures such as change ofcrop pattern or reduction of losses in the current water conveyance system arenot considered as adequate and efficient responses for sustainable use of waterresources.

5. An interesting point achieved with case study is the remarkable consistencyrecognized between the current water management policies in the Gediz RiverBasin and B/C indicator. In other words, canal maintenance is observed to bethe most preferred alternative for both. This implies that a special emphasis isdevoted to B/C ratio in real life applications. However, environmental, social andeconomic sustainability should be all satisfied in management strategies. In thisregard, the developed methodology is a valuable tool for the assessment of waterresources systems and illustrates an efficient implementation of water resourcesmanagement approach.

6. The WEAP model is a potentially useful tool for planning and managementof water resources, and it provides a comprehensive, flexible and user friendlyframework for evaluation of management strategies.

7. For water resources management in developed countries, similar approacheshave been widely used, but have not yet been effectively implemented for otherriver basins of Turkey. It is recommended to increase the number of similarstudies that will also incorporate groundwater resources, water quality, industrialand domestic water demand, if adequate and accurate data is available.

References

Al-Omari A, Al-Quraan S, Al-Salihi A, Abdulla F (2009) A water management support system forAmman Zarga Basin in Jordan. Water Resour Manag 23:3165–3189

American Society of Civil Engineers (ASCE) (1998) Sustainability Criteria For Water ResourcesSystems. (Task Committee on Sustainability Criteria, Water Resources Planning and Man-agement Division, ASCE and Working Group, UNESCO/IHP IV Project M-4.3). ASCE,Reston, VA


Cetinkaya CP, Fistikoglu O, Fedra K, Harmancioglu NB (2008) Optimization methods applied forsustainable management of water-scarce basins. J Hydroinform 10:69–95

Chen Y, Zhang D, Sun Y, Liu X, Wang N, Savenije HHG (2005) Water demand management: a casestudy of the Heihe River Basin in China. Phys Chem Earth 30:408–419

De Voogt K, Kite G, Droogers P, Murray-Rust H (2000) Modeling water allocation betweenwetlands and irrigated agriculture: case study of the Gediz Basin, Turkey. International WaterManagement Institute, Colombo

Food and Agriculture Organization (FAO) (1979) Yield Response to Water—Yield andWater—FAO Irrigation and Drainage Paper 33, Chapter 5. http://www.fao.org/AG/AGL/aglw/cropwater/parta.stm. 27 January 2009

Guimarães LT, Magrini A (2008) A proposal of indicators for sustainable development in themanagement of river basins. Water Resour Manag 22:1191–1202

Harmancioglu NB (2004) Indicators for sustainable management of water resources (in Turkish).In: Proceedings of IV. National hydrology congress. ITU, Istanbul, pp 9–18

Harmancioglu NB, Ozkul S Baran T (eds) (2005) Network on governance, science and tech-nology for sustainable water resource management in the Mediterranean: the role of DSStools—NOSTRUM-DSS Project National Report II—Turkey. Project code: INCO–MPC-1-509158, Izmir

Harmancioglu NB, Fedra K, Barbaros F (2008) Analysis of sustainability in management of waterscarce basins: the case of the Gediz River Basin in Turkey. Desalination 226:175–182

Jeong CS, Heo JH, Bae DH, Georgakakos GP (2005) Utility of high-resolution climate modelsimulations for water resources prediction over Korean Peninsula: a sensitivity study. HydrolSci J 50(1):139–153

Koch H, Grünewald U (2009) A comparison of modelling systems for the development and revisionof water resources management plans. Water Resour Manag 23:1403–1422

Lévite H, Sally H, Cour J (2003) Testing water demand management scenarios in a water-stressedbasin in South Africa: application of the WEAP model. Phys Chem Earth 28:779–786

Loukas A, Mylopoulos N, Vasiliades L (2007) A modeling system for the evaluation of waterresources management strategies in Thessaly, Greece. Water Resour Manag 21:1673–1702

Moriasi DN, Arnold JG, Van Liew MW, Bingner RL, Harmel RD, Veith TL (2007) Model evalua-tion guidelines for systematic quantification of accuracy in watershed simulations. Trans ASABE50(3):885–900

OPTIMA (2006) Regional case study: Gediz River Basin, problem analysis, “Optimisationfor Sustainable Water Resources Management” (OPTIMA), co-funded by the EuropeanCommission FP6 Programme, 2002–2006. Project document: Deliverable D08.1. http://www.ess.co.at/OPTIMA/

Ozkul S (2009) Assessment of climate change effects in Aegean river basins: the case of Gediz andBuyuk Menderes Basins. Clim Change 97:253–283

Pallotino S, Sechi GM, Zuddas P (2005) A DSS for water resources management under uncertaintyby scenario analysis. Environ Model Softw 20:1031–1042

Silay A, Gunduz O (2007) Demirkopru reservoir water budget and irrigation performance assess-ment (in Turkish). In: Proceedings of V. National hydrology congress. Middle East TechnicalUniversity, Ankara, pp 103–112

SMART (2005) Regional case study: Gediz River Basin, Turkey, “Sustainable Management ofScarce Resources in the Coastal Zone” (SMART), supported by the European CommissionFP5 Programme (Contract number: ICA3-CT- 2002), 2002–2005. Project document: DeliverableD05.1. http://www.ess.co.at/SMART

Stockholm Environment Institute (SEI) (2007) WEAP Water Evaluation and Planning System.(User Guide for WEAP21). Stockholm Environmental Institute, U.S. Center. http://www.weap21.org. 20 December 2008

Varis O, Kajander T, Lemmela R (2004) Climate and water: from climate models to water resourcesmanagement and vice versa. Clim Change 66(3):321–344

Yates D, Sieber J, Purkey DR, Huber-Lee A (2005) WEAP21—a demand, priority, and preferencedriven water planning model. Part 1: model characteristics. Water Int 30:487–500

Walmsley JJ (2002) Framework for measuring sustainable development in catchment systems.Environ Manag 29(2):195–206

http://www.fao.org/AG/AGL/aglw/cropwater/parta.stm

http://www.fao.org/AG/AGL/aglw/cropwater/parta.stm

http://www.ess.co.at/OPTIMA/

http://www.ess.co.at/OPTIMA/

http://www.ess.co.at/SMART

http://www.weap21.org

http://www.weap21.org

an indicator based assessment for water resources ... · an indicator based assessment for water...

Documents