analysis integrated basin management: water and food ... · this paper presents a basin scale...

Analysis

Integrated basin management: Water and food policy options for Turkey☆

Serkan Gürlük a,⁎, Frank A. Ward b

a Uludag University, Department of Agricultural Economics, 16059 Bursa, Turkeyb New Mexico State University, Department of Agricultural Economics and Agricultural Business, Las Cruces, NM 88003, USA

a b s t r a c ta r t i c l e i n f o

Article history:Received 17 January 2009Received in revised form 28 March 2009Accepted 4 May 2009Available online 1 June 2009

Keywords:Water framework directiveBasin analysisClimate changeFood security

This paper presents a basin scale analysis of the Nilüfer River Basin of Turkey, where agricultural, urban, andenvironmental users compete for scarce water in an environment where climate change and food securitypresent large and growing challenges. It presents results of a basin scale dynamic nonlinear programmingmodel that addresses economic efficiency, climate change, and food security. Its approach can be applied toother water-stressed regions operating in environments of economic and hydrologic constraints on wateruse. Basin scale modeling approach provides a general framework for formulating water managementpolicies, consistent with the principles underlying the European Union Water Framework Directive.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

Water and its management influence human health, communities'welfare, and ecosystem sustainability. Despite water's essential char-acteristics, its scarcity has increased as a result of population growth,emergingenvironmentalvalues, andglobalwarming. Such impacts havebeen magnified as water demands for consumptive uses such asirrigation, municipal water supply and industry continue to increase.The Intergovernmental Panel on Climate Change (IPCC) reported thatgreenhouse gas emissions increase 70% between 1970 and 2004 (IPCC,2007). Changing climate conditions will have a considerable impact onthe hydrology and water resources of river basins.

Climate change will affect food security planning, especially ineconomies dependent heavily on irrigated agriculture. Every nationwishes to ensure access by its people to strategic food staples, andattempts to secure access to food on a reliable basis through interactionswith local markets and home resources (FAO, 2006). So the agriculturalsector is a special target of food security planning indeveloping countriessuch as Turkey. Impacts of climate change on food securitywill be higherin developing countries where irrigation, farming, and food processingare least efficient or most vulnerable to unexpected changes. Well-designedwater resource policies have considerable potential to improvethe allocation ofwater aswell as contributing to food security objectives.Improved water management and irrigation strategies, better basinmanagement and land use planning techniques can be instrumental in

addressing impacts of climate change. Managing to meet environ-mental demands for water can be an important contributor to economicefficiency. That is, protecting the water environment and waterecosystems from degradation can be a good investment, especiallywhere tourism-based outdoor recreation and ecotourism are significantcontributors to national income.

Efficient water resources management is based on improving theeconomic productivity of available water. Achieving economicefficiency requires understanding the availability of water and anotion of how much of it will be needed, in what quantity, for howlong, and for which purposes. Thus, it is necessary to have economicinformation and methods to increase total economic benefits byallocating water more efficiently (Ward and Lynch, 1996). Integratedriver basin optimization models provide this kind of mechanism.

The EuropeanWater FrameworkDirective (WFD), adopted in 2000,introduced an integrated approach towater management in Europe. Itestablished a common approach to protecting the water environmentand to setting environmental objectives for all waters of the EuropeanUnion (EU), and also provided a framework for designing future EUwater legislation. The main objective of the WFD is for member statesto achieve good water status for both surface and ground water and toprevent degradation of existing quality where good water status hasbeen achieved. Economic theories take important part in theWFD, andwill play a uniquely central role in characterizing how and to whatextent the WFD is implemented across Europe. Basin scale economicanalysis in theWFD is onemethod to provide important information tosupport major changes in the management of Europe's waters.

In recent years, much work has been conducted on integratedhydrologic–agronomic–economic models. Examples include modelsdesigned to support improved irrigation efficiency (Cai et al., 2003),tracking hydrologic and institutional constraints (Dai and Labadie,

Ecological Economics 68 (2009) 2666–2678

☆ The authors are grateful for financial support for this work by the Scientific andTechnological Research Council of Turkey, the Rio Grande Basin Initiative, and NewMexico State University Agricultural Experiment Station.⁎ Corresponding author.

E-mail addresses: [email protected] (S. Gürlük), [email protected] (F.A. Ward).

0921-8009/$ – see front matter © 2009 Elsevier B.V. All rights reserved.doi:10.1016/j.ecolecon.2009.05.001

Contents lists available at ScienceDirect

Ecological Economics

j ourna l homepage: www.e lsev ie r.com/ locate /eco lecon

fward

Pencil

fward

Typewritten Text

For non-commercial and educational use only. Please ask publisher for use permission.

fward

Typewritten Text

fward

Typewritten Text

fward

Typewritten Text

fward

Typewritten Text

2001), analysis of improved information at the basin scale to supportthe WFD (Van Ast and Boot, 2003), insights into measures forreducing nitrogen pressure to support both the EU Nitrates Directiveand the WFD (Fassio et al., 2005), nutrient management in the RhineBasin (Van der Veeren and Lorenz, 2002), and climate changeassessment at the basin and regional scales (Krysanova et al., 2007).

Many similar researches have been carried out in the world'svarious river basins. Some of them are the following: Analysis of theColorado River Basin, USA examined the economic value of competingwater allocation measures throughout the basin (Booker and Young,1994), while a study of the Rio Chama, USA examined economic valuesand tradeoffs between recreation and hydropower (Ward and Lynch,1996). An early celebrated analysis was conducted for the MissouriBasin, USA that examined reservoir operation rules for Corps ofEngineer projects (Lund and Ferreira, 1996). More recent work for thePo Basin, Italy conducted an integrated analysis of water regulation inconnection with the WFD (Bazzani et al., 2004), while integratedanalysis of the Syr Darya River, Central Asia addressed basin-scaleimplications of irrigation-induced salinity (Cai et al., 2003). Hanleyet al., 2006 estimated the benefits of water quality improvements forthe Motray and Brothock catchments in Scotland. In the Murrumbid-gee Basin, Australia Xevi andKhan (2005) examined tradeoffs betweenenvironmental and farming uses of water; for the Rio Grande, USA,Ward and Velazquez (2008) described a basin scale optimization toinformmanagement choices affecting water quantity and quality. Twovery recent works have examined basin-scale tradeoffs and choices inTurkey, one for the competition for industrial and irrigationwater usesin the Gediz Basin (Harmancioglu et al., 2008), and another for a seriesof hydroelectric and irrigation reservoirs in the Euphrates Basin inTurkey and Syria (Tilmant et al., 2008).While each of these studies hasmade considerable advances in basin-scale analysis, none of thosestudies take into consideration principles of economic efficiency ofwater resources with climate change and the related challenge postedby food security. This paper's aim is to address these existing gaps inthe literature. Current paper presents a basin scale optimizationanalysis that addresses food security, climate change, and environ-mental values ofwater. It describes thedevelopment and application ofa basin scale optimization model to support water policy analysis forNilüfer River Basin, Turkey.1 The approach described in this paper isconsistent with the principles described by the WFD for integratingeconomics into water management and policy (Heinz et al., 2007).

Turkey has applied for membership in the European Union (EU). Ifthat application is approved, Turkey will need to conform to EUlegislation in sectors such as agriculture, energy, economy, environ-ment, and transportation. Conforming to, applying, and implementingmeasures required by the EUWFDwill be required. Likemost of the EUmember states, Turkey faces a number of unprecedented challengesrelated to:

• Comprehensively assessing compliance with existing water man-agement laws and institutions

• Identifying cost effectiveness of various watermanagementmeasuresand assessing the size and distribution of their economic impacts

• Adapting existing institutions to ensure future water resourcesmanagement and planning is effective at the basin level

• Establishing open communications among people and institutionsthat contribute to the basin planning process

Basin scale analyses consistently and comprehensively applied, theaim of this paper, could contribute to the water planning process in

Turkey. Such analysis could provide a foundation for increasing thenation's water-related economic benefits. It could also demonstrateimplementation of the challenging water planning principles estab-lished by the EU WFD.

2. Study area and issues

The province of Bursa is located in northwest Turkey. The NilüferBasin is its most important renewable water resource supply. TheNilüfer River rises in the UludagMountains, and flows about 160 km tothe sea. It flows towards the city of Bursa, and is controlled by twoimportant reservoirs, the Nilüfer and Doganci.

According to 2000 census, with a population of 1,194,687, Bursa isTurkey's fourth largest city, as well as one of the most industrializedand culturally important metropolitan centers in the country. Aftersplitting Bursa, the Nilüfer River turns west and flows to the MarmaraSea. The Basin covers an area of about 1539 km2, of which 54% consistsof irrigated agriculture.

Karaer and Küçükballı (2006) state that the Nilüfer River has a poorwater quality in its many parts. Untreated domestic wastewaters,industrial discharges and agricultural activities contributed to the totalannual organic loads. That study revealed the importance of construc-tion, operation, maintenance and legislation of wastewater collectionand treatment programs, as well as the need for tighter control ofnutrient loads for the preservation of the Nilüfer's water quality.

The basin has experienced considerable economic and populationgrowth since the 1970s. The City of Bursa has an average householdsize of 3.8 (Census 2000). The Nilüfer River provides water to about428,000 households. The upper part of theNilüfer Basin has highwateruse in agriculture, partly induced by low water prices charged toirrigators. Water prices in irrigation are typically insufficient to coveroperation and maintenance costs (Çakmak et al., 2006). In the lowerreaches of the Nilüfer River, agricultural production was the mainsource of income until considerable water-using industrial growthappeared in the 1970s. The growth of these industries combined withincreased population has resulted in conflicts over water allocationand quality. Industrial pollution on the Nilüfer River is high, especiallydownstream of Bursa. In addition to industrial pollution, urban house-hold discharges have also contributed to water pollution. About 60% ofthe basin's population is connected to a sewage system (Karaer andKüçükballı, 2006). Therefore, a considerable amountof domesticwastewater is unprotected bywastewater treatment plants and is dischargeddirectly into the Nilüfer River and its tributaries, presenting seriouswater quality issues.

Reservoir water levels of the basin have historically determinedhaphazardly with little consideration to economic or environmentalimpacts. The issues described above characterize the importance ofdeveloping an integrated model that systematically accounts forinteractions of water supply, political constraints, and economic valuesof water. A basin scale model that optimizes the economic performanceof water allocations would help to understand contribution of variouswater sectors, includingurban, agricultural, and environmentalvalues ofwater. It would also support and inform the recovery of water servicecosts based on the economic analysis, a principle supported by the EUWFD.

3. Methods

An optimization model was developed to address the issuesdescribed above, with the intent of producing a single unified frame-work for policy analysis. The model consists of 9 river nodes, 9 inflownodes, 4 diversion nodes, 4 surface water return flow nodes and 5reservoir release nodes (Fig. 1). That model's economics accounts foragricultural demands for basin's major irrigated areas, recreationalvalues from the water environment, and urban-industrial demands forthe City of Bursa.

1 The model presents a long-run analysis of the economic benefits associated withagricultural, urban, and recreational uses of water. For agriculture, production costsinclude establishment, capital, operating, and replacement costs of all enterprises. Forurban uses, costs include capital, operating, and replacement costs of water supply anddistribution. For recreation, costs include management costs of supporting visitsassociated with reservoir fluctuations.

2667S. Gürlük, F.A. Ward / Ecological Economics 68 (2009) 2666–2678

3.1. Data sources

Data on headwater inflows to the basin were obtained from localState HydraulicWorks and Directorate ofWater and Sewage System of

Bursa. Table 1 indicates flow capacities of headwaters. Hydrologic dataare from the State Hydraulic Works (www.dsi.gov.tr/english) whileurban water use data source is the Directorate of Water and SewageSystem of Bursa (www.buski.gov.tr). Agronomic crop yield and water

Fig. 1. Nilufer Basin schematic.

2668 S. Gürlük, F.A. Ward / Ecological Economics 68 (2009) 2666–2678

use data were obtained from the Agricultural Directorate of Bursa(www.bursatarim.gov.tr). Recreational water use data are provided bythe Directorate of National Parks based on the Ministry of Environ-mental and Forestry (www.cevreorman.gov.tr).

Water uses include agricultural demands for the basin's three majorirrigation areas, urban demands for the City of Bursa and environmentaldemands for reservoir-based outdoor recreation. The hydroeconomicmodel of the Nilüfer Basin (HMNB) is a dynamic nonlinear program-ming model designed to maximize the present value of total neteconomic benefits from the water uses and environments describedabove subject to various hydrological and policy constraints.

Themodel simulates irrigationwater use demands through use of afarm management model based on production costs and returns forthe basin's major irrigated crops. In the model farmers are assumed tochoose a farm income (profit) crop mix, as well as a quantity of land,and quantity of surface water used for known crop prices, crop yields,and farm production costs. For urban uses, the total benefit of wateruse, total revenue fromwater sales plus consumer surplus is based on alinear demand function. That urban demand is calibrated with aknown price elasticity of demand, current price, and current averageuse per household. The equations used for themodel are similar to theones presented by Ward and Velazquez (2008). Environmental(recreation) benefits are based on a quadratic function of all recreationenvironments, defined as reservoir volume. These volume-basedbenefits increase up to the point of each reservoir'smaximumcapacity.The parameters used are based on an economic analysis of outdoorrecreation in that basin conducted by Gürlük and Rehber (2006).

3.2. Economic benefits

Water has an economic value in all its competing uses and is treatedin our analysis as an economic good. Both water use (depletions) andthe water environment (non-depletions) have an economic value touserswho arewilling to pay for it.2Water userswill expand their use aslong as the benefits from use of an additional cubic meter depleted orstored in a reservoir exceed the additional costs incurred. Benefits aredetermined by water users' total willingness to pay (WTP). They aredefined as the maximum amount of income that a person will pay inexchange for an improvement in conditions, or the maximum amounta personwill pay to avoid a decline in those conditions (Briscoe, 2005;Haab and McConnell, 2002; Young, 2005).

For agricultural uses, the WTP for water depletions is measured bythe contribution of water to farm income (Bruinsma et al., 2003; Lienand Hardaker, 2001). For urban uses, water's value is measured byprice per unit water times the number of units sold to the householdplus any unpriced consumer surplus. For environmental benefits, WTP

is measured as the price that could be charged to visitors of allreservoir-based recreational activities. Environmental values of thewater environment can be separated into use and non-use categories.Use values are use-related benefits such as scenic beauty, and on-siterecreational activities while non-use values include benefits such asoption, existence or bequest values. In this study, several importantnon-use values are not measured, although they are likely affected byfluctuations in reservoir levels as well as variations in stream flows.Good examples of analysis of non-use values are the works of thestudies works by Greenley et al. (1981), Boyle et al. (1994), Carey andZilberman (2002), Cooper et al. (2004), and Turpie et al. (2003).

3.2.1. Use-related benefitsBenefits related to water use were measured for agricultural and

urbanuses at the Basin'smost importantnodes. For agricultural uses,weconstructed water demand functions through the use of farm manage-ment income optimization models. These models were estimated andcalibrated to produce optimized cropping patterns consistent withhistorical observed patterns (Michelsen et al., 1999; Ward andVelazquez, 2008; Vreke, 1990; Ward et al., 2001). The Nilüfer Basincontains three important irrigated areas, the Nilüfer Irrigations District(NID), the Misi Irrigation District (MID) and the Cayirkoy IrrigationDistrict (CID).

The land area of NID is suitable for a wide variety of produce,including strawberries, cherries, grapes, and several vegetables. The NIDmakes up the largest part of the irrigated area, consisting of 54% of thebasin's total irrigated area in production. MID is downstream of NID. Inaddition to being an agriculturally productive area, the MID containsconsiderable scenic beauty and outdoor recreational facility. It is themajor organic farming center of Bursa City, and provides irrigation of awide range of produce such as olives, peaches, apples, and cherries.Abovementioned agricultural products are themost cultivated productsof the Bursa Region. CID has also been a historically importantagricultural production center for Bursa City. This irrigated area isproductive for both produce and food staple crops. Recent industrializa-tion and rapidurbanizationnearbyhas resulted in environmental issues,with consequent higher production costs due to water purificationexpenses. The CID makes up only 7.5% of the basin's total irrigated land.

For urban uses in the City of Bursa, a linear demand schedule wasspecified to characterize the demand for water. The estimated priceelasticity was −0.028, based on long-term observations on price anduse patterns. Using that elasticity combined with the 2007 averageprice and total use, we computed the integral of the linear demandwater use to measure total benefits of urban use. A linear demandfunction results in a quadratic total benefit function, for which thosetotal benefits are maximized at the level of water consumption thatoccurs at a zero price. The following equationwas used to characterizethose urban benefits:

Buut = B0u + B1uxut + B2ux2ut ð1Þ

where Xut at the uth urban use node (Bursa City) is water use and theBs are constant parameters. Urban water customers use water forhigh-valued applications such as drinking, cooking and hygiene, sourban benefits rise quickly for supplies allocated to these uses, startingfrom a position of no use. For these high-valued necessities, there arefew substitution opportunities available. These circumstances leadone to expect that B1u is large and positive. When lower-valued uses ofwater increase, such as outdoor landscapes, car washing, sidewalkcleaning, urban marginal benefits fall rapidly, which means that is B2uis large and negative.

Low-value urban water applications exist in the conditions ofabundant water supplies and low prices. Where severe droughts andurban water rationing occur, as has occurred in several areas of Turkeysince 2002, marginal benefits of urban use rise rapidly. More detailedanalysis is shown in various microeconomic theory and water resources

Table 1Headwater flows of the Nilufer River basin, Turkey.

Source Ave annual flow (1000 m3)

Uludag Headwater 520,000Yaylalar 52,000Karadere 117,390Aras 146,900Karan 260,000Maden 98,925Yaylacik 175,770Bostanli 357,800Hasanaga 124,700

2 Crop yields are based on ET, so irrigation benefits only occur when the cropdepletes the water. Irrigation return flows get used again downstream. Urban benefits(wtp) are based on the quantity of water treated and diverted. However urban use bythe City of Bursa has a low consumption rate, returning about 80% of the waterdiverted, suitably treated, back to the stream. Recreational uses are based on reservoirvolumes, from which only a small part is depleted as evaporation.


economics text books such as Griffin (2006), Intriligator (1971),Nicholson (1998), Silberberg and Suen (2001).

3.2.2. Environmental benefitsWater ecosystem services are an important environmental

resource. These services provide various use and non-use benefits.In this study environmental benefits are measured by theWTP for theimprovement of the basin's recreational facilitates. These facilitiesprovide numerous activities at each site such as sport fishing, scenicviewing, picnicking, and bird watching. Based on previous environ-mental economic analysis in the basin, a benefits transfer protocol wasset up to characterize the benefits of environmental quality improve-ment. In this method, results of earlier research are applied system-atically to approximate a total environmental benefits functionapplicable to the target sites. A number of authors such as (Batemanet al., 2002) describe this kind of transfer system. Benefits transfersare conducted because of the high administration costs of carrying outprimary survey surveys at each site separately.

We used the results of publishedwork byGürlük (2006), inwhich aregional Contingent ValuationModel (CVM)was applied to determinethe demand for improving the recreational activities described above.The CVM provides data to compute total gross economic values ofselected reservoir levels. The CVM assumes that visitors of outdoorwater based recreation sites receive environmental benefits fromwater at each of the basin's reservoir, and are willing to pay for thewater provided. More information and details on CVM are described inBateman et al. (2002), Cameron (1992), Carson et al. (1998) andWard(2005).

For each reservoir in the Basin, gross recreational economicbenefits for an improved water-based environment are formulated as:

Bert = B0e + B1ezrt + B2ez2rt ð2Þ

where zrt is the storage volume of water in the rth reservoir held instorage for the tth year. The B coefficients are based on earlierpublished work by Gürlük (2006). In that study, the recreationalfacilities described above are investigated for the economic value oftheir water-related improvements through use of the CVM. Thequadratic functional form for benefits was applied to reflect totalbenefits at any reservoir storage volume increases up to a point inwhich incremental recreational benefits are zero.

3.3. Costs

3.3.1. Use-related costsFor each use-related node, increased water use typically require

water supplier to incur additional costs to make that water use suitablefor human use. The greatest costs for depletingwater at irrigation nodesare those incurred for operation and maintenance of the irrigationdistrict. Urban nodes have considerable costs for purification to makethe water safe for human consumption.

Urban delivery cost data are obtained from the City of Bursa'swater utilities, while agricultural water costs were obtained from theAgricultural Directorate. These costs are included in the objectivefunction as negative terms for which costs are subtracted frombenefits, described in a subsequent section. Based on these costs, thetotal cost of water delivered to any agricultural or urban node is:

XCuut = δuXut ð3Þ

where management cost per unit volume is defined as δu. Because ofthe high costs of purification, those per unit costs are considerablyhigher for urban uses than for irrigated agriculture. As describedabove, the variable Xut is the volume of water depleted from the riverin the tth period. That quantity of water is unknown in advance of a

model run, for which its optimized values are determined by themodel's solution.

3.3.2. Environmental costsEnvironmental costs produced by a range of water-related activities

continue to be a large and growing factor influencing outcomes ofTurkey'swater policy decisions. Twoprincipleswere used formeasuringenvironmental costs: (1) opportunity costs, equal to gross environ-mental benefits displaced by a decision, and (2) environmentaloperations costs, equal to the cost of additional resources required tosupport securing and protecting increased environmental benefits. Wecombined both bymeasuring each reservoir's environmental benefits asgross environmental benefits minus added marginal gross environ-mental costs needed to secure a higher quality environment. In thisbasin, data are scarce on costs of protecting the water for environment.As a first approximation, we measured those costs as marginal costsincurred by sitemanagers tomaintain fishing and picnicking facilities inthe face of reservoir volume increases. These environmental operationscosts were measured as:

XCert = δerZrt ð4Þ

where δer is the marginal cost of managing larger volumes of water atany reservoir site in the basin, while Zrt is the reservoir's storage volumein period t. For this study, the term δer was measured as the additionalcosts tomanagers associatedwith supportingmarginalvisitors attractedby larger water quantities held in a reservoir. We used data on thesemanagement costs, available at Gürlük (2006). A more comprehensivetreatment of marginal costs of environmental improvements wouldinclude anyoperations costs needed to preserve, protect, or improve anynatural environment.

3.3.3. Resource benefits and costsThe EU Water Framework Directive may be one of the first set of

published principles explicitly calling for an assessment of resourcebenefits and costs as components of the full cost to be recovered.However, the definition and practical assessment of resource benefitsand costs remains controversial, partly because it requires considerablemethodological development. (Brouwer, 2004; Görlach and Interwies,2004; Pulido-Velazquez et al., 2006). Oneway to bind a resource benefitis by a resource cost avoided. The value of that avoided cost canbe measured in principle as a willingness to pay to avoid the risk ofirreversible damages to a resource's environmental or use capacity.Economic cost analysis of irreversible risks was originally developed byCiriacy-Wantrup in the early 1950s, with special attention directed tothe cost of species extinction. Methods for measuring these costshave been debated many times since Wantrup's original work, oftenusing the principle of the safe minimum standard (SMS) as a targetof environmental and natural resource policy (Ciriacy-Watrup, 1952;Bishop, 1978; Crowards, 1998).

We account for resource costs in the current analysis by the simpleconstraint that the critical zone is avoided. Our implementationmethod is to constrain the basin model to return in the last period allbasin reservoirs to a water level with at least as high as its level in thefirst period. This method assumes that no basin reservoirs currentlyoccupy the critical zone.

3.4. Net benefits

Both agriculture and urbanwater produce economic net benefits tousers (use-related benefits). For these nodes total net benefits aredefined by subtracting use-related costs from use-related benefits:

XNBuut = XBuut − XCuut ð5Þ

In conditions of plentiful streamflows, net benefits can bemaximized at each use node, while volumes and flows at other


nodes can be ignored. However, in water stressed periods, no singlenode is likely to have sufficient water to be able to maximize thatnode's net benefits because of the competition for and high value ofwater at competing nodes.

The net economic benefits of the water environment at reservoirsare defined as:

XNBert = XBert − XCert ð6Þ

Where the first right hand side term is the gross recreation benefitsassociated with various reservoir levels and second right hand sideterm is the management cost of managing larger numbers of peopleand related facilitates associated with greater reservoir storagevolumes. Increases in reservoir volumes increase the environmentalbenefits at the reservoir up to the point where the marginal cost ofmanaging environmental improvements equals the marginal benefitof additional reservoir volume. However use related net benefits,which require water depletions, typically reduce environmental netbenefits. Especially in water-stressed basins, water use and the waterenvironment compete both for water and for other resources. As is thecase in many of the world's basins, the Nilüfer Basin's environmentalnet benefits (supported by reservoir stocks) conflict with use-relatednet benefits (supported by flow depletions). So the policy challenge isto identify the pattern of reservoir storage and releases that maximizethe overall net benefits over the basin's three sectors.

3.5. Discounted net benefits

Discounted net present value is defined in its typical algebraicform:

XNPV =Xu

Xt

XNBuut

1 + ruð Þt +Xe

Xt

XNBeet1 + reð Þt ð7Þ

Discounted net present value, measured in this way, includes thesummed flow of net-use related benefits and environmental netbenefits. Our discounting is based on the standard algebraic computa-tion.Weused a 5% real (inflation adjusted) rate, reflecting our estimateof Turkey's recent real cost of private capital. The discount rate is a veryold and highly debated issue. Rowlatt et al. (1998) and Pearce andUlph(1999) present good examples of this ongoing debate.

3.6. Solving the model

The model is formulated as a dynamic nonlinear program, maxi-mizing discounted net present value summed over water uses, water

locations, and time periods. The objective function sums benefit fromwater flow depletions and environmental stocks. The General AlgebraicModeling Software (GAMS)was used to compute the solution (GAMS®,2007). The model's dimensions are 4532 variables, 3697 equations and15,080 nonzero elements for the 20 yearmodel. Discounted net presentvalue over themodel's time horizon is maximized, inwhich constraintsare defined by the Basin'swater supply, agronomic cropwaterdemands,hydrologic continuity, and land constraints.

4. Results

4.1. Overview

Results are presented for conditions defined by depleted reservoirconditions starting in 2007, and include 20-year projections. We focuson two policy options implemented under each of two states of nature,for a total of four scenarios. The scenarios include two food securityprograms and two climate change conditions, all of which affecthydrologic, agronomic, and economic outcomes.While results are basedon long term average historical inflow levels to the basin, the methodhas a completely flexible structure suitable for future analysis of a rangeof hydrologic conditions and policy choices that could face the Basin'spolicy makers. The model offers considerable economic insight intoimpacts of changing hydrologic conditions and of institutional adjust-ments for coping with those altered water conditions.

4.2. Hydrologic, agronomic and economic performances of four programs

4.2.1. With food security and with climate changeTable 2 shows results of a food security program under conditions

of future climate change. Food security is defined and implemented inthe model by the requirement that food staple crops are required on aset part of the basin's total irrigated land in production. Climatechange conditions are defined by reduced basin headwater flows,falling by 20% of the long term historical record beginning in the tenthmodel year (2017). That is, we implement food security requirementsin all years, while climate change conditions are faced only beginningin the eleventh year.

Hydrologic outcomes are shown in the first four rows. Theseoutcomes reflect the economic benefit-maximizing system operationof reservoir storage, reservoir release and water use patterns thatoccur when a policy of food security implemented under conditionsstressed by climate change starting in 2017. Results show about 13% ofthe water depletions allocated to urban uses with the remainder ofdepletions assigned to agriculture. Among the irrigation districts, theNilüfer Irrigation District (NID) and Misi Irrigation District (MID) are

Table 2Hydrologic, agronomic, and economic performance of basin water management: with food security and with climate changea.

Sector Urban Agriculture Environmental Total

User City of Bursa Nilufer Irrig. Dist. Misi Irrig. Dist. Cayirova Irrig. Dist. All reservoirs Basin

Water outcomes (1000 m3/year)Headwater flows 1,575,462Water volumes 180,115 180,115Water depletions 46,219 186,045 114,171 21,893 368,329Flows to the sea 1,249,366Land outcomes (1000 ha)Land in staple cropsb 15.4 18.1 4.7 38.2Land in producec 26.1 26.1 4.2 56.5Total land in production 41.6 44.3 8.9 94.7Economic outcomes (1000 €/year)d

Benefits 922,572 444,909 217,731 35,755 33,643 1,654,610Costs 47,035 98,495 60,444 11,591 530 218,093Net benefits 875,538 346,414 157,287 24,165 33,113 1,436,517

a All tabled entries are averages for 2007–2026.b Staples: wheat, barley, beans, corn, potatoes.c Produce: watermelon, olives, cantaloupe, tomatoes, peaches, grapes, strawberries, cherries, apples, pears.d 1€=1.70 TL.


the largest water users. Çayırova Irrigation District (ÇID) has less landarea and less water use. Its higher water supply costs give rise tohigher effective water prices and reduced use per acre.

A national food security requirement is implemented by therequirement that at least 30% of the basin's irrigated land is allocatedto food staples: wheat, barley, beans, corn and potatoes. Under thatrequirement, land under irrigation producing staple crops appears inall irrigation districts. However, not all districts produce an equalpercentage of food staples, since the model is designed to minimizethe cost of implementing the food security requirement. As a result,staple production is allocated to those areas that can produce it atminimum economic cost of other water related benefits displaced.Generally, staple crops in Turkey produce a lower income for farmersthan is earned by growing produce. So food security is achieved at acost of farm income displaced from the considerably more profitableproduce crops.

From the view of the basin, constrains imposed by food securityrequirements combined with reduced basin inflows under a climatechange scenario produce the lowest net economic benefit among fourwater programs we examined. While urban water use consumesconsiderably less water than agricultural irrigation, it produces amuch higher economic value per unit of water depleted because of themuch lower price elasticity of demand for urban than for agriculturaluse reflecting widespread international trends in irrigated comparedto urban areas. Urban water uses in this basin produce about 60% oftotal net economic benefits, reflecting the intense competition for

water between urban and agricultural uses in the world's dry places.By comparison, agricultural and environmental (reservoir-basedrecreation) water use produces produce only 36% and 4% of totalBasin's net economic benefit, respectively.

4.2.2. With food security and without climate changeTable 3 shows hydrologic, agronomic, and economic outcomes for a

scenario with food security implemented in an environment whenfuture climate change is avoided. In these conditions of comparativelyplentiful water, water use increases by of 2.4%, 13.1% and 5.9% for thethree irrigation districts NID, MID and ÇID, respectively. However,comparativewateruse rates amongsectors stayabout thesameas for theprevious scenario. Water flowing to the sea increases by 20% comparedto the previous scenario. This finding illustrates the importance ofrecycled water as a water conservation measure, consistent with similarfindings of (Önder and Önder, 2006).

The requirements of national food security have an importantinfluence on water use and its allocation. Large amounts of land areaallocated to staple crops occur in this scenario's results despite itslower economic return and reduced total economic benefit. Comparedto the first scenario, results show greater land allocated to non-stapleproduce in the districts with higher overall profitability, such asNilüfer and Misi. Land allocated to fresh produce increases by 100% inthe NID while the equivalent percentage is 31 at the MID. In spite ofthese results, MID produces more economic benefit than NID. Thereason for this lower agricultural benefit at the NID is because

Table 3Hydrologic, agronomic, and economic performance of basin water management: with food security and without climate changea.

Sector Urban Agriculture Environmental Basin total

User City of Bursa Nilufer Irrig. Dist. Misi Irrig. Dist. Cayirova Irrig. Dist. All reservoirs

Water outcomes (1000 m3/year)Headwater flows 1,853,485Water volumes 160,674 160,674Water depletions 47,963 190,453 128,014 23,177 389,607Flows to the Sea 1,506,374Land outcomes (1000 ha)Land in staple cropsb 17.7 14.1 6.4 38.2Land in producec 52.6 34.4 3.3 90.4Total land in production 70.3 48.6 9.8 128.6Economic outcomes (1000 €/year)d



Table 4Hydrologic, agronomic, and economic performance of basin water management: without food security and with climate changea.

Sector Urban Agriculture Environmental Basin total

User City of Bursa Nilufer Irrig. Dist. Misi Irrig. Dist. Cayirova Irrig. Dist. All reservoirs

Water outcomes (1000 m3/year)Headwater flows 1,575,462Water volumes 180,115 180,115Water depletions 46,219 191,045 122,698 14,703 374,666Flows to the sea 1,243,734Land outcomes (1000 ha)Land in staple cropsb – 0.2 0.0 0.1 0.2Land in producec – 66.3 43.7 5.3 115.2Total land in production – 66.5 43.7 5.3 115.5Economic outcomes (1000 €/year)d

Benefits 922,572 546,283 373,283 46,574 33,643 1,922,355Costs 47,035 101,141 64,958 7784 530 221,448Net benefits 875,538 445,142 308,325 38,790 33,113 1,700,907



increasing land used for producing staple crops while land in staplecrops decreases at the MID. The urban benefits of water have thehighest percentage of total net benefits compared to the first scenario.Total net benefits increased by about 14% without climate change,about €61 million (79.3 million USD) annually.

4.2.3. Without food security and with climate changeTable 4 shows the overall performance at the basin scale without a

food security requirement in which of 30% of overall irrigated land isallocated to food staple crops. This scenario presumes that the nation'sfood security objectives will be met by other means, such as increasedfood imports, possibly paid for by exports of higher-valued freshproduce. Consequently, avery small part of lands in thebasin take part instaple cropproduction due to economic objectives of rational producers,grown on land well-suited economically for food staples. For thisscenario, land allocated to fruits and vegetables increases by more than100% compared to the first program and increases by 28% compared tothe second. Compared to the first two programs, this scenario increasesagricultural economic performance in income from € 560,979 (729,272USD) and € 620,823 (807,069 USD) to a much higher €825,369(1,072,979 USD). Eliminating the food security requirement of con-straining domestic production of staples increases total economicbenefits considerably, as larger amounts of land are allocated to high-valued produce. Total net benefits increase by 61% and 41%, respectivelycompared to benefits achieved under the first two scenarios. Economicbenefits contributed by urbanwater uses see almost no change due to aconsiderably lower price elasticity of demand for urban use than foragricultural water use. The share of urban water uses making up totalbasin-wide economic benefits are between 51% and 60%.

4.2.4. Without food security and without climate changeThe final scenario analyzed is completely free of constraints on

food security through domestic staple production and completelylacking water supply limits caused by climate change, with resultsshown in Table 5. This scenario sees an increase in total basin-scaleeconomic value of water use considerably compared with the otherthree scenarios analyzed. Urban water use decrease by less than 1%.Compared to the second scenario with food security, this scenariobrings greater amounts land into output of fresh fruits and vegetables.However agricultural benefits of this program are higher than thesecond program by an amount of 47% thanks to much larger inflows ofwater to the basin. Compared to the third programwith food security,this program producesmore total net economic benefits by an amountequal to €91.2 million. This difference indicates the sum of cost ofrequiring food security combined with strong future water suppliesdue to a lack of climate change impacts.

4.2.5. Tradeoffs among programsComparing the hydrologic, agronomic, and economic consequences

of a rangeof scenarios illustrates the importance andutilities of thebasinto policymakers associated with basin scale analysis. The first and thirdprogram presents policies under the climate change conditions (withversus without food security) while the second and forth programspresentpolicies under the food security conditions (with versuswithoutclimate change). Developing countries often face a felt need to achieveself-sufficiency in production of food staples for political, economic, andhumanitarian reasons. This self-sufficiency is believed to be important,even if its costs are considerable in income lost from other uses of itsagricultural infrastructure. Therefore, information on these costs isimportant for informing food and water policy debates.

The comparison of the first-third and second-fourth programspresents important information on the economic cost of food securitypolicies in the Nilüfer Basin. The economic cost of food security is€264.4 million (343.72 million USD) annually in the Basin in climatechange conditions. Without climate change, food security costs thenation €294.6 million (382.98 million USD) per year in basin scalewater-related economic benefits displaced. What this means is thatunder a climate change condition of reduced headwater flows,sustaining a policy of food security incurs more expense in terms ofbenefits displaced. With these considerably higher costs, alternativemeasures to ensure food security, such as increased long termtechnological investments on agriculture, are worth exploring as amatter of national food policy. However the investment costs ofdeveloping additional technological infrastructure can be considerablefor developing countries such as Turkey. According to RegionalEnvironmental Center for Central and Eastern Europe (REC), which isan important EU environmental organization, Turkey can haveadvantages from fully implementing EU directives for agriculturaland environmental sectors. The EU funds and laws on water andenvironment will provide directly and indirectly to Turkey. Althoughthe technological and institutional investments are considerable(€264.4–294.6 million) for Turkey, the annual value of these invest-ments is estimated by between €3.1and €15 billion (4.03 billion USD–19.5 billion USD) for Turkey (REC, 2002).

Another important finding from our work can be seen by com-paring the economic value of water for urban use compared toagricultural uses. Water has a much higher economic value for urbanuses than for agriculture for all the four scenarios presented. Water inurban use creates about €19 per cubic meter while this value is under€2 per cubic meter for use in irrigated agriculture. Both of these areaverage values per existing cubic meter, much higher than equivalentmarginal values per additional cubic meter. This outcome shows theimportance of efficient water pricing for agricultural uses as ameasure

Table 5Hydrologic, agronomic, and economic performance of basin water management: without food security and without climate changea.

Sector Urban Agriculture Environmental Total

User City of Bursa Nilufer Irrig. Dist. Misi Irrig. Dist. Cayirova Irrig. Dist. All reservoirs Basin

Water Outcomes (1000 m3/year)Headwater flows 1,853,485Water volumes 160,674 160,674Water depletions 47,963 198,209 133,952 26,619 406,743Flows to the sea 1,491,142Land outcomes (1000 ha)Land in staple cropsb – 0.2 0.0 0.1 0.2Land in producec – 68.9 47.8 9.7 126.5Total land in production – 69.1 47.8 9.8 126.7Economic outcomes (1000 €/year)d




to allocate water efficiently among urban, agricultural, and environ-mental uses of water, results corroborated by Çakmak et al. (2006).

5. Conclusions

This paper has examined the policy-informing role that canbe played by the development and use of basin scale models. Bycombining and integrating elements of hydrology, agronomy, andeconomics, use of these models has considerable potential to supportthe design of better-performing policies. The framework developed inthis paper was applied to assess impacts of two policy instrumentsunder each of two conditions. Both a more food secure as well as a lessfood secure future are examined under conditions of abundant versusscarce future water supplies.

Our findings suggest that economic impacts of climate change areconsiderable both with and without requirements of food security.Under all scenarios, the average economic value of water varies widelyamong sectors. Urban water use presents a considerably higher neteconomic benefit than either agricultural or environmental values.However, current environmental values of water may be low in Turkeybecause of low average per capita incomes, for which future growthcould considerably increase the economic value of sustaining waterenvironments.

Principles underlying the EU Water Framework Directive provideguidance to increasing the economic value of water facing watermanagers at the basin scale. Those principles present opportunities toprotect the water for environment, for which the implementationmethods described in this paper, provide a framework for achievinggood water status in other river basins.

Several important assumptions influence our results: The modelassumes the presence of market-like arrangements, such as waterrentals or transfers, by which scarce water moves to its highest-valueduses, constrained only by hydrologic relations and by existing waterallocation rules. This paper has presented only two policy alternativesin each of two environmental indicators including water supply. Infuture work, we hope to consider more policy choices and a largernumber of environmental indicators.

The model is currently limited to a yearly time step. Monthly timestep would permit a more precise analysis for informing importantcurrent policy debates like climate change, intraseasonal water storageand allocation, and awider range of alternative measures for promotingfood security. The model also has a limited treatment of environmentalvalues of water, currently addressing only outdoor recreation visitors'use values from reservoir level fluctuations. Important non-use valuessuch as option, existence, and bequest values of the complete waterenvironment are ignored. The model is deterministic. A more refinedmodel will account for stochastic inflows, prices, and water demands aswell as for risks of any of these factors failing to measure up to waterusers' and managers' expectations. All these limits point to avenues bywhich future research could improve the model's capacity to supportwater policy design at the basin scale.

Despite the limits of our model, the importance and complexity of21st century water policy debates assign considerable importance tothe use of basin scale analysis for informing debates, such as the oneconducted for this paper. Basin scale analysis provides a comprehensiveframework for informing the design of measures that produce anefficient distribution of economic benefits and costs of water programs.

Appendix A. Integrated watershed model for Nilüfer Basin, Turkey

A.1. Introduction

The model was developed to support analysis of policy optionsaffecting the use of the water resources of the Nilüfer River Basin,Turkey (the Basin) for multiple uses, including urban water supply,irrigated agriculture, and the environment. A related objective was to

assemble a comprehensive database for the Basin informing policydebates on the development, conservation, use, and management ofwater resources. It was designed to analyze and assess selected policyoptions based on their cost, water demands, affects on water supply,and long-term sustainability.

A.2. Hydrology

The essential principle of the hydrology model is mass balance,both for surface flow interactions and reservoir levels. The hydrologymodel described below uses mass balance principles to account forheadwater flows, river flows, reservoir levels, water from surfaceapplied to various uses, and the impact of surface flows on current andfuture reservoir storage's levels.

A.2.1. Headwater runoffInflows into the Basin are defined as total annual flows at the

various headwater gauges. Inflow at hth headwater gauge and year t,Xht, equals total source supplies:

Xht = Sourceht : ðA1Þ

A.2.2. StreamflowAs shown in this paper's figure, the Nilüfer Basin has several

stream gauges. Flow at each vth gauge in period t, Xvt, equals the sumof flows over any upstream node whose activities directly influencethat flow. These include: (1) headwater inflows; (2) upstream rivergauges; (3) upstream diversions; (4) upstream surface return flows;(5) upstream reservoir releases. Total flows, which cannot be negative,are defined for each of those six types of nodes, respectively, as:

Xvt =Xh

BhvXht +Xv

BvvXvt +Xd

BdvXdt +Xr

BrvXrt +XL

BLvXLt :

ðA2Þwhere the set v defines all gauges, and Xvt is the flowat any river gaugenode (element of the set v). Each of the six vectors of B coefficientstakes on values of 0 for non-contributing upstream sources, 1 forsources that add flow, and−1 for sources that reduce flow. So, positivesigns in an equation (+) require adding flows, and subtractions (−)occur whenever a B coefficient is negative. For example, the first term,Xh

BhvXht ; sums contributions over the set (h) of headwater nodes. The

vector Bhv contains 1 s for all immediately upstream headwater gaugesthat contribute to a river's flow and 0 otherwise, where Xht are flows atall headwater gauges. The second right-hand side term,

Xv

BvvXvt ;

sums contributions over the set (v) of relevant upstream river gaugeelements. The vector Bvv typically contains a single 1, and the restzeros. The third term

Xd

BdvXdt ; sums riverflow reductions over the set

(d) of upstream diversion nodes. By accounting for upstreamdiversions, the Bdv coefficients are 0 for non-diverting locations andfor diversions that do not affect the given node's flow, but −1 whereupstream diversions directly reduce that flow. The last three termssimilarly account for; upstream surface return flows in the set (r),unmeasured use in the set (m), and upstream reservoir releases in theset (L) that affect river flows.

A.2.3. Water divertedAgricultural uses are supplied by direct stream diversions. There is

little groundwater pumping for irrigated agriculture in most areas ofTurkey. Still, in many of the world's arid regions, the historical recordoften shows zero flow in dry periods when there is high demand.The following equation, a “wet water” condition, requires that nodiversion exceed available river flow at the point of diversion.Therefore, for the river to be wet, each diversion must be less thanthe sum of all six classes of upstream sources: (1) headwater inflows;


(2) upstream river gauges; (3) upstream diversions; (4) upstreamsurface return flows; (5) upstream reservoir releases. A diversion (dsubscript), which cannot be negative, is:

XdtVXh

BhdXht +Xv

BvdXvt +Xd

BddXdt +Xr

BrdXrt +XL

BLdXLt :

ðA3Þ

The right hand side terms are the sum of all contributions to flow atthe point of diversion from upstream sources. The various B terms,which indicate presence (1) or absence (0) of upstream flow sourcesfor a given node, are used to configure the basin.

A.2.4. Water appliedLike diverted water, total water applied to a use at any node in

period t, Xat, is a choice variable. Water applied can come from twosources: a stream diversion, Xdt or water pumped, Xpt.3 Total waterapplied is:

Xat =Xd

BdaXdt +Xp

BpaXpt : ðA4Þ

Both sets of coefficients Bda and Bpa are identity matrices toconform like nodes in the basin. For each agricultural node in thebasin, total water applied to farmlands is expressed as:

Xat =Xc

Bac

Xu

BuaLuct : ðA5Þ

Total irrigation water applied from surface flows at each ath waterapplication node in the tth year equals total water demands. Thesedemands are summed over crops (c) based on known waterapplication amounts per unit land by crop, Bac. The result is multipliedby an identity matrix, Bua, that conforms nodes and the quantity ofland irrigated at the uth use node by the cth crop in the tth year, Luct.The optimal solution of a single model run determines the totalquantity of irrigated land, Luct, which determines the total demand forirrigation water applications, Xat.

A.2.5. Water consumedAny use node's consumptive use, Xut, is an empirically determined

proportion of total water applied, Xat. For irrigation, consumptive useis the quantity of water lost through plant evapotranspiration (ET) toany future use in the system. That use, which cannot be negative, ismeasured as:

Xut =Xa

BauXat : ðA6Þ

The parameters Bau are elements indicating the proportion of totalwater applied that is used consumptively. Setting Bau at 1.00 for thecurrentmodel implementation achieved the best hydrologic balancingwhen comparing model-predicted use against actual river flows andland under irrigation. For agricultural nodes, water use ismeasured as:

Xut =Xc

Buc

Xu

BuuLuct : ðA7Þ

Irrigation ET at the uth agricultural node in the tth year is derivedfrom total quantity of land in crop production. That water use ismeasured as the sum over crops (c) of empirically estimated ETamounts per unit land by node, crop and Buc, times an identity matrix,Buu, that conforms nodes. The result is multiplied by the quantity of

land irrigated. The amount of land irrigated by use, crop, season, andtime is determined by the model's optimal solution.

A.2.6. Gross surface returns to riverFor agricultural nodes, total surface returns to the river are mea-

sured as:

Xrt =Xc

Bcr

Xu

BurLuct : ðA8Þ

A.2.7. EvaporationReservoir evaporation is measured as:

Xet =Xr

BreZart : ðA9Þ

That is, any period's evaporation from a reservoir equals the, Bre,multiplied by the average surface area exposed at the reservoir thatperiod, Zart. At any given reservoir, the term Bre is difficult to altersignificantly through human action. However, for a systemof reservoirs,storing more water at locations where Bre is lower is one measure forreducing the system's evaporation. Becausemany factors cause reservoirlevels to rise and fall in the Nilüfer basin, especially the system'soperation,measuring evaporation from its reservoirs is complex, hard tovalidate, and subject towidespread, longstanding and vigorous debates.

A.2.8. Reservoir storageEach rth reservoir's water stock in the Nilüfer Basin is tracked for

the tth year. That year's water stock, Zrt, equals its stock in theprevious year, minus the net release (outflow minus inflow) from thereservoir, XLt, which contributes to flow at the downstream node inthat year. A second term subtracts the year's evaporation from itsreservoir's contents. The evaporation quantity, Xet, accounts for thefact that a reservoir's exposed surface area depends on its contents. A(0–1) vector of coefficients. BLr keeps track of each rth reservoir'slocation in the basin, assuring that streamflowgoing into it adds to thereservoir's contents, while outflows released from it reduce thereservoir's contents while also adding to streamflow in the down-stream reach. For the basin as a whole, the term BeL is an identitymatrix that conforms reservoir nodes with reservoir release nodes.Reservoir contents are:

Zrt = Zrt−1 −Xe

BeLXet −XL

BLrXLt ðA9aÞ

Contents of the rth reservoir in the initial period (0), Zr0, aredefined by beginning watershed conditions, Br0:

Zr0 = Br0: ðA9bÞ

The upper bound on each reservoir's water contents (storage) isdefined as:

Zrt b Krt ðA9cÞ

This equation guarantees that the rth reservoir's actualwater storagecontents in each tth period never exceed its capacity, defined above asKrt. If capacity is expanded, reservoir storage contents are set to a higherlevel. Policy makers often have an interest in considering an optimalreservoir storage capacity, different from the current capacity, so theactual storage capacity is specified as a choice variable instead of aparameter (fixed coefficient). Treating a reservoir's storage capacity assomething that can be changed has the potential of offering greaterflexibility to water policy makers. It could see growing use in basins oftheworld currently lacking adequate storage and/or facing the potentialconsequences of greater shortages brought on by climate change.

3 In the current model implementation for the Nilüfer Basin, pumping is constrainedto be zero.


A.3. Land use

Land use patterns affect the demand for water. For irrigatedagriculture, total land in production is expressed as:

Xc

Luct b RHSut : ðA10Þ

That is, irrigated land in production by node, crop, technology, andtime, summed over crops and technologies cannot exceed availableland (RHSut) by node and time period. In many dry rural irrigatedregions of the world, like the Nilüfer, water is often more limiting thanland. We used the maximum current irrigated land capacity for eachirrigation node as the upper limit on available land. However, moreacreage will likely become available if greater long term watersupplies can be secured and if institutions adjust to permit the extrawater to be used by agriculture.

A.4. Institutions

A.4.1. Environmental flowsA minimum annual water delivery to the Sea of Marmara is

specified by an environmental constraint in which saltwater intrusioninto the Basin's irrigated lands are avoided by keeping the flow intothe sea above a minimum level. That constraint is written as:

XMvt N XM0

vt ðA11Þ

where XvtM = annual deliveries to the Sea must exceed a minimum

amount shown on the right hand side.

A.4.2. Food securityFood security is required in the model by requiring that the total

irrigated land in production summed over all grain crops from all cropuse nodes exceeds a minimum amount of land. This constraint isspecified in the following form:

Xu

Xc

Luctzetv: ðA12Þ

This constraint assures that a guaranteed minimum amount of theBasin's irrigated land is set aside to assure food security by the basinsupplying a set amount of basic grains.

A.5. Economics

A.5.1. BenefitsIn the Nilüfer Basin, water decisions can produce both use-related

benefits and environmental benefits. Both are defined by the totalwillingness to pay by those who benefit from either kind of use.For agricultural uses, the willingness to pay is measured by thecontribution of water to net farm income. For urban nodes, it ismeasured by price per unit water times the number of units sold tothe customer (total water bill) plus any remaining consumer surplus.Consumer surplus is measured as the area beneath the demandfunction and above actual price charged. For environmental benefits,willingness to pay is measured as the maximum price that could becharged to visitors at the Basin's water-based recreation sites. In thecurrent implementation of the model, those sites are limited to thebasin's reservoirs. Important excluded environmental values includebenefits produced by instream flows at non-reservoir nodes as wellas any environmental values, such as option, existence, or bequestvalues influenced by variations in reservoir levels or by other waterdecisions.

A.5.1.1. Use-related benefits. Water uses produced by diversions inthe tth period and uth use, XBut, create economic benefits by beingapplied to the following quadratic total benefits function:

XBuut = Bou + B1uXut + B2uX2ut ðA13Þ

where Bou, B1u and B2u are parameters for the constant, linear andquadratic terms, respectively, for thebeneficialuseof surfaceflowateachof theu nodes,Xut.B2ub0 reflects a downward sloping demand schedule.For urban uses, a downward sloping water demand means that for anygiven period, greater per household use only occurs if the price per unitused falls. For agricultural uses, it means that for a given acreage, givencrop prices, given irrigation technology, increases in water applied peracre produce declining increments in net farm income. More details aredescribed on the computation of net farm income below.

For urban use nodes, incremental benefits from added use beginhigh as basic human requirements, such as drinking and sanitation,are met (i.e., B1u is large and positive). Incremental urban benefits fallrapidly as household water is applied to lower valued uses, such asoutdoor landscapes, in the face of greater supplies and lower prices(B2u large and negative). Similarly, displaced incremental urbanbenefits rise rapidly as droughts or other urban shortages becomemore severe and opportunities are reduced for painless conservation.

Total and marginal benefits for urban water uses are based onestimated price elasticities of demand for the City of Bursa. Increasedwater scarcity reduces water use, and increases the marginal value ofremaining water as urban water buyers move back along their waterdemand functions, substituting various water conserving measuresfor water.

The equation above describes use-related benefits per household forurban uses and benefits per acre for agricultural nodes. That equation isthen modified to account for future changes in population or irrigatedacreage. These alternative futures are accounted for by scaling theequationupbynumberof households for urbanuses in theNilüferBasin.

A.5.1.2. Environmental benefits. For this analysis, the economicbenefits of environmental quality are measured as the willingness topay for changes in quantities of water supplied for water-basedrecreation at the basin's six mainstem reservoirs. More details onmethods for measuring the value of water in recreation are describedin Ward et al. (1996) and in Ward and Beal (2000). For any givenreservoir node, those benefits are measured as:

XBert = Boe + B1eZrt + B2eZ2rt ðA14Þ

The B coefficients used for this analysis are based on the publishedwork by Gürlük (2006), with updates based on more recent visitationcounts from the year 2007. Generally, total gross recreation benefits atany reservoir increase up to a point where beaches and other facilitiesare flooded by water. The quadratic functional form was chosen toreflect the observation that further volume increases beyond thatpoint at which the quadratic function tops out will reduce visitationand total recreation benefits.

A.5.2. CostsIncreased stream diversions or depletions typically require addi-

tional costs to be incurred to make suitable for human use theincreased water used. For agricultural groundwater pumping nodes,the largest incremental costs are those incurred for energy and forrelated operation, and maintenance. For urban pumping nodes thereare also considerable additional costs for purification to make thewater safe and healthy for human consumption.

A.5.2.1. Use-related costs. Energy, operation, and maintenance costper acre foot pumped at a pumping node is defined as δpeu, whiletreatment costs at the same node is δteu per acre foot. Similar energy


and treatment costs are defined for surface diversion nodes as δdeu forenergy, operations, and maintenance, while δdtu is the equivalent costsurfacewater treated per acre foot. For urban uses, treatment costs areconsiderably higher than for agriculture, but urban treatment costsare typically lower for pumped water than for diverted river water.Based on these four kinds of costs, the total cost of water delivered toany agricultural or urban node is:

XCuut =Xp

δpeu + δptuh i

Xpt +Xd½δdeu + δdtu�Xdt : ðA15Þ

where Xpt is the number of acre feet pumped and Xdt is the number ofacre feet diverted in the tth period. Both of these quantities areunknowns whose values are determined by the model's solution. TheCity of Bursa has the only significant pumping in the Basin.

A.5.2.2. Environmental costs. Environmental costs are everywherebecoming an increasingly important factor in determining outcomesof major water policy decisions. Two possible concepts for definingenvironmental costs presented themselves (1) environmental oppor-tunity costs, equal to environmental benefits displaced by a decision,and (2) environmental operations costs, equal to the cost of additionalresources required to support the protection of greater environmentalbenefits at a given site. Either concept for measuring environmentalcosts would have worked, and in principle both should produce thesame optimized water policy. We chose the second alternative inorder that net environmental benefits could be measured at eachnode. These net environmental benefits are measured as environ-mental benefits minus added costs needed to assure a higher qualityenvironment. Environmental operations costs were measured as:

XCert = δerZrt ðA16Þ

where δer is the added cost of managing larger volumes of water at areservoir site, while Zrt is the acre feet of reservoir volume in period t.For the current analysis the term δer was measured as the additionalmonetary budget expense required to support additional visitorsattracted by larger quantities of water when those larger volumes canbe stored. That same term reflects the management costs saved whensmaller water volumes are stored. Measuring the added environ-mental cost of protecting and securing a higher quality environmentin this way is a considerable simplification of the marginal coststructure of environmental improvements. More generally, thatmarginal cost would include any operations costs needed to preserve,protect, or improve a natural environment. Considerable research isrequired on this important issue.

A.6. Net benefits

A.6.1. Use-related net benefitsUse-related net benefits are the following simple algebraic

subtraction of use-related costs from use-related benefits:

XNBuut = XBuut − XCuut ðA17Þ

With adequate streamflows produced by sufficient snowpack, use-related net benefits will be maximized by the model at each use nodeas well as its sum being maximized over all use nodes.

For irrigated agricultural nodes, net benefits are measured as netfarm water-related income:

Yuct = PucYielduc − Costuc½ �Luct ðA18Þ

That is, net farm income at the uth basin node for the cth crop inthe tth period equals net income per acre multiplied by the number ofacres. Annual net income per acre, Yuct, equals crop price, Puc, timescrop yield, Yielduc, minus total production costs, Costuc.

A.6.2. Environmental net benefitsEnvironmental net benefits are computed with a similar algebraic

subtraction:

XNBert = XBert − XCert ðA19Þ

When there are adequate starting reservoir volumes, theseenvironmental benefits will also be maximized by selecting reservoirvolumes at which environmental benefits exceed environmental costsby the largest amount. When volumes are low and/or when stream-flows are scarce each reservoir node's marginal environmental netbenefits will be nonzero.

A.7. Discounted net present value

Discounted net present value is expressed in its standard algebraicform:

XNPV =Xu

Xt

XNBuut

1 + ruð Þt +Xu

Xt

XNBeut1 + reð Þt ðA20Þ

That is, benefits from water uses and water environments aresummed together.

For irrigatedagriculture, discountednet present value is expressed totake account of alternative crops and alternative irrigation technologies.It is expressed as:

XNPVA =Xu

Xc

Xt

Yuct1 + ruð Þt ðA21Þ

That is, the present value of total water-related farm income for theNilüfer Basin sums income over nodes, crops, and time periods, whichdiscounts future incomesmore heavilywhen there is a higher discountrate.

References

Bateman, I.J., Carson, R.T., Day, B., Hanemann, W.M., Hanley, N., Hett, T., Jones, M.,Loomes, G., Mourato, S., Ozdemiroglu, E., Pearce, D.W., Sugden, R., Swanson, J., 2002.Economic Valuation with Stated Preference Techniques. Edward Elgar Publishing,Cheltenham, UK.

Bazzani, G.M., Di Pasquale, S., Gallerani, V., Viaggi, D., 2004. Irrigated agriculture in Italyand water regulation under the European UnionWater Framework Directive.WaterResource Research 40, W07S04. doi:10.1029/2003WR002201.

Bishop, R.C., 1978. Endangered species and uncertainty. The economics of a safeminimum standard. American Journal of Agricultural Economics 60, 10–18.

Booker, J.F., Young, R.A., 1994. Modeling intrastate and interstate markets for ColoradoRiver Water Resources. Journal of Environmental Economics and Management 26,66–87.

Boyle, K.J., Desvousges, W.H., Johnson, F.R., Dunford, R.W., Hudson, S.P., 1994. Aninvestigation of part — whole biases in contingent-valuation studies. Journal ofEnvironmental Economics and Management 27, 64–83.

Briscoe, J., 2005.Water as an economic good. In: Brouwer, R., Pearce, D. (Eds.), Cost–BenefitAnalysis and Water Resource Management. Edward Elgar Publishing, Northampton,MA, pp. 46–70.

Brouwer, R., 2004. The concept of environmental and resource cost. Lessons learnedfrom ECO2. In: Brouwer, R. (Ed.), Environmental and Resource Cost and the WaterFramework Directive: An Overview of European Practices. InHolland, Amsterdam.Riza Working Paper 2004 112x.

Bruinsma, J., Hrabovszky, J., Alexandratos, N., Petri, P., 2003. Crop production and inputrequirements in developing countries. European Review of Agricultural Economics10, 197–222.

Cai, X., McKinney, D.C., Lasdon, L.S., 2003. Integrated hydrologic–agronomic–economicmodel for river basin management. Journal of Water Resource Planning and Manage-ment 129, 4–17.

Cameron, T.A., 1992. Combining contingent valuation and travel cost data for thevaluation of non-market goods. Land Economics 68, 302–317.

Carey, J.M., Zilberman, D., 2002. A model of investment under uncertainty: modernirrigation technologyand emergingmarkets inwater. American Journal of AgriculturalEconomics 84, 171–183.

Carson, R.T., Flores, N.E., Hanemann, W.M., 1998. Sequencing and valuing public goods.Journal of Environmental Economics and Management 36, 314–323.

Ciriacy-Watrup, S.V., 1952. Resource Conservation Economics and Policies. University ofCalifornia Press, Berkeley, CA, USA.


Cooper, P., Poe, G.L., Bateman, I.J., 2004. The structure of motivation for contingentvalues: a case study of lake water quality improvement. Ecological Economics 50,69–82.

Crowards, T.M., 1998. Safe minimum standards: costs and opportunities. EcologicalEconomics 25, 303–314.

Çakmak, B., Uçar, Y., Aküzüm, T., 2006. Water resources management, problems andsolutions for Turkey. International Congress on River Basin Management, Ankara,Turkey.

Dai, T.W., Labadie, J.W., 2001. River basin network model for integrated water quantity/quality management. Journal of Water Resources Planning and Management-ASCE127, 295–305.

Fassio, A., Giupponi, C., Hiederer, R., Simota, C., 2005. A decision support tool for simulatingthe effects of alternative policies affecting water resources: an application at theEuropean scale. Journal of Hydrology 304, 462–476.

GAMS Development Corporation, 2007. GAMS Official Home Page. http://www.gams.com/.

FAO, 2006. Food security strategies. The Asian experience. FAO Agricultural Policyand Economic Development Series, vol. 3. http://www.fao.org/docrep/W7442E/W7442E00.htm.

Griffin, R.C., 2006. Water resource economics. The MIT Press, NY, USA.Görlach, B., Interwies, E., 2004. Assessing Environmental and Resource Costs in theWater

Framework Directive: The Case of Germany. Ecologic for the Federal EnvironmentalAgency, Berlin, Germany. Accessed December 2007. Available from http://www.ecologic.de/download/projekte/1950-1999/1970-01/erc_germany.pdf.

Greenley, D.A., Walsh, R.G., Young, R.A., 1981. Option value— empirical-evidence from acase-study of recreation and water-quality. Quarterly Journal of Economics 96,657–673.

Gürlük, S., 2006. The estimation of ecosystem services' value in the region of Misi RuralDevelopment Project: results from a contingent valuation survey. Forest Policy andEconomics 9, 209–218.

Gürlük, S., Rehber, E., 2006. Evaluation of an integratedwetlandmanagement plan: caseof Uluabat (Apollonia) Lake, Turkey. Wetlands 26, 258–264.

Haab, T.C., McConnell, K.E., 2002. Valuing Environmental and Natural Resources.Edward Elgar Publishing, Cheltenham, UK.

Hanley, N., Colombo, S., Tinch, D., Black, A., Aftab, A., 2006. Estimating the benefits ofwater quality improvements under the Water Framework Directive: are benefitstransferable? European Review of Agricultural Economics 33, 391–413.

Harmancioglu, N.B., Fedra, K., Barbaros, F., 2008. Analysis for sustainability in manage-ment ofwater scarce basins: the case of the Gediz River Basin inTurkey. Desalination226, 175–182.

Heinz, I., Velazquez, M.P., Lund, J.R., Andrev, J., 2007. Hydro-economic modeling in riverbasin management: implications and applications for the European Water Frame-work Directive. Water Resource Management 21, 1103–1125.

Intriligator, M.D., 1971. Mathematical Optimization and Economic Theory. Prentice HallPublishing, NJ, USA.

IPCC, 2007. Climate Change Information Sheet. www.ipcc.ch.Karaer, F., Küçükballı, A., 2006. Monitoring of water quality and assessment of organic

pollution load in theNilüfer Stream,Turkey. EnvironmentalMonitoringandAssessment,114, 391–417.

Krysanova, V., Hattermann, F., Wechsung, F., 2007. Implications of complexity anduncertainty for integrated modelling and impact assessment in river basins. Environ-mental Modelling & Software 22, 701–709.

Lien, G., Hardaker, J.B., 2001. Whole-farm planning under uncertainty: impacts ofsubsidy scheme and utility function on portfolio choice in Norwegian agriculture.European Review of Agricultural Economics 28, 17–36.

Lund, J.R., Ferreira,M.,1996. Operating rule optimization for theMissouri River ReservoirSystem. Journal of Water Resources Planning and Management 122, 287–295.

Michelsen, A.M., Taylor, R.G., Huffaker, R., McGuckin, T., 1999. Emerging agriculturalwater conservationprice incentives. Journal of Agricultural and Resource Economics24, 222–238.

Nicholson, W., 1998. Microeconomic Theory. The Dryden Press, TX, USA.Önder, S., Önder, D., 2006. Evaluation of water resources on the basis of river basins and

the probable changes to occur in basin management in the future due to globalclimate change. International Congress on River BasinManagement, Ankara, Turkey.

Pearce, D.W., Ulph, D.,1999. A social discount rate for the United Kingdom. In: Pearce, D.W.(Ed.), Economics and Environment: Essays on Ecological Economics and SustainableDevelopment. Edward Elgar Press, Celtenham, UK.

Pulido-Velazquez,M.,Andreu, J., Sahuquillo, A., 2006. Economicoptimizationof conjunctiveuse of surface and groundwater at the basin scale. Journal of Water Resoures Planningand Management 122, 454–467.

REC, 2002. Turkey's environment. The Regional Environmental Center for Central andEastern Europe. Working Report, Szentendre, Hungary.

Rowlatt, P., Spackman, M., Jones, S., Jones Lee, M., Loomes, G., 1998. Valuation of deathsfrom air pollution. National Economic Research Associates (NERA) Report, London,UK.

Silberberg, E., Suen, W., 2001. The Structure of Economics: A Mathematical Analysis.McGraw-Hill Press, NY, USA.

Tilmant, A., Pinte, D., Goor, Q., 2008. Assessing marginal water values in multipurposemultireservoir systems via stochastic programming. Water Resources Research 44.

Turpie, J.K., Heydenrych, B.J., Lamberth, S.J., 2003. Economic value of terrestrial andmarine biodiversity in the Cape Floristic Region: implications for defining effectiveand socially optimal conservation strategies. Biological Conservation 112, 233–251.

VanAst, J.A., Boot, S.P., 2003. Participation in Europeanwater policy. Physics and Chemistryof the Earth 28, 555–562.

Van der Veeren, R.J.H.M., Lorenz, C.M., 2002. Integrated economic–ecological analysisand evaluation of management strategies on nutrient abatement in the Rhine basin.Journal of Environmental Management 66, 361–376.

Vreke, J., 1990. Modelling agriculture to support regional water management. EuropeanReview of Agricultural Economics 17, 317–334.

Ward, F.A., 2005. Environmental and Natural Resource Economics. Prentice-Hall Publish-ing, NJ, USA.

Ward, F.A., Beal, D.J., 2000. Valuing Nature with Travel Cost Models. Edward Elgar,Cheltenham, UK.

Ward, F.A., Lynch, T.P., 1996. Integrated river basin optimization: modeling economicand hydrologic interdepence. Water Resources Bulletin 32, 1127–1137.

Ward, F.A., Velazquez, M.P., 2008. Efficiency, equity, and sustainability in awater quantity–quality optimization model in the Rio Grande Basin. Ecological Economics 66 (1),23–37.

Ward, F.A., Roach, B.A., Henderson, J.E., 1996. The economic value of water in recreation:Evidence from the California drought.Water Resources Research 32 (4), 1075–1081.

Ward, F.A., Young, R.A., Lacewell, R., King, J.P., Fraiser, M., McGuckin, J.T., DuMars, C.,Booker, C., Ellis, J.F., Srinivasan, R., 2001. Institutional Adjustments for Coping withProlonged and Severe Droughts in the Rio Grande Basin. (Accessed November 30,2007). Available from wrri.nmsu.edu/publish/techrpt/tr317/.

Xevi, E., Khan, S., 2005. A multi-objective optimization approach to water management.Journal of Environmental Management 77, 269–277.

Young, R.A., 2005. Economic criteria for water allocation and valuation. In: Brouwer, R.,Pearce, D. (Eds.), Cost–Benefit Analysis andWater Resources Management. EdwardElgar Publishing Limited, MA, USA, pp. 13–45.

Internet resources

www.dsi.gov.tr/english.www.buski.gov.tr.www.bursatarim.gov.tr.www.cevreorman.gov.tr.


analysis integrated basin management: water and food ... · this paper presents a basin scale...

Documents