Simulation of Water Supply and Demand in the Aral Sea Region

8/2/2019 Simulation of Water Supply and Demand in the Aral Sea Region

1/13

Simulation of Water Supply and Demand inthe Aral Sea Region

by P . Raskin, E. Hansen, and Z. Zhu, M. IWRA

Stockholm Environment Institute-

Boston Center89 Broad Street, 14th Floor

BOSTON M-A 021 IOU.S.A.andD. Stavisky,

Institute of Geography, Academy of Sciences, Moscow, Russia(currently Stockholm Environment Institute - Boston Center)

ABSTRACT

The Aral Sea, a huge saline lake located in the arid south-central region of the former U.S.S.R., isvanishing because the inflows from its two feed rivers, the Amudarya and Syrdar'ya, have diminishedradically over the past three decades. The loss of river flow is the result of massive increases in riverwithdrawals, primarily for cotton irrigation in the basins. A microcomputer model, the Water Evaluationand Planning System (WEAP), has been developed for simulating current water balances and evaluatingwater management strategies in the Aral Sea region. WEAP treats water demand and supply issuesin a comprehensive and integrated fashion. The scenario approach allows flexible representation of theconsequences of alternative development patterns and supply dynamics. For the Aral regions complexwater systems, a detailed water demand and supply simulation was performed for the 1987-2020

period, assuming that the current practices continue. The analysis provides a picture of an unfolding

and deepening crisis. Policy scenarios incorporating remedial actions will be reported in a separatepaper

I N T R O D U C T I O N

The Aral Sea, a saline lake located in the aridsouth-central region of the former U.S.S.R. is vanish-ing (Fig. 1). Once the fourth largest lake in the worldby area, the Aral Sea today is nearing half of its

surface area in 1960, less than one-third its previoussize by volume. If current patterns continue, the lakewill diminish to several residual lifeless brine lakesnext century.

The Aral is shrinking because the flows from itstwo feed rivers, the Amudarya and Syrdarya, havedecreased from over 50 km3 per year thirty years agoto a mere trickle. The loss of river flow is the resultof massive increases in river withdrawals, primarilyfor irrigation, along the river basins. The two riversbegin at the Pamir and Tianshan plateaus, plunge

downward into the desert of the Central Asian re-publics and terminate at the Aral Sea. Since the 1960san immense system of dams and reservoirs has beendeveloped in the region. Today, the Aral basin is anastonishingly complex web of canals, impoundments,

irrigation fields, and water engineering facilities. Thewaters in the two rivers are the lifeblood of the

agricultural economies in five Central Asian republicsof the former U.S.S.R.: Turkmen, Uzbek, Tadzhik,Kirgiz, and Kazakh, supporting 7.6 million hectaresof irrigated crops. The current patterns of water use

and the recession of the lake has generated multipleenvironmental and economic problems [l-5]. Thescale of these problems is substantial, covering anarea of 3.5 million km* and affecting some 35 millioninhabitants in the five republics. There is an inter-national consensus that the situation is not ecologi-cally sustainable and comprehensive strategies foraltering water development patterns are needed.

Beyond the deterioration of the lake and the lossof its fishing industry, there are other serious impacts.For example, the recession of the sea has created a

huge area

-

about 30,000 km2 -

of salt on the for-mer lake bed, Toxic to humans and deleterious tocrops, the salt is whipped up by winds and carriedover wide areas. The ecology of the river deltas hasbeen seriously degraded as the surrounding water

0250-8060/92/$3.30 Water International, 17 (1992) 55-67 0 IWRA/Printed in the U.S.A.


2/13

urkmcn S.S.R.

Figure 1. A map of the Aral Sea Region.

table falls along with the sea, and river flow dimin-ishes. In addition, regional climate may be changingas the modulating influence of the Aral diminishes

with its size, with summers and winters apparentlybecoming more severe [2]. Shorter growing seasons,compounded by soil salinization and salt storm de-posits, would expand water shortages by further in-creasing the requirements for water. Last, but notleast, there is great concern that deteriorating waterquality will lead to a deepening public health crisis.

Regional climate may be

changing as the modulatinginfluence of the Aral

diminishes

A microcomputer model, the Water Evaluation andPlanning System (WEAP), was developed for evalu-ating alternative water development policy options incomplex systems such as the Aral Sea region [6].Employing the scenario approach, the WEAP model

provides a structured approach to integrated waterdemand-supply analysis.

This paper presents results of a business-as-usualsimulation of the regions water supply for the 1987-

2020 period, assuming that the current practices

continue. Development and evaluation of alternativewater policy scenarios will be reported in futurepapers. In this paper, we focus on illustrating the

magnitude of the problem and the challenge fordevising sustainable water strategies for the Aral re-gion.

C UR R ENT WATER DEMAND AND

S U P P L Y

Comprising lowland deserts and mountains, the

Aral region has a climate characterized by high evapo-transpiration and severely arid conditions. Annual

precipitation is less than 100 mm in the southwestdeserts and about 200 mm approaching the foothillsof the southeastern mountains. However, the regionhas favorable thermal conditions for the growth ofcotton and other heat-loving crops: the average noon-time temperature during growing seasons (May-Sep-tember) reaches 20-4X and the average daily tem-perature in July is 35C [7]. Although thin andinfertile, soil in the region is easily tilled and pro-ductive for certain crops with the application ofsupplementary water. These favorable conditions haveprovided the natural base for intensive irrigated ag-ricultural development, particularly the large scaleproduction of cotton in the Aral region.

The Amudarya and Syrdarya basins have some30 primary tributaries (Figs. 2 and 3). More than 20

56 Water International


3/13

ZerafshanIScheme ofthe

Amudarya Basin

KKAU-Tahlatash

TuyamuyunResv

Afganlstan

Pyandz

C - - - I -

- - local riierA

or tributsly- reservoir

r) - river node-L

1 -Horm-Tash-Saka 7 -Tash-Klich-Bay2 -Tash-Tash-Saka 8 -KKAR-Klick-Bay

n groundwater 3 -- Horm-Urgench-Arna 9 -Tash-Klich-Bozsu6 4 5

- Horm-K ch-Bay Harm-Okty-Ama HormPh-Ama 12 10 11 -0 - distribution system -- - Tash-Sovet-Yab KKAR-Kch-Bozsu TashDgumbay-Saka

Figure 2. Scheme of the Amudarya Basin.

One of the most complicated

human water development

systems in the world

large and middle sized reservoirs and 60 canals of

different sizes have been constructed in the two basinssince the 1950s [8-l0]. The Karakum canal, con-structed in 1950s as a centerpiece of Soviet plans toexpand cotton production, diverts water from Amu-darya with a maximum flow of 320 m3 per secondover 840 kilometers to the vast Karakum desert. Inaddition, approximately ten per cent of supplies arefrom groundwater sources. The regions water systemis one of the most complicated human water devel-opment systems in the world.

In designing the-schematic representation of thetwo basins, we have aimed for as much detail aspossible in characterizing both demand and supplysources, subject to the availability of field data. Re-ferring to Figs. 2 and 3, the representations consistof the following main elements:

l Distribution Systems A distribution system rep-resents water users in a common geographic areawith shared water sources. In the current repre-sentation, distribution systems are identified with

irrigation systems that are used for allocatingwater in the Aral ,region. There are 23 distribution

systems identified for Amudarya, and 6 for Syr-

darya. Irrigation systems at the lower Amudarya

area are further separated into twelve districts(indicated by the naming convention adminis-trative district/irrigation system, e.g., Horezm/Tash-Saka). Water demand in each distribution

system is subdivided by major sectors: irrigation(further partitioned by crop type and irrigation

technique), industry (by type), municipal (by ur-ban and rural), fishery, and livestock.

*Main River and Tributaries These are the pri-mary water conduits in the region. Stream flowsare estimated along every tributary and the main

rivers on a monthly basis. Account is taken ofinflows, outflows, withdrawals, evaporative losses,and groundwater interactions. There are five typesof river nodes: reservoir node, withdrawal node,diversion node, confluence node and tributary node.

Vol. 17, No. 2 (1992) 57


4/13

Scheme of theSyrcfarya Basin High Narin

ARTUR 5 CHAKIR

Low Syr

ChardaraResV

Farhad

Right Joint

Mld Syr- - local riverA

or tributary

-restwoir

;: riier nodegroundwater0 -distribution sys t em

Figure 3. Scheme of the Syrdarya Basin.

Each is simulated according to its operating rule.For instance, WEAPs reservoir operating ruletakes into account a reservoirs inflow, storage

capacity, surface evaporation, withdrawal at thereservoir, hydroelectric generation, and down-

stream release requirements. In-stream flow re-quirements for maintaining, for example, envi-

ronmental quality also may be specified.

l Local Supply Sources These include run-of-riverpumping stations, groundwater aquifers, rainwatercollection, and reservoirs on rivers that are hy-drologically independent of the main river system.In WEAP, withdrawal demands are met by local

sources with residual requirements assigned to anyriver linkages.

9 Links between Distribution Systems and SupplySources Transmission links between demand sitesand supply sources are identified in the system

representation. Each distribution system may be

supplied by a maximum of twelve sources withlinks to ten local sources, one tributary nodeand one main river node. Capacity constraintsand conduit losses are taken into account.

Karasu Right

ShaydansayF

Narin Gnte

1987 Water Demand

Water accounts have been estimated for the year1987, the most recent year for which comprehensivedata is available. Water demands for that year aresummarized in Table 1, broken down by each sectorfor each distribution system. The total water demandfor the Aral region is 97.32 km3. Of this total, 53.55km3 is demanded from the Amudarya basin, and43.77 km3 from the Syrdarya basin. Water demandsare dominated by the agriculture sector, accountingfor 82 per cent of the total demand.

The regions irrigated areas by type of crop aresummarized in Table 2 [11]. The total irrigated areaof the region in 1987 was 7.6 million hectares, with4.3 million hectares located in the Amudarya basin,and 3.3 million hectares in the Syrdarya basin. Waterdemand shares by crop types in the two basins arepresented in Fig. 4. Cotton is the major crop, ac-counting for 51 per cent of the agricultural water

demand in Amudarya, and 34 per cent in Syrdarya.The Soviet Union has been the second largest cottonproducer in the world, producing over 90 per cent ofits fiber in the Aral region. Clearly, strategies for



5/13

Table 1. 1987 Water Demand of the Aral Region (Unit: km3).DistributionSystem

Amudarya Basin

Agriculture Industry Municipal Livestock Fishery Total

Pyandz 1.11 0.05 0.22Vahsh 2.48 0.08 0.30Kafirnigan 1.40 0.03 0.25Surh-Sherabad 3.19 0.00 0.27Afghanistan 0.00 0.00Karakum 7.22 1.69

Kashkadraya 5.56 0.05Bukhara-Zerafshan 8.89 0.66Cardzou 2.39 0.17Horezm 2.60 0.05Tashaus 2.59 0.00KKAR 7.37 0.20

0.000.41

0.400.650.160.150.050.12

0.02 0.02 1.420.03 0.19 3.080.01 0.45 2.140.01 0.03 3.500.00 0.00 0.000.02 0.07 9.41

0.01 0.05 6.070.11 0.04 10.350.00 0.09 2.810.00 0.05 2.850.00 0.05 2.690.03 1.51 9.23

Amu Total 44.80 2.98 2.98 0.24 2.55 53.55Percentage 84 6 6 0 5 100

Syrdarya Basin

High Narin 2.11 0.01 0.03 0.00 0.07 2.22Fergana Valley 12.48 0.31 1.31 0.00 0.07 14.17Middle Syrdarya 7.45 2.14 0 .33 0.00 0.12 10.04

CHAKIR 5.17 2.30 1.39 0.00 0.12 8.98ARTUR 2.11 0.21 0.15 0.00 0.12 2.59Lower Syrdarya 5.26 0.07 0.11 0.00 0.33 5.77

Syr Total 34.58 5.04Percentage 79 12

Aral Total 79.38 8.02Percentage 82 8

Note: percentage figures may not total correctly, due to rounding.

3.32 0.00 0.83 43.778 0 2 100

6.30 0.24 3.38 97.326 0 3 100

Table 2. 1987 Irrigation Areas (Unit: 1,000 hectares).

Cotton Rice Wheat Maize Cereals Potato Veg. Melon Fodder Vineyd. Orchard Total

Amudarya Basin

PyandzVahshKafimiganSurh-SherabadKashkadrayaBukhara-ZerafshanCardzou

KarakumHorezmTashausKKAR

55.0123.261.6

199.3310.5443.9175.4

507.2120.6208.7237.7

Amu Total 2443.1 190.0 157.9 116.4 45.7 28.9 105.6Percentage 57 4 4 3 1 1 2

4.:1:7E20:1::

29:00.0

125.7

1.9 1.94.2 4.22.1 2.1

12.328.0 E18.1 1I3:o17.5 11.1

50.5 32.1

2;; 2.40:9 13.214.9

1.94.22.16.1

18.111.9

EE0:9

i.392:12.94.15.1

::0:9:::

3.78.34.17.4

13.717.7

7.8

22.44.4

96::

1.7 29.6

::: 66.333.2;;

74.4

10:9 129.7184.211.4

32.9 E4.9 55:8

13.5 0.011.2 173.6

103.8 746.72 17

I:.;5:915.730.937.1

8.7

25.31.3

10.41.2

11.7 116.026.3 260.013.1 130.024.2 361.032.1 585.044.0 811.0

8.1 241.0

23.4 697.0

99.:230.5

13:8 286.8588.6153.6 215.4 4307.0

4 5 100

Syrdarya Basin

High Nat-inFergana ValleyMiddle SyrdaryaCHAKIRARTUR

Lower SyrdaryaSvr Total

82.0 2.3 2.8 2.8787.6 10.9 3:: 40.3 23.2288.5 54.4 17:5 23.5 11.8142.9 49.3 10.8 15.9 5.328.9 32.9 5.6 5.6 0.0

78.3 88.9 26.9 26.9 0.01408.4 238.6 95.3 115.0 43.0

43 7 3 3 1

2.8 2.6 44.1 17.511.2 3:.;11:o 12.6 272.8 3:; 103.94.0 10.7 207.4 2017 30.96.3 16.2 162.3 14.3 32.41.3 2.9 ::: 79.6 4.3 8.62.8 5.5 7.3 186.2 7.4 14.9

28.4 73.7 43.7 952.5 93.3 208.11 2 1 29 3 6

173.11365.6680.4462.7173.1

445.13300.0

100

Aral Total 3851.4 428.6 253.2 231.3 88.7 57.2 179.4 147.5 1699.2 246.9 423.5 7607.0Percentage 51 6 3 3 1 1 2 2 22 3 6 100

Vol. 17, No. 2 (1992) 59


6/13

Amudarya Basin (44.8 km3 in total)

Syrdarya Basin (34.6 km~3 in total)Vineyard (1 9%)

Vegetables (2.3%)

Othe r Cerea ls (1%)

Rice (16.6%)

Figure 4. 1987 agricultural water demand shares.

rectifying the water situation in the Aral region arecoupled to strategies for cotton: how much, whattype, what technologies? Fodder crops account forthe second largest requirement, at 29 per cent and19 per cent of agricultural water demands in theAmudarya and Syrdarya basins, respectively, It isalso notable that water-intensive rice production ac-counts for 19 per cent and 12 per cent, respectively,of agriculture water demands. The demands for ag-riculture are built up at the distribution level bymultiplying irrigation areas by water application rates[12]. Estimated on-farm water application rates areincluded in Table 3. These figures are comparable to

U.S. rates. In Arizona, where the climatic conditionsare similar to the Aral region, the on-farm annualwater application rates are of the same order ofmagnitude: 14,000 m3 for cotton, 9,000 m3 for corn,and 12,000 m3 for potatoes [13].

Water demands for industry (Table 1) are far lessthan for agriculture, approximately 6 per cent inAmudarya and 12 per cent in Syrdarya. Dependingon economic development strategies in the future,

Water demands for industry. . . are far less than for

agriculture

industrial demands may become more significant withtime. Industrial demands are built up at the distri-bution system level from estimates of productionoutput and water use rates. Industrial water demandsare currently dominated by the electric power sector.

Municipal water demands comprise about 6 per

cent of total demand in the Aral region, as estimatedfrom population and water use data at the adminis-trative district level and allocated to distribution

systems. The final two water demand sectors areLivestock and Fishery. As reported in Table 1, knownwater demands for livestock are quite small, whilefisheries account for some 3 per cent of overall waterdemands.

These water demands discussed above are for finaluse. They represent the water required by the finaluser for crop growth, industrial processes, domesticconsumption, and so on. To convert these final de-mands to the actual water supply requirements, WEAP

allows for three adjustments to water demands. Thefirst adjustment takes into account the distributionlosses in each distribution system. For an irrigationsystem, a considerable amount of water delivered tothe field will not be used by the crop root zone dueto field evaporation and deep percolation. The secondadjustment accounts for water recycling or reuse. Thisrefers to processes by which water may be used inmore than one application before discharge. For ex-ample, irrigation water may be routed for reuse inmore than one field. The effect of recycling is to

reduce the water required from primary water sources.The third adjustment is for water transmission loss.This refers to the evaporative and infiltration lossesof water in the canals and conduits carrying the waterto a distribution system. Unfortunately, at this stage,our data are insufficient to distinguish the distributionlosses from transmission losses, and these two factorsare combined in the current estimates. The totalwithdrawal requirements in the two basins in 1987were estimated as 127 km3 (70 km3 for Amudaryaand 57 km3 for Syrdarya), or 130 per cent of theestimated final demand.

198 7 Water Supply

Major surface and groundwater sources are iden-tified in Figs. 2 and 3. In WEAP, surface water istracked from the flows entering the system throughvarious river nodes. Stream virgin flow data of 1987,which was a wet year for the Aral region, is collectedin Table 4 [14]. The total surface water resources in

the region comprised 132 km3, of which 84 km3 werefrom the Amudarya basin, and 48 km3 from theSyrdarya basin. The 1987 virgin flow figures ofAmudarya and Syrdarya are equivalent to four timesand 2.3 times, respectively, the average virgin flow ofthe Colorado River. On average, the annual surface



7/13

Table 3. 1987 On-Farm Water Application Rates (Unit: m3/ha/yr).

Cotton RiceOther

CerealsPotato &Vegetables Melons Fodder Vineyard Orchards

Amudarya Basin

Pyandz 8700 249 00 7200 11600Vahsh 8700 24900 7200 11600Kafimigan 9900 26800 8200 13200Surh-Sherabad 8200 27900 7000 10500Kashkadraya-Karshi 9100 30800 8300 11700Bukhara-Zerafshan 10100 32400 9100 12800

Cardzou 10100 32400 9100 12800Karakum 10600 33400 9200 13500Horezm 8300 29200 7900 10500Tashaus 8300 29200 7900 10500KKAR 7500 28000 7600 9600

56004900620058007000

74007500620062005800

11400 7330 853011400 7330 853012900 8230 963010300 7610 851011300 8070 907012500 8960 9960

12500 8960 996013300 9590 1079010300 7330 823010300 7330 82309500 6540 7440

Syrdarya Basin

High Narin 7400 22700 7700 9900 5500 9700 6300 8400Fergana Valley 8500 248 00 7700 11400 6000 11200 7100 8400Middle Syrdarya 8700 29700 8100 11000 6800 10800 7770 8670CHAKIR 8500 24800 7700 11400 6000 11200 7100 8400ARTUR 7400 27900 7200 9500 5700 9400 6640 7440Lower Syrdarya 7500 26700 7400 8900 5600 8800 6240 6940

Table 4. 1987 Surface Water Sources of the Aral Region (Unit: million m3).Jan. Feb. Mar. Apr. May June July Aug. Sent. Oct. Nov. Dec. Year

Syrdarya Basin

Toktogal res. 313 290 375 467 1457 2514 3134 1861 832 611 500 442 12797Karasu left 16 12 13 18 48 65 83 70 49 40 31 24 470Karasu right 45 54 72 256 415 335 213 123 83 76 76 60 1809

Shaydansay 2 27 17

2216 13 11 5 5

65

112Karadarya trib. 191 215 554 1325 2086 2190 1814 1039 468 446 522 436 11284Kassansay res. 7 5 5 15 69 77 53 28 12 10 11 9 301Abshirsay : 4 14 67 75 51 27 11 12 8 28 4KurvasayIsfayramsay 29 2;

:, 2: 0 0 0 0 0 0 0 : 027 :; 119 201 153 78 59 46 869

Shahimardan 13 10 11 10 2: 67 54 34 27::

21 335Isfarasfara I1 8 24 123 137 52 24 18 16 49 0sokh 32

2: 2:75 194 375 412 171 75 43 1512

Right tributaries 13 12 22;;

207 220 167 152 93 41:;:

30 1077Aksu total 16 16 15 17 20 34 52 32 20 24 19 15 281Kattasay 0 0 0 0 0 0 0 0 0 0 0 0Sanzar 3 2 5 16 16 7 : 3 : 3 5 72Shirinsay 3 3 4 3 3 6 4 4

:48

CHAKIR total 266 259 446 1098 2160261::

2312 164681;

519 424 377 12941

Aris and Bugun 172 136 208 449 391 361 309 210 125 107 96 95 2658Lower Syrdarya 56 41 89 145 61 16 15 10 7 9 6 8 46 4

Syr Total 1193 1120 1889 3997 7197 8949 8993 5972 2863 2091 1898 1641 47803

Amudarya Basin jVahshPyandzKunduzKafimigan trib.Surhan and SherabadMurgabTedjenArtek

KashkadrayaGuzadaryaZerafshan

4 5 8 41 41079 975

122 127166 17390 9375 60

0 0

:: 3:3 2

118 95

5972036

149561321137

:;160

17178

11432929

27 299869123113757

26567237

21373884

350141399618227

23:33

42 6

36035780

63814631010

1409

25:29

984

44466803

6161109610

59

9191

161298

4178 2136 11175544 3525 2119

513 290 193693 349 339384 168 159

40 65 880 0 0

23 0 22

103 55 386 6 9

1158 560 287

7981674

140266137

88

2:35

21;

6371253

12521310099

029

307

163

2166537602

3535774147591264

19121 2

1431203

5716

Amu Total 2157 1983 4195 7027 9692 13908 15152 12643 7153 4372 3380 2658 84320

Aral Total 3350 3103 6084 11024 16889 22857 24145 18615 10016 6463 5278 4299 132123

Vol. 17, No. 2 (1992) 61


8/13

water resources of the Aral region account for some120 km3[15].

As seen in Table 5, the regions groundwater with-drawal in 1987 accounted for 12.3 km3, 4 km3 in theAmudarya basin and 8.3 km3 in the Syrdarya basin.Evaluating the role of ground water in the regionswater budget is complex and, due to limited data,detailed physical interactions between surface waterand groundwater are not included at the current stage.

Groundwater patterns need more clarification in fu-ture analysis.

Water losses on river sections from evaporationand infiltration and returned water from demand sitesare taken into account in WEAP. Reservoir waterstorage and release are simulated by user-definedoperating rules. Characteristics of the main reservoirsin the region are summarized in Table 6 [16-18].

P R O J E C I I O N S

An important concept of WEAP is the distinctionbetween a business-as-usual scenario and alterna-tive policy scenarios. The business-as-usual scenarioincorporates currently identifiable trends in economicand demographic development, water supply avail-ability, water use efficiency, water pricing policy, andother aspects. No new water conservation measuresor supply projects are included in the business-as-usual scenario. The business-as-usual analysis pro-vides a reference against which the effects of alter-

native policy scenarios may be assessed.

Water Demand Projection Hy drological Fluctuat ions

In the past three decades there have been tremen-dous efforts in water demand projections [19-22]. In

Table 5. 1987 Groundwater Sources of the Aral Region (Unit:million m3).Syrdarya Basin

Hydrological fluctuation patterns are important inestimating future water availability. WEAP is designedto incorporate historic fluctuations to represent futurepatterns. However, time series data for many elementsof the Aral basin are not available. River flows havebeen altered with the extensive irrigation developmentand many hydrological records cannot serve as proxies

for historic hydrological patterns. Therefore, whileWEAP is designed to utilize historic time series datafor the general cases, a second, simpler option hasalso been built into the model for the Aral Sea case.

High Narin 1000Fergana Valley 4800

Middle Syrdarya 1000CHAKIR 1000ARTUR 25 0Lower Syrdarya 25 0Syr Total 8300

Amudarya Basin

Pyandz 173Vahsh 275Kafimigan 459Surhandarya 416Kashkadarya & Karshi 29 9Zerafshan & Buhara 1030

Cardzou 41 4Karakum 591Lower Amudarya 343

Amu Total 4000

Aral Total 12300

general, water demand forecasting approaches fallinto four broad categories, each with advantages andlimitations: time extrapolation, single coefficientmethods, multiple coefficient methods, and probabi-listic analysis.

WEAP provides a flexible and detailed structurefor water demand forecasting. It is designed to allowthe inclusion of a full array of possible demand-sidemeasures. A multiple-level structure is used in WEAP

to manage demand data: Sector, Subsector, End-use,Device, and Use-rate. For example, under the agri-culture sector, irrigation areas for each crop are definedat the Subsectorlevel; fractions of irrigation area ineach subregion are measured at the End-use level;irrigation techniques used in each subregion are iden-tified at the Dev ice level; and water use rates aredefined at the bottom level. At each level, activitiescan be driven by user-specified development targets.

The full complexity of the WEAP demand fore-casting structure is being used to develop a range ofpolicy scenarios for the Aral region. However, therapidly changing political and economic situation inthese Central Asian republics - and limited sourcesof credible data - hamper our exercises. In this paperour task is more straightforward: to introduce thecurrent water accounts and a business-as-usualreference projection based on the continuation ofcurrent patterns. For the latter purpose, we rely pri-marily on population growth as demand driving vari-able. These results provide a benchmark for the more

complex policy-oriented demand scenarios.

In the simpler method, five categorizes of water-type years, Very Wet, Wet, Normal, Dry, and Very

Dry , are used to represent hydrological patterns. Thesefive water-type years correspond to different hydro-

Many hydrological records

cannot serve as proxies forhistoric hydrological patterns



9/13

Table 6. Characteristics of Selected Reservoirs.

Year ofReservoir River Basin Construction

Aumdarya Basin

Tuyamuyun Amu-darya 1985Nurek Vahsh 1975Kattakurgan Zerafshan 1952/1968South-Surhan Surhan-dar ya 1964Chimkurgan Kashka-darya 1963Pachkamar Guza-darya 1968

Syrdarya Basin

Toktogul Narin 1974Chardara Syr-Dar ya 1965Kayrakkum Syr-Darya 1956Andigan Kara-Darya 1980Charvak Chirchik 1970Ahangaran Ahangaran 1974Tuyabuguz Ahangaran 1960CHAKIR rsv. CHAKIRBugun Bugun 1970Kassansay Kassansay 1956Karkidon Kurvasay 1963Dgizak Sanzar 1967Kattasay Kattasay 1965Nayman Kirgizata 1966

MaximumSurfaceArea

(km*)

65 09884.564.649.214.2

28 4 90 0

5135940.3

8.120.769.163.511

9.512.52.93.2

MaximumStorage(lo6 m)

723010500

90 080050026 0

195005700403017901990

18026 0

243037027 02l8

90

::.5

Dead EvaporationVolume Rate( lo6m) (mm/year)

2390 2 0 0 06000 1000

60 200024 0 2000

50 200010 2000

5500 10001000 20001480 2000

150 2000300 2000

10 20002000::20002;: 20002000

20002000i0 20001.5 2000

logical occurrence probabilities in conventional fre-

quency analyses. The frequency analysis of an annualinflow record at a representative river point providesa sequence of water-type years. This sequence may

then be adjusted to explore alternative assumptionsn future hydrological patterns. From the monthly

inflow record at the selected river point, averagemonthly inflows for each water-type year are calcu-lated and the ratios of monthly fluctuations for thefour nonnormal years to the normal year are then

computed. For every supply source, the base year (thefirst year in the planning period) monthly inflows are

input as data, while values for the future year monthlyinflowsare set by the water-type sequence by applyingappropriate monthly fluctuation coefficients to the

base year inflows.

In this study, monthly inflow data of 1950-1982 atthe Tupolang (on Amudarya River) and the Narin

(on Syrdarya River) gauging stations were used inestimating the two basins water-type sequences dur-ing the 19882020 period. Through frequency anal-

yses, the five water-types, Verywet,Wet, Normal,Dry, and Very Dry defined in this study, correspond,respectively, to occurrence probabilities of 0- 1 O%, 1 O-

30%,30-75%, 75-95%, and 95-100%. Because manysmaller tributaries dont have time series data, we canonly assume that the two defined sequences are rea-

sonable approximations for the entire Amudarya andSyrdarya basins. Though this method assumes hy-drological homogeneity across each of the two basins,it reduces the requirements for historical data whilepermitting explorations of future water patterns that

Vol. 17, No. 2 (1992) 63

deviate from historical patterns due, for example, toclimate alternations.

Simulation Results

Like other streamflow simulation models, the prin-ciple of mass balance guides the water flows throughthe system in WEAP [23-26]. At each river node, theincoming water is balanced by the outgoing waterplus the retained water at the node. Outgoing wateris the water diverted, either for demands or otherpurposes, plus the flow conveyed downstream. Be-

tween nodes, evaporation from the stream surface,interaction with groundwater aquifer, and return flowsfrom distribution systems affect the water balance.Each system element, such as a reservoir, has a defined

governing rule in passing, releasing, and allocatingwater. Unlike these models, however, WEAP addressesboth the supply and demand issues in an integratedfashion. Demands drive the water allocations amongsupply sources and demand sites. Detailed demandmanagement strategies as well as the full range ofsupply development options are incorporated in the

model. WEAP provides optional water allocationschemes, one based on priorities and another basedon equitable allocation, and flexible reports in varioustabular and graphic forms [6].

Table 7 presents the annual average water demandcoverage

-

the ratio of supply available to demand-

at each demand site in selected future years. When

the coverage value is one, the demand is fully supplied;otherwise, only the indicated portion of the demand


10/13

Table 7. Projected Demand Coverage in Selected Years.

Amudarya Basin

PyandzVahshKafimiganSurh-SherabadAfghanistanKarakumKashkadraya

Bukhara-ZerafshanCardzouLower Amudarya

Syrdarya Basin

High NarinFergana ValleyMiddle SyrdaryaCHAKIRARTURLower Syrdarya

1987 1995 2000

Very Wet Very Dry Normal

1.00 1.00 1.001.00 0.97 1.001.00 0.84 0.961.00 1.00 1.001.00 1.00 1.001.00 0.94 1.001.00 0.81 1.00

1.00 0.52 0.951.00 0.31 0.960.99 0.11 0.62

Wet Very Dry Normal

1.00 1.00 1.001.00 0.88 1.001.00 0.76 0.880.98 0.76 0.941.00 0.81 0.990.77 0.30 0.44

.

2010 2020

Normal Normal

1.00 1.000.99 1.000.95 0.941.00 1.001.00 1.001.00 1.001.00 1.00

0.94 0.930.90 0.890.61 0.62

Very Dry Dry

0.66 0.740.85 0.910.60 0.640.57 0.690.74 0.840.25 0.22

is met. Coverage is less than or equal to one, since

supplies are driven by demands in the model andredundant water is not sent from supply sources todistribution systems.

In the Amudarya basin, upstream distribution sys-tems would be mostly satisfied in the selected years,while downstream areas after Kashkadarya canal (Fig.2) such as Bukhara-Zerafshan and Cardzou wouldface water shortages. In the assumed Very Dry yearof 1995, only 3 1 per cent of Cardzous demand and52 per cent of Bukhara-Zerafshans demand could be

met. For the Lower Amudarya, users could onlyexpect to get 11 per cent of required water in theVery Dry year and about 61 per cent of supply in the

Norm al years. For the Syrdarya basin, the situationwould be more serious than for the Amudarya basin.During the Very Dry and Dry years, water supplyshortages would occur in almost every distributionsystem. The Lower Syrdarya users, even in the Wetyear of 1987, could not fully satisfy their waterrequirements. They could only satisfy 44 per centsupplies during theNorm al years and no more than

30 per cent supplies during the Dry and Very Dryyears. The shortages for downstream users may bealleviated to a small degree if upstream users areforced to reduce their withdrawals, but this wouldonly spread the unmet demand problem with theoverall water shortage situation remaining. Whilewater allocation in the region has been a source ofcontention since 1980s these projections suggest thatthe problems, if current patterns are allowed to persist,will only deepen. Withdrawal treaties between theupstream and downstream users along the two riverbasins, similar to the Colorado River Compact, areurgently needed in the near future.

The simulated annual stream flows entering theAral Sea from the two rivers are displayed in Table8. The Aral Sea inflow is projected to average 3.32

km3 from 1990 to 2000,2.99 km3 from 2000 to 2010,and 2.54 km3 from 2010 to 2020, a continuingdownward trend. When looking at monthly streamflows in drier years, as in Fig. 5, the seriousness ofthe situation is underscored. There would be twoextremely low-flow periods, January-March and June-September, during which no stream flow would enter

the Aral Sea. In a drier year, there would be almost6 consecutive dry months. As can be seen from thefigure, most of the annual stream flow would reachthe Aral Sea in spring, with little during the summer

seasons. These undesirable patterns suggest that bettersystem operation is needed for the regions waterstorage and regulating facilities.

By applying these projected stream flows to theAral Sea, we have calculated the water budgets of theSea and simulated the future changes of water leveland surface area of the Aral (Fig. 6). The Aral Seassurface area would decrease from its 1987 level of40.78 km2 to 9.41 km2 in 2015, while its water levelwould drop from 40 meters to 26.8 meters. It is clearthat without any action to reduce the demands or to

increase the supplies in the future, the sea wouldcontinue to shrink at roughly the same rate as it didin the 1980s devolving into one or several residualbrine lakes.

DIRECTIONS FOR POLICY

SC EN A R I O S

One of the primary objectives of our study is toexamine alternative future development scenarios forthe Aral region. Using the business-as-usual pro-

jections as a point of departure, the next step in theproject involves the creation of a number ofpolicyscenarios, or alternative water futures incorporatinga wide range of possible measures that alter business-



11/13

P

Table 8. Projected Yearly Flows Entering the AralSea (Unit: km3).Amudarya Syrdarya

Water-type Flow to Ara l Water- type Flow to Aral Tota l to Aral

1987 Very Wet 6.28 Wet 2.51 8.791988 N or ma l 1.96 Normal 1.18 3.141989 Normal 1.89 N o r m a l 1.14 3.031990 N or ma l 1.88 Very Wet 2.13 4.011991 We t 2.53 Normal 1.17 3.701992 Very Wet 3.51 Normal 1.13 4.641993 N or ma l 1.93 Wet 1.19 3.121994 Normal 1.87 Very Wet 2.38 4.25

1995 VeryDry 0.69 Very Dr y 1.23 1.921996 N or ma l 1.86 Dry 0.91 2.77

1997 N or ma l 1.85 Normal 1.07 2.921998 N or ma l 1.85 N o r m a l 1.09 2.941999 Normal 1.84 N o r m a l 1.08 2.922000 N or ma l 1.84 Normal 1.08 2.922001 N or ma l 1.83 Normal 1.07 2.902002 Very Wet 3.36 We t 1.13 4.492003 N or ma l 1.87 Normal 1.13 3.002004 Normal 1.82 We t 1.17 2.992005 Dry 1.09 Dry 0.97 2.062006 We t 2.45 Normal 1.05 3.502007 Normal 1.81 We t 1.12 2.932008 Normal 1.80 N or ma l 1.12 2.922009 Dry 1.06 N or ma l

1.08 2.14Very Dr y2010 N or ma l 1.79 Q.83 2.622011 Very Wet 3.26 Normal 0.50 3.762012 Dry 1.06 Dry 0.89 1.952013 Dry 1.04 Dry 0.18 1.222014 Normal 1.78 Dry 0.17 1.952015 Dry 1.03 We t 0.57 1.602016 We t 2.39 N or ma l 1.10 3.492017 We t 2.38 Normal 1.04 3.422018 Normal 1.76 Normal 1.04 2.802019 We t 2.37 Dry 0.88 3.252020 N o r m a l 1.75 Dry 0.16 1.91

as-usual projections. Policy scenarios will includeactions in three areas: changing demand patternsthrough efficiency .improvement and economic re-orientation, better managing the existing system anddeveloping new local water sources. Each of thesecategories of intervention encompass many separatemeasures, such as pricing policies, investment strat-egies, and technological and operational options. Forexample, irrigation efficiency can in principle beimproved through various technologies (sprinkler, drip,or trickle systems), through improved water applica-

tion scheduling or through land leveling and con-touring.The feasibility of any or all of these in theAral region is being studied in detail.

It is noted that our study has focused on thepotential for local solutions to address the problemsof the region. ,Nonlocal water supply enhancementsconsidered in the past include artificially increasingrainfall, increasing the rate of glacial melting, trans-

These exercises provide alaboratory for experimenting

with alternative futures

Vol. 17, No. 2 (1992)

ferring Caspian Sea water and transferring Siberianriver water. Each of these proposals has met withgreat concern about environmental impacts. More-over, critics of the most advanced of these proposals,the north-south Siberian water transfer, have raisedquestions about the cost-effectiveness of such a large-

scale project, the politics of inter-republic resourcetransfers and the impacts on local cultures. The project

is currently suspended.

We anticipate many alternative scenarios, each eval-uated on economic and environmental criteria. The

scenarios will begin with the business-as-usual scenario

reported here that quantifies the degree of water short-ages over time, the increasing pressure on the lake,and the scale of required remedial efforts and go onto anAral Sea stabilization scenario (just stabilizingthe sea requires significant improvements in todayswater-use efficiency and local supplies); and an AralSea restoration scenario requiring radical changes inthe future water and economic strategy for the area inorder to return inflow above equilibrium levels. In

each case, the feasibility and costs will be assessed.These exercises provide a laboratory for experi-

menting with alternative futures for the Aral Sea region.It is hoped that such glimpses of the future will help

steer current policies in a sustainable direction.

65


12/13

Projected Monthly Flows Entering the Sea

E -( 1.8 -0 1.7= 1 .6 -gt o 15 - .g 1 .4 -3 1.3 -B 1.2 -i i2 1.1 -lzE l-g 0.9 -

0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 -

&l 1995

0

JAN F E B MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

MonthFigure 5. Projected monthly flows enter ing the Aral Sea.

ACKNOWLEDGMENTS

We have benefitted from useful discussions andcontributions from A.E. Asarin, I.D. Cigelnaya, V.Dukhovny, N.F. Glazovsky, I.G. Horst, A.N. Krenke,

Y.M. Malisov, A.V. Meleshko, I? MickIin, I? Rogers,R. Razakov, G.V. Sdasyuk, I.S. Zonn. Any errors or

omissions in our analysis are, of course, the respon-sibility .of the authors alone.

REFERENCES

1. Micklin, Philip I?, Description of the Aral Sea: AWater Management Disaster in the Soviet Union,Science, Vol. 241, 1988, pp. 1171-1176.

2. Glazovsky, Nikita F.,The Aral Sea Crisis: the Source,the Current Situation, and the Ways to Solving It,

presented at the conference: The Aral Crisis: Causes,Consequences and Ways of Solutions, Nukus, SovietKarapalpak Republic, U.S.S.R., October 1-7, 1990.

3. Kotlyakov, V.M., The Aral Sea Basin: A CriticalEnvironmental Zone, Env ironm ent , January/Feb-ruary, 1991.

66

4.

5.

6.

7 .

8.

9.

10 .

11.

12.

white, Gilbert E, Editorial: A Lesson from the AralSea:Environment , January/February, 199 1.

Precoda, Norman, Requiem for the Aral Sea,AMBZ O, Vol. 20, May, 1991.

Stockholm Environment Institute - Boston Center,User Guide for W EAP Version 91.10, September,1991.

Zhitomirkaya, O.M., Climatological Description of the

Aral Sea Region, Hydro-Meteorological Publisher,Leningrad, U.S.S.R., 1964.SREDAZHYPROVODHLOPOK, Advanced Scheme of

Complex Use and Preservation of Water and LandResources ofAmudarya River- General Principles,Tashkent, U.S.S.R., 1984.

SREDAZHYPROVODHLOPOK, Advanced Scheme ofComplex Use and Preservation of Water and Land

Resources of Syrdar'ya River- General Principles,Tashkent, U.S.S.R., 1987.

SOYUZHYPROVODHOZ, Scheme of Complex Useand Preservation of Water and Land Resources of

Aral Sea Basin - General Principles, Moscow,

U.S.S.R., 1989 and 1990.U.S.S.R. Department of the Census, Statistical Year-

book of Republics, Moscow, U.S.S.R., 1987.SREDAZHYPROVODHLOPOK, Crop Irriga t ion Ratesfor Amudarya and Sy rdarya Basins, Tashkent,U.S.S.R., 1970.

Water International


13/13

Aral Sea Projections in Businks-as-usual Case1930 - 2020

80 , I

q Surface Area$ Flow Entering the Sea0 Sea level

q o+60-

50 -

40 -

30 -

20 -

10-

Surface Area 40.78

0011930 1940 1950 1960 1970 I II1980

1990 2000 2010 2020

YearFigure 6. Base case projection for the Aral Sea.

13. U.S. Department of Commerce, Bureau of the Census,Farm and Ranch Irrigation Survey (1988)," 1987Census ofAgricult ure, Vol. 3, Part 1, May 1990.

14. U.S.S.R. Department of Hydro-meteorology, Hy dro-logical Yearbook, Obninsk, U.S.S.R., 1979-1987.

15. Chembarisov, E.I. and B.A. Bahritdinov, Hy drochem -istry of River and Drainage Water in Central Asia,Ukituvchy, Tashkent, U.S.S.R., 1989.

16. SAO-HYDROPROEKT, Operational Rules of NurekReservoirs on Vahsh Riv er; Tashkent, U.S.S.R., 1989.17. SAO-HYDROPROEKT, Rogun Hydropower Station

on the Vahsh River, Technical Project, Vol. 2, Part1, Tashkent, U.S.S.R., 1972.

18. SAO-HYDROPROEKT, Tuyamuyun Gate on Amu-darya River, Technical Project, Vol. 2, Part 1, Tash-kent, U.S.S.R., 1978.

19. Howe, C.W. and EP. Linaweaver, The Impact of Priceon Residential Water Demand and Its Relation toSystem Design and Price Structure, Water Resources

Research, Vol. 3, 1967, pp. 13-32.

20. Boland, John J. et al., An Assessment of Municipal andIndust rial W ater Use Forecast ing Approach, A report

submitted to the Institute for Water Resources, U.S.Army Corps of Engineers, May 1981.

21. Davis, William Y. et al., IWR-MAIN Water Use Fore-casting Syst em, Version 5.1, Users Manual and Sys-tem Description, Prepared for the Institute for WaterResources, U.S. Army Corps of Engineers, December1987.

22. Kindler, J. and C.S. Russell, edited with B.T. Bower etal., Modelling Water Demands, Academic Press, Lon-don, 1984.

23. Strzepek, K.M., L.A. Garcia, and T.M. Over, MITSIM2.1, River Basin Simulation, Vol. I, User Manual,Center for Advanced Decision Support for Water andEnvironmental Systems, University of Colorado, 1989.

24. Loucks, D.P. and K.A. Salewicz, IR IS, An Interact iv eRiver Syst em Sim ulation Program, General Int ro-duction and Description, November 1989.

25. Hydrologic Engineering Center, U.S. Army Corps ofEngineers, HEC-5 Sim ulation for Flood Cont rol andConservation Syst ems, User Manual, 1982.

26. Hydrologic Engineering Center, U.S. Army Corps of

Engineers,HEC-3 Reservoir Sy st em Analysis for Con-servation, Programmers Manual, 1976.

Simulation of Water Supply and Demand in the Aral Sea Region

Documents