Transboundary Water Disputes in Central Asia: Using Indicators of Water Conflict in Identifying Water Conflict Potential

Valery Votrin, Master's thesis
Vrije Universiteit Brussel
Faculty of Medicine and Pharmacy
Master programme in Human Ecology
Academic year 2002-2003

Abstract

While the literature about the likelihood of an acute conflict over freshwater resources continues to grow, little work has been done to identify key drivers of such conflict and to establish indicators of water conflict potential. Such indicators need to incorporate a wide range of physical, social, economic and environmental variables to develop a comprehensive model covering basin specific situation. This method can be applicable in those cases of internationally shared rivers where progress towards successful resolution of a dispute was slow or unachievable. Considering all factors having influence on the river basin regime is becoming crucial to elaborate meaningful and workable framework to provide resolution to the growing water conflict.

The Syr Darya and Amu Darya Basins in Central Asia have been long known as “hot spots” where the potential for escalating existing tensions over water into a violent conflict was high. This study contributes to the identification and development of potential indicators of water conflict in Central Asia, while analysing some of them within the Geographical Information System (GIS). Unlike other studies that involve global level analysis, the thesis focuses on Central Asia’s regional and basin specific information. Taking the basin as a unit of research, the proposed study offers a model of relationship between water resources and social, political, economic and environmental patterns in the Amu Darya Basin. A parallel analysis of water conflict related variables aims at finding combinations of indicators that provide an indication of potential water conflict.

Table of Contents

Introduction

Hypothesis
Methodology

1      Water Use in Central Asia

1.1       Geographical background
1.2       Historical background
1.3       Existing water infrastructure
1.4       Water scarcity and access
1.5       Water quality

2      Regional Water Co-operation and Conflict in Central Asia

2.1       Water management during the Soviet Union
2.2       Post-independence water management
2.3       Legal framework for current water management in Central Asia
2.4       International actors
2.5       Recent water conflicts
2.6       Current water disputes
    2.6.1        The Syr Darya Basin
    2.6.2        The Amu Darya Basin
2.7       Potential water disputes
2.8       Multidimensional nature of Central Asia’s water disputes

3      Indicators of Water Conflict

3.1       Measuring water conflict potential
3.2       Testing indicators of water conflict
3.3       Role of GIS
3.4       Attempts at applying water conflict indicators to Central Asia
3.5       Water Event Intensity Scale
3.6       Aggregating data for the Amu Darya Basin
    3.6.1        Population
    3.6.2        Runoff
    3.6.3        Dams
    3.6.4        Minority groups
    3.6.5        GDP
    3.6.6        Overall relations
    3.6.7        Freshwater treaties

4      Water Conflict Potential in the Amu Darya Basin

4.1       Population density/runoff
4.2       Planned water infrastructure
    4.2.1        The Rogun Dam
    4.2.2        Golden Century Lake
4.3       Further internationalisation potential
    4.3.1        Possible secession of Karakalpakstan
    4.3.2        Role of Afghanistan
4.4       GDP
4.5       Overall relations/Water events
4.6       Freshwater treaties

5      Discussion

5.1       River boundaries
5.2       Trade complications
5.3       Sarez Lake: A natural threat to the basin’s stability
5.4       Impact of international law
5.5       Vox populi?
5.6       Summing up results
5.7       Conclusions and recommendations

References

Appendix

Table I. Overall relations between the Amu Darya Basin countries (excluding Afghanistan) in 1995-2003

Table II. Water-related events in the Amu Darya Basin in 1995-2002

List of Tables

Table 2.1: Water allocations under the Almaty Agreement
Table 3.1: Water Event Intensity Scale
Table 4.1: GDP per capita in the Amu Darya Basin countries
Table 5.1: Water conflict potential in the Amu Darya Basin

List of Figures

Figure 1.1: Map of Central Asia
Figure 1.2: Layout of water infrastructure in the Aral Sea Basin
Figure 4.1: Population density in the Amu Darya Basin (1995)
Figure 4.2: Population density in the Aral Sea Basin
Figure 4.3: Estimated annual runoff in the Amu Darya Basin
Figure 4.4: Number of dams per country in the Amu Darya Basin
Figure 4.5: Power stations in the upstream sections of the Amu Darya
Figure 4.6: Uzbekistan’s administrative units (including Karakalpakstan)
Figure 4.7: Overall relations between the Amu Darya Basin countries (excluding Afghanistan) in 1995-2003
Figure 4.8: Water-related events in the Amu Darya Basin in 1995-2002

Acknowledgements

I am greatly indebted to Nguyen Hahn Quyen of Institute of Geography, Vietnam National Centre for Natural Sciences and Technology, and to Aaron T. Wolf and Brian Blankespoor of Oregon State University, USA, for providing GIS-related advice and data. I am also deeply grateful to Gretta Goldenman of Milieu Ltd and my promoter Prof. Marc Pallemaerts for their guidance and comments. My thanks are also due to Alexey Tolmachov for technical support.

To Elena,
sine quo non

Water!.. Everywhere you turn here, you're involved with the lack of water!.. Why is there so little of it?.. The water was there. It dries up. And never again is there water…
Frank Herbert

- Dune -

Introduction

Over the last decades, there has been a growing speculation about the likelihood of an acute conflict or even war over freshwater resources. Scholars increasingly point out that the 21^st century might see the battles fought due to water scarcity. Indeed, water is the only resource h

aving no substitute, and the demand for it is constant and burning. All forms of the Earth’s life, including humans, need water to survive. The fact that we inhabit the ‘water planet’ soothes little as only less than 3% of total water resources on the Earth are freshwater. Its distribution is obviously uneven, with some nations suffering severe droughts every year and the others blessed with water abundance.

It is no wonder then that through the whole history of human race, water allocation and quantity fuelled tensions between various states, with particular role played by transboundary river resources. And although history shows that full-scale wars over water, proving to be neither strategically rational nor hydrographically effective, have never been fought (Wolf, 1998), water continues to be a source of intense disputes worldwide. The problem grows harder when it comes to the relationships between two or more countries over river water as a result of the “internationalisation” of a basin through political change. The number of international basins has grown from 214 in 1978 to 263 today. These international basins cover 45.3% of total land surface, affect about 40% of the world’s population, and account for about 60% of global river flow. Nineteen basins are shared by 5 or more riparian countries, with only the Danube being shared by 17 riparians, whereas five basins – the Congo, Niger, Nile, Rhine and Zambezi – are shared by between 9 and 11 countries (Wolf, 2001).

In this situation, such long-negotiated instrument of international water law as the 1997 UN Convention on the Non-Navigational Uses of International Watercourses, is of little help as it provides for equally contradictory ‘equitable use’ and ‘no significant harm’ principles: while the former is favoured by upstream countries, downstream riparians insist on emphasising the latter because it protects their own rights. It is also difficult to enforce the Convention in the absence of any international enforcing mechanisms. More importantly, the Convention hardly weighs out a variety of political, social, economic, demographic and environmental factors that encompass each shared river basin.

Negotiators whose task is to provide timely diplomatic intervention, or apply means of the so-called ‘preventive diplomacy’, in order to avoid the escalation of a dispute into open conflict need to be aware which basin is prone to water conflict well in advance. To do that, they need to identify potential indicators of conflict that incorporate a wide range of physical, social, economic and environmental variables, including those which can be analysed within a Geographic Information System (GIS), and to develop a comprehensive model to explore specific linkages between them. In particular, this method can be applicable in those cases of internationally shared rivers where progress towards successful resolution of a dispute was slow or unachievable. Considering all factors having influence on the river basin regime is becoming crucial to elaborate meaningful and workable framework to provide resolution to the growing water conflict.

Since the collapse of the Soviet Union, Central Asia has become a tangle of unresolved transboundary water disputes. Water is the most critical resource in Central Asia and it “has more often been the source of competition rather than the focus of conservation” (Hogan, 2000b). The absence of mechanisms to handle the water problems has already resulted in various accusations of improper water use. Consequently, the whole region becomes the site of potential conflict that requires a framework which should incorporate a great many variables to identify the proneness to water conflict and to allow for the possibility of preventive diplomacy. Such method which has never been used towards the specific problem of Central Asian water disputes can provide solutions based on a more holistic approach to natural resources, while recognising the historical, geopolitical and natural characteristics of the region.

This study seeks to identify potential indicators of water conflict and analyse some of them within the Geographical Information System (GIS). Unlike other studies that involve global level analysis, the thesis will focus on Central Asia’s regional and basin specific information. An attempt to answer a question of what indicators better point to water conflict will be taken. Chapter 1 focuses on the background of water use in Central Asia. Chapter 2 overviews main regional water conflict/co-operation trends and highlights major hot spots in the Amu Darya and Syr Darya Basins. Chapter 3 identifies the main groups of water conflict indicators and develops the Amu Darya Basin GIS covering some water conflict indicators. Chapter 4 provides specific details of each indicator. Finally, Chapter 5 discusses the main findings of the study.

Hypothesis

The study aims at testing the working hypothesis that proneness of a specific river basin to water conflict can be identified through a variety of indicators representing physical, political, environmental and socio-economic processes occurring in this specific region, thus providing better chance for using preventive diplomatic actions.

Methodology

The study’s methodology is based on the concept of international river basins offered by the Transboundary Freshwater Dispute Database (TFDD), Oregon State University (TFDD, 2002). The database contains a large collection of water treaties and agreements and is a sound foundation for transboundary water conflict research. For the purposes of this study, the Amu Darya Basin is examined. Taking the basin as a unit of research, the proposed study offers a model of relationship between water resources and social, political, economic and environmental patterns in Central Asia. There is also chance to explore the use of GIS such as ESRI Arc View for social research. A parallel analysis of water conflict related variables focuses on finding combinations of indicators that provide an indication of potential water conflict.

Back to Table of Contents

1  Water Use in Central Asia

1.1       Geographical background

Central Asia lies in the heart of the Eurasian continent and is comprised of the five former Soviet republics – Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan and Uzbekistan, as well as of northern Afghanistan and China’s Xinjiang Uighur Autonomous Region. All these share basins of the two major rivers in the region: the Amu Darya and the Syr Darya which form the Aral Sea basin. The Amu Darya catchment basin constitutes 62% of the region’s surface water resources and the Syr Darya forms the remaining 30%. The Basin’s population of over 35 million occupies about 1.5 million square kilometres and population density varies from about 10 persons per square kilometre in the desert plains to over 300 in the valleys and foothills of the mountains (Smith, 1995).

With domination of low-lying deserts, the climate in Central Asia is hot and dry, with low and irregular precipitation. Sharp daily and seasonal differences in temperature are typical, with long hot summers and cool moist winters. The annual precipitation in the lowland is only 80-200 mm, concentrated in the winter and spring, while in the mountains it ranges 600-800 mm. The region has very different climatic zones with distinct water demands for irrigation. Agricultural, industrial and personal needs can only be satisfied through diversion of water from the Syr Darya, Amu Darya and Zeravshan rivers and their tributaries.

The Amu Darya is Central Asia’s largest river and has the highest water bearing capacity of the region draining the catchment of 692,300 km². It originates in the Pamir mountains and forms the Pyandj river at the Tajik-Afghan border. Near town of Termez in Uzbekistan the Pyandj is joined by the Surkhandarya to form the Amu Darya. The Pyandj is augmented by a number of major tributaries including the Vaksh and Kafirnigan. From its headwaters, the Amu Darya flows 2540 km west across Tajikistan, Uzbekistan and Turkmenistan and finally crosses the Uzbek region of Karakalpakstan to discharge into the Aral Sea. Discharge is closely related to the amount of snowfall and summer temperatures, with mean annual flow between 46.9 and 108.4 km³ per annum with an average of 78.5 km³ (O’Hara, 2000).

The Syr Darya is the longest (2,212 km) river in Central Asia, having less catchment of 219,000 km². It rises in the mountains in Kyrgyzstan and has two major tributaries: the Naryn fed by over 700 glaciers in the Tien Shan, and the Kara Darya sourcing in the Ferghana and Alay mountains. After confluence in eastern Uzbekistan, these form the Syr Darya which crosses into Tajikistan and then re-enters Uzbekistan and finally flows into Kazakhstan where it discharges into the Aral Sea. Its discharge is smaller than the one of the Amu Darya, ranging from 21.4 to

Figure 1.1: Map of Central Asia
54.1 km³ per annum, with the average of 37.2 km³ (Vinogradov, 1996; O’Hara, 2000).
Source: http://www.grida.no/aral/maps/geog.htm

The Zeravshan, the third largest river in Central Asia, begins in the Pamir mountains in Tajikistan and crosses high mountain valleys before entering the flat plains in Samarkand region of Uzbekistan. As a result of intensive evaporation and consumption, the river gradually disappears in Kyzylkum desert before reaching the Amu Darya. Mean annual flow of the Zeravshan river is 5.2 km³ (Smith, 1995).

1.2       Historical background

The major states of the 19^th century’s Central Asia, the Bukhara Emirate and the Khiva and the Kokand Khanates, were put under the control of the Russian Empire by the end of the century, when the Anglo-Russian Agreement curtailed further Russian expansion in the vicinity of the northern borders of British India. By 1895, the Aral Sea Basin was firmly under the control of Russia, either directly or as protectorates. Thus, the scene for the future colonial relationship was set that persisted under the Soviets and indirectly led to the Aral Sea environmental catastrophe.

The 1917 Bolshevik Revolution toppled the Provisional Government (in power between March-November 1917) and proclaimed Russia the Soviet Republic, eventually to be replaced by the Union of Soviet Socialist Republics. The creation of Central Asian states, in their modern borders, was finished by 1924. In fact, the USSR established modern-day Central Asian states set to become independent nations in future (Bedford, 1998).

In 1991, after the Belovezh Agreement was signed by the leaders of the three most powerful Soviet republics – Russia, Ukraine and Belarus, the Soviet Union ceased to exist and the Soviet republics became independent. The independence appeared a somewhat of a shock for the five Central Asian states, as there was little domestic pressure for them to leave the Soviet Union and no real history as independent nations. Closely woven together economically, they had no experience of dealing with important economic decisions, which have always been taken in Moscow. Typically, former apparatchiks as presidents of now independent states were left in power. Nevertheless, necessary cosmetic changes were made such as banning their local Communist parties, reinventing themselves as people’s representatives, etc. Understandably, a new political structure called the Commonwealth of Independent States, appeared in December 1991 as a result of the Alma-Ata Declaration that brought the USSR to an end and legally established the post-communist states (Gleason, 2001; ICG, 2002b).

In their description of Central Asian statehood, Menon and Spruyt (1999) give several important particularities of state formation after the collapse of the Soviet Union in Central Asia to keep in mind. The states were formed by Stalin’s administrative diktat that assigned largely arbitrary borders to the republics and allotted such territories to “titular” nationalities, which is why state boundaries and ethnic composition in Central Asia lack correspondence; Central Asian states are late developers who have traditionally opted for interventionist economic policies and authoritarian government; clan, religious, ethnic and regional affinities here have not been displaced by centralising, high-capacity states; and, finally, these states lack any experience with democratic multi-party systems.

1.3       Existing water infrastructure

With irrigation agriculture being the largest water consumer in the region, all Central Asian rivers are utilised and heavily managed. Between 1960s and 1980s, an extended network of dams, reservoirs and canals has been built in the Aral Sea Basin. The largest water storage facilities are the Toktogul Reservoir in Kyrgyzstan controlling the flow of the Naryn river, and the Nurek Dam on the Vaksh river in Tajikistan, dubbed “the largest earth-fill dam in the world”. Only the Syr Darya basin has 22 operating reservoirs (Toryanikova and Kenshimov, 1999), including the Naryn-Syr Darya cascade of dams which consists of five reservoirs: three upper reservoirs with the over-year regulation - the Toktogul (projected total volume is 19.5 km³), the Charvak (2.0 km³), the Andijan (1.9 km³) - and also two channel reservoirs with the seasonal regulation - the Kairakkum (4.03 km³) and the Chardara (5.7 km³), with the aggregate active storage capacity of 24.1 km³ (Khamidov et al, 1999). Both rivers have a total of 274 water diversion structures and 612 km of main canals (Elhance, 1997).

Figure 1.2: Layout of water infrastructure in the Aral Sea Basin
Source: Tsusui and Hatcho, 1995

However, due to the lack of maintenance and high silt content in the Amu Darya water (up to 6 kg of silt and sand per 1 cubic metre), many important water facilities like, for example, Hauzkhan, Kurtli and Geok-Tepa reservoirs in Turkmenistan, are currently half silted up (Mouradov, 2002). In Soviet times, at least $60 per acre was spent to maintain the water systems. Uzbekistan currently spends less than $25 per acre. Tajikistan, recovering after the civil war, spends $4 (Wines, 2002). As a consequence, of 55 m³ of water to be used for irrigation, only 38-44 m³ of water reach the fields and plants actually receive 25-26 m³ due to the outdated irrigation technology (Yegorov, 2001).

Among the irrigation canals, the Karakum Canal, the world’s largest irrigation canal, is the most significant. Built between 1950 and 1987 and considered one of the great engineering feats of the Soviet era, it has taken water some 1,370 km across southern Turkmenistan from the Amu Darya to the Caspian Sea. Though currently in crisis due to the very poor maintenance, the Karakum Canal still supports 3.5 million hectares of rangeland and about 1 million hectares of cropland, providing also power generation, water for industrial and domestic use and partly navigation (Hannan and O’Hara, 1998).

Other canals include: the North and Grand Ferghana Canals transporting water from the Syr Darya to the Ferghana Valley; the Karshi Canal providing water to 1.2 million hectares in Uzbekistan’s Karshi Steppe; the Amu-Bukhara Canal irrigating land in the Bukhara Region in Uzbekistan from the Amu Darya; and the South Hungry Steppe and Kirov Canals irrigating the Golodnaya (Hungry) Steppe from the Syr Darya (Nanni, 1996).

Soviet-built irrigation schemes brought about widespread and rapid land degradation which only in Turkmenistan resulted in land being abandoned at a rate of over 46,000 hectares per annum in the 1970s (O’Hara and Hannan, 1999). Nevertheless, some of the projects are still under construction, for example, the huge Rogun Dam upstream of the Nurek. Given the current location of reservoirs, about 60% of total storage capacity of the Amu Darya and 9% of total storage capacity of the Syr Darya are controlled by Tajikistan, whereas Kyrgyzstan controls 58% of total storage capacity of the Syr Darya largely because of the Toktogul. Downstream Uzbekistan and Turkmenistan have very few water storage facilities and are entirely dependent for water on the upstream countries.

1.4       Water scarcity and access

The distribution of water in Central Asia is notably uneven. While upstream Tajikistan and Kyrgyzstan enjoy water abundance, downstream nations of Uzbekistan, Turkmenistan and partly Kazakhstan experience sharp water shortages. But the problem is predominantly consumption. At the heart of economic systems of Uzbekistan and Turkmenistan is cotton, which is the major hard currency earner for those countries. Cotton is a thirsty culture requiring intensive irrigation. With plans to triple cotton and rice production and almost complete dependence on upstream water resources, water overuse is common in downstream riparians and virtually no effort has been made to reduce water use. In addition, people in Central Asia consume 110 to 120 billion m³ of water per annum for domestic needs, which is several times higher than in the Middle East. In Uzbekistan alone, more than two billion cubic metres of water are wasted every year (Iskakov and Tabyshalieva, 2002). Water is cheap; in some regions its price is about 65 cents per “Olympic swimming pool” (Wines, 2002). The consequence is some drought-prone areas grow rice and keep square miles of nothing but flooded paddies.

Central Asia has become notorious worldwide as a site of the most dramatic environmental disaster, the Aral Sea Crisis. Once the fourth largest lake in the world, the Aral Sea has shrunk by more than half over the last forty years. The major reason for that was increased discharge of inflowing water for irrigation. Grand Soviet schemes for the production of cotton, rice and other irrigated cultures required huge amounts of water. However, large irrigation canals designed to provide massive agricultural expansion were on arid lands badly suited to irrigation where the soil is often much permeable and seepage is great. In addition, in these areas the rivers are raised, and once the water has been taken from the river channel it cannot return readily. The result of this overexploitation of water resources was that by 1995-1996 total annual inflow to the Aral Sea has dropped to 7 km³, the surface area of the sea was reduced to 33,000 km³, and the level had fallen by 37 m. The sea became saline and devoid of fish (Tanton and Heaven, 1999).

Throughout the sea’s former basin, water has mobilised deep salt reserves, raised the water table and waterlogged the fields as a result of over-irrigation. In Turkmenistan, 95% of irrigated lands suffer from salinisation, 50% - in Uzbekistan, about 30% - in Kazakhstan, 15% - in Tajikistan. Due to the mismanagement, several new lakes have appeared in the last decades, like the Aidar-Arnasai lake located in the middle course of the Syr Darya and formed in 1969 as a result of the effluent drain from Toktogul and Chardara reservoirs (Sievers, 2002; CAREC, 2002).

Intensive irrigation development in the deltaic areas was also accompanied by increased anthropogenic pressure in form of shrub felling, technogenic erosion and deflation. The period between 1978 and 1982 was marked by intensified process of desertification, as a result of the combination of excessive water withdrawals with a series of dry years (1974-1977). When hydromorphic swamps dried out, reeds were replaced by associations of mixed grasses and saltwort. Between 1982 and 1996, there was a further dramatic reduction in river discharge, with a total stoppage of inflow in case of the Syr Darya. In the driest years, the Amu Darya flow was almost nil. As a result, salt and dust storms became a major new environmental hazard for the area (Saiko and Zonn, 2000).

Despite that, water consumption goes at the same pace as before. The most populated nation in the region (25 million in 2002), Uzbekistan alone uses three-fifths of regional water supplies. In fact, industrial consumption of water in Uzbekistan and Turkmenistan is twice than in Kyrgyzstan and Tajikistan (Smith, 1995). Wasteful use of water on Uzbekistan’s very large irrigated area has contributed to a dramatic decline of the combined flow of the Amu Darya and Syr Darya and acted as a major factor in the Aral Sea Crisis. Uzbekistan showed itself to be a staunch proponent for this wasteful use of water arguing that it is vital for increasing cotton harvests. But excessive water stimulates vegetative growth that does not necessarily increase crop yields and may actually reduce them (Lerman, Garcia-Garcia and Wichelns, 1996). Consequently, agriculture has never been a component of a negotiating set for downstream countries as any sweeping agricultural reform would have resulted in foreign currency losses and Turkmenistan and Uzbekistan perpetuated the cotton monoculture to ensure social, political, and economic stability in the short term (Weinthal, 2001).

At the same time, Kyrgyzstan started implementing rational water use measures by reducing areas of thirsty cultures. As a result, in 2000, out of projected 7,641,600 m³ of water to be diverted from rivers, 6,866,200 m³, or 89.8%, was actually diverted, and 4,88,660 m³, or 86.5%, of projected 5,648,800 m³ was used for irrigation (Apasov, 2001).

Consequently, water overexploitation is the major reason for downstream riparians’ water scarcity. Uzbekistan and Turkmenistan use their water inefficiently and hence they experience water scarcity in a time of water abundance (Wegerich, 2001).

1.5       Water quality

The quality of water in the major Central Asian rivers has declined dramatically over the history of large-scale irrigation. Huge amounts of salt, fertiliser, herbicides, and pesticides found their way to the rivers as the return flow from the fields. Of 36-40 km³ of total annual return flow, about 50%, or 18-20 km³, bring to rivers about 115 million tons of salt and other harmful components, dramatically deteriorating water quality. In the Amu Darya Basin alone, a total of 84 million tons of salt is discharged into the river, transported with the water which is used to irrigate the fields (Kobori and Glantz, 1998; Spoor, 1998; Dukhovny and Stulina, 2001).

The groundwater table has risen and became as well contaminated with high levels of salts and other minerals. Groundwater quality ranges from a maximum of 0.5 g/L of total dissolved salts to 6 g/L, 20 times than in North America (about 300 mg/L). Total dissolved salts in drinking water reach the level of 3.5 g/L, with the salt limit set by the Uzbek government to be 1 g/L. In terms of chemical contamination, about of 65% of drinking water samples taken in Karakalpakstan proved not to correspond to standards (Small, van der Meer and Upshur, 2001).

In the Syr Darya basin, the chemical most commonly found both in river water and in the fields is butifos which has been widely used in the 60s-80s as defoliant with intensity of 1-1.5 kg per hectare. This organo-phosphorous agent with an acute oral toxicity affects central nervous system, liver and kidneys. Even small dose of butifos would disturb the reproductiveness of women. Since 1986, the production and use of butifos has stopped; however, it is still found in the water and at the bottom of the Amu Darya and Syr Darya rivers, as organo-chlorine pesticides (BHC, DDT) are (Ishida et al, 1995).

High levels of pesticides are found in the tissues of fish in the Amu Darya and Syr Darya. Samples of cane, rice, millet and wheat grown around the Syr Darya have been found to contain dangerous levels of benzopyrene, a carcinogen produced by car exhausts, oil and coal furnaces and manufacture of asphalt (Vinogradov and Langford, 2001).

Chemical pollution of drinking water has caused high cancer incidence, and substantial dioxin residues have been found in mothers’ milk, particularly in Karakalpakstan. In the midstream and downstream areas of the Amu Darya and Syr Darya, the incidence of waterborne diseases such as typhus, paratyphoid, cholera and viral hepatitis has increased enormously. According to Uzbekistan’s Academy of Sciences, only 8 percent of Uzbekistan’s rivers are “clean”, around 15 percent of river water is of “satisfactory” quality, and 41 percent is “bad”. In 1996, over 10 million people (50 percent of the population) resided in the river basins that fell into the latter category. Around 36 percent of river water is considered “dangerous” or “extremely dangerous”, particularly in Karakalpakstan and in the lower delta of the Zeravshan river, where 24 percent of the Uzbek population lives. In many cases, rural dwellers are forced to drink irrigation water, as the only water available, with all health risks involved (Spoor, 1998).

The Aral Sea crisis – or “syndrome” as Klötzli (1997) put it – has severely affected the area’s human ecology where millions of people are dependent on water and soil that appear to be highly contaminated. A new model of regional co-operation other than that used in the Soviet times was needed to address specific environmental, socio-economic and political problems of the region. Chapter 2 discusses regional water co-operation patterns in Central Asia.

Back to Table of Contents

2   Regional Water Co-operation and Conflict in Central Asia

2.1       Water management during the Soviet Union

Within the Soviet Union, inter-republican water resources were managed on the basis of water use plans. These plans were developed by local Ministries of Land Reclamation and Water Management and then sent to Moscow to the Ministry of Land Reclamation and Water Management of the Soviet Union for approval. These plans and schemes provided for annual water withdrawal limits with respect to each tributary, reservoir or canal and the limits were calculated against annual crop requirements. A number of bilateral agreements was signed between the republics to correct water allocation, such as the an agreement between the Turkmen SSR and the Uzbek SSR on water quotas from the Amu Darya, between the Kyrgyz SSR and the Uzbek SSR on the use of waters of the Sokh river, etc. Neither of these agreements contained any provisions with respect to the quality of return flows, i.e. the drainage water disposed of into the rivers. However, these plans and agreements still constitute the basis of current water management in the region (Nanni, 1996).

Under the Soviet system of water allocation, water quotas imposed by Moscow favoured downstream countries at the expense of the upstream riparians: water-abundant Kyrgyzstan and Tajikistan were supposed to supply irrigated agriculture economies of Uzbekistan and Turkmenistan with water in spring and summer when water should be available for cotton fields. In autumn and winter, when Kyrgyzstan and Tajikistan experienced peaks in electricity demand, they were supplied with Turkmen and Uzbek gas and Kazakh coal to satisfy energy consumption. They also received electricity from downstream countries during winter to be compensated for the hydropower produced in summer. Maintenance and operating costs of dams and reservoirs were covered by Moscow.

2.2       Post-independence water management

After independence, the need for all riparians to enter into an agreement regulating water allocation in the Basin has become apparent. Such agreement signed on 18 February 1992 in Almaty did not go far from water quotas set up under the Soviet Union. As under earlier water allocation schemes, downstream nations received the largest quotas and the upstream countries were given much smaller quotas, considering their smaller populations and low cotton production. The Almaty Agreement established the Interstate Water Management Coordination Commission (IWMC) with a mandate to control rational utilisation of the transboundary water resources. IWMC’s decisions regarding intake limits and rational utilisation of water are obligatory for all users. It is responsible for governing the two inter-republican Basin Water Management Bodies (Basseinoe vodnoe ob’edinenie – BVO): BVO Amu Darya and BVO Syr Darya. Hence, the five preferred to continue with the BVO management system put in place during the Soviets (IGC, 2002b).

Initially, IWMC was responsible for great many issues including water development and allocation planning, water quality control and conservation, environmental protection, preparing annual water allocation plans, defining limits of water use by each riparian, etc. With the establishment of other intergovernmental institutions between 1993 and 1995 such as the Interstate Council on the Aral Sea Basin and the International Fund for the Aral Sea, functions of the IWMC became somewhat duplicated and its relationship with other intergovernmental bodies remain unclear (Vinogradov and Langford, 2001; ICG, 2002b).

The Almaty Agreement attempted to secure the existing situation where water was apportioned to allow maximum utilisation whereas the international concept of equitable and optimum utilisation was kept aloof. The Agreement also lacked the provision about dispute settlement. According to it, water disputes are to be settled by the Ministers of Water Resources of the five states. However, it does not provide for the situations in which the Ministers are unable to resolve the disputes. In absence of any inter-republican dispute settlement body, this seems to be serious flaw (Nanni, 1996; Vinogradov and Landgford, 2001). Furthermore, the problem is also actual functioning of the water management bodies, BVOs, who lack funding and legal powers. According to the Almaty Agreement, they have to submit a budget to the ICWC for approval. Once a budget has been approved, the five members states are supposed to contribute a proportion of their budget based on the percentage of river water allocated. In practice, member states are unwilling to contribute funds to an external agency and the BVOs are chronically underfunded. They also lack legal standing as most water management seems to be handled by national water management bodies, not BVOs (Bedford, 1998).

Table 2.1. Water allocations under the Almaty Agreement

Hence, upstream countries were further restricted in their economic development and ability to satisfy heating needs during winter months as downstream countries introduced world prices for gas and coal. Unable to afford them, Kyrgyzstan increased electricity production at Toktogul reservoir that caused sharp reduction in water to downstream Uzbekistan and Kazakhstan for irrigation during cotton season. After serious tensions in 1997, the countries have come to enter a framework barter agreement in 1998. Under this, Uzbekistan and Kazakhstan would provide Kyrgyzstan with gas and coal during winter in return for irrigation water during spring and summer. However, barter agreements are in constant breach due to a number of reasons: they are ready usually in spring when Uzbek and Kazakh fields are in dire need of water; the parties lack trust and do not keep their commitments; and there is lack of control mechanisms. Also, several years of severe drought have affected the situation dramatically, causing Kyrgyzstan to reduce water for irrigation in summer and triggering floods in Uzbekistan during winter. Attempts to adjust quotas to reflect this have so far failed. Downstream countries have shown little understanding of upstream riparians’ demands to expand their water use (ICG, 2002b).

2.3       Legal framework for current water management in Central Asia

Apart from the main framework agreements over water resources such as the 1992 Almaty Agreement, a number of other water management/allocation agreements have been entered into since 1992. Below is the list of some of them:

• Agreement between the governments of Russia and Kazakhstan on the joint use and protection of transboundary water resources, Orenburg, 1992
• Agreement on the creation of the International Fund for the Aral Sea, 1993
• Programme on the joint actions on the improvement of environmental situation in the Aral Sea Basin, 1994
• Declaration on the problems of sustainable development in the Aral Sea Basin, Nukus, 1995
• Statement of leaders of Kazakhstan, Kyrgyzstan and Uzbekistan about the energy/water use, Bishkek, 1996
• Declaration on the creation of the Interstate Commission on Sustainable Development and the need of preparation of a Convention on Sustainable Development, Almaty, 1997
• Framework Agreement between the governments of Kazakhstan, Kyrgyzstan and Uzbekistan on joint use of the Syr Darya River Basin water/energy resources, Bishkek, 1998
• Agreement between Kazakhstan, Kyrgyzstan, Tajikistan and Turkmenistan on co-operation in the field of environmental protection and Agreement on biodiversity conservation, Bishkek, 1998
• Agreement between Kazakhstan and Kyrgyzstan on the interstate use of hydrological facilities at the Chu and Talas rivers, 2000
• Annual intergovernmental water/energy agreements (CAREC, 2002).

However, actual regional co-operation over water resources, other than entering into numerous agreements, is “glaringly absent” (Spechler, 2001). Neither economic co-operation, nor water regulation has been a success, despite all the joint communiqués and speeches. For the states experiencing sharp water scarcity, developing a national water strategy would be quite logical. Yet none of Central Asian states has developed one, though Kazakhstan, Kyrgyzstan and Tajikistan have started working on it. Implementing existing agreements appears to be another problem. Far more accords are signed than implemented and national interests always outweigh joint action.

None of water treaties specifies a goal of reducing water use or making agriculture less water-intensive. The sceptical attitude of downstream countries to multilateral co-operation deters them from any environmental and financial commitments (Klötzli, 1997).

Modern transboundary watercourse law largely based on the 1997 UN Convention on the Law of Non-Navigational Uses of International Watercourses urges riparians not only to create legal agreements to manage their shared resources, but also to find “joint management mechanism” and to cede sufficient sovereignty to them to make them effective. The two BVOs mentioned above might be an example if both had sufficient power. And none of Central Asian states have become a party to the 1997 Convention, although Kazakhstan did accede to the Convention on the Protection and Use of Transboundary Watercourses and International Lakes.

2.4       International actors

From the first days of independence, bilateral donors, international agencies and private foundations have funded dozens of projects in order to resolve difficult water situation in Central Asia. In doing so, technical solutions were preferred to political and economic ones. Especially active were the World Bank, United Nations Development Programme (UNDP), the EU Programme of Technical Assistance to the CIS (TACIS), and the United States Agency of International Development (USAID) who spent millions of dollars to help resolve the Aral Sea crisis. Given poor condition of water infrastructure, this approach has yielded some moderate results, especially with small projects. Effort to tackle water from political perspective, however, has resulted in problems. A lack of willingness of Central Asian states to co-operate has buried great many initiatives, like, for example, an attempt by the Organisation of Security and Co-operation in Europe (OSCE) to persuade the five to discuss the region’s water problems at a water conference to be held in London. This effort came to nothing with Turkmenistan’s president’s response that international conference in London was not the right place to discuss Central Asia’s water. In his turn, Uzbekistan’s president said that his country had a thousands years of experience in managing water problems and he preferred bilateral discussions to a multilateral conference (Eggleston, 2000; ICG, 2002b).

In trying to find a solution to the desiccation of the Aral Sea, AralGEF, a project funded by the Global Environmental Facility (GEF) and designed to create small but viable wetlands and fisheries on the place of the Aral Sea through restoration of a modest flow to the old seabed, and the UNDP Aral Seashore Rehabilitation and Capacity Development Project are so far most prominent. According to the World Bank, in order to successfully restore “modest flow” to the Amu Darya, agricultural runoff should be entirely restored. Thus, huge lakes that developed over the decades of water negligence from the excess water of the Toktogul reservoir and that now support local agriculture, fisheries, recreation areas and biodiversity habitats would thus be bound to disappear. And the main success of UNDP’s project on rehabilitation of the Aral Seashore consisted in providing 16,000 residents of Karakalpakstan with safe drinking water and planting thousands of trees that withered immediately because they were unsuited for local climate.

Part of the problem is the failure of Central Asian states to support the donor projects in a meaningful way, either administratively or financially. Neither donors seem to need anything more than paper reports. Even National Environmental Action Plans funded by the World Bank, UNDP and TACIS are based on the ineradicable idea, the only one familiar to the region’s old guard of water bureaucrats, that plan, not action, is needed to save the situation. This approach endorsed by donor staffs inherently contradicts the overarching goal to help Central Asian states to move away from a planned economy and to embrace the market economy and decentralisation (Sievers, 2002; ICG, 2002b).

2.5       Recent water conflicts

Not surprisingly, this situation has already led to numerous small-scale local conflicts, began in the late 1980s when the central authorities weakened their grip on Central Asia. In 1990, the outbreak of conflict in the Kyrgyz town of Osh, on the border with Uzbekistan, claimed over 300 lives and was provoked by fierce competition for water together with high population density, limited arable land and ethnic dimension (large population of Uzbeks living in the area).

Since summer 1993, there have been serious water tensions between Kyrgyzstan and Uzbekistan. Kyrgyzstan was blamed by the Uzbek authorities for releasing too much water from the Toktogul reservoir. Extra water did not reach the Aral Sea but was dumped instead into the Aydarkul depression, the large ‘sinus’ which has developed as a result of years of negligence (Klötzli, 1994). In 1997, Uzbekistan has deployed 130,000 troops on the Kyrgyz border to guard the reservoirs straddling the two countries (Hogan, 2000b; Grozin, 2001). In June 2001, the Kyrgyz parliament adopted a law classifying water as a commodity, and the government followed up by announcing that the downstream countries would be charged for the water they use. Uzbekistan’s response was to cut off all deliveries of gas to Kyrgyzstan and accuse Kyrgyzstan of failing to honour the barter agreement to provide Uzbekistan with water in return for oil and gas. Although weaker in political and military terms Kyrgyzstan acknowledged this failure, Uzbekistan would be emboldened to behave in a more aggressive manner towards its neighbours. The two were on the verge of violent conflict for several times (Khamidov, 2001).

2.6       Current water disputes

Central Asia’s two major rivers, as well as their tributaries, have become a focus for growing competition among their riparians, with the Syr Darya being a particular point of tensions. The Amu Darya is rapidly becoming a locus of disputes as the governments of Turkmenistan and Uzbekistan become more hostile towards each other competing also for water, and Afghanistan is about to demand its share. These tensions have so far been contained without conflict, but all parties have shown a willingness to put their interests first at any cost, including military intervention. Due to their reliance on agriculture, Uzbekistan and Turkmenistan view irrigation as a key security issue (ICG, 2002b).

2.6.1    The Syr Darya Basin

The Syr Darya is shared by four states and, in case of Uzbekistan, is shared twice as after flowing from Kyrgyzstan and crossing the Uzbek part of the Ferghana valley the river flows into the Tajik territory in the western Ferghana Valley and then pours again into Uzbekistan. After crossing the Hunger Steppe, the Syr Darya runs into Kazakhstan.

As indicated above, Kyrgyzstan and Uzbekistan have had particularly discordant history over the use of water from the Syr Darya. Since independence, Kyrgyzstan faces serious economic problems, mainly because of a shortage of energy supply from Russia, Kazakhstan and Uzbekistan. The primacy of energy production over the irrigation needs downstream has already created a major discord between Uzbekistan and Kyrgyzstan (Klötzli, 1994). The 1998 Syr-Darya Framework Agreement between the two has been broken by both sides. The implementation of such barter agreement runs across one major problem – all barter agreements are delayed until the late spring when the downstream countries urgently need water for irrigation. As this might be the case, Kyrgyzstan would have had an incentive to produce less electricity. However, due to Kyrgyzstan’s uncertainty whether enough gas would be provided, it produces electricity to protect itself giving rise to a vicious circle (IGC, 2002).

Uzbekistan intensified the tension more than once by acting in a unilateral manner. In July 1997, it cut off 70 percent of downstream flow, which caused a riot among the Kazakh farmers whose 100,000 hectares were threatened (O’Hara, 1998; Hogan, 2000b). History of altering water flow by upstream riparians is no more soothing. In summer 1999, Tajikistan released 700 million cubic meters of water from its Kairakum reservoir without warning its downstream neighbours. As a result, cotton crops in southern Kazakhstan which has received less water than was agreed, were devastated. The situation was seriously aggravated by Kyrgyzstan’s concurrent move to reduce the flow to southern Kazakhstan in retaliation for Kazakhstan’s failure to supply coal under the barter agreements. After months of talks, the incident was finally settled (Hogan, 2000a).

At issue is also the Naryn-Syr Darya cascade of dams in Kyrgyzstan. Every year Uzbekistan insists on releasing water from it to improve downstream agriculture. Several times, the conflict was on the verge of war. In 1997, Uzbekistan deployed 130,000 troops on the Kyrgyz-Uzbek border, near the Toktogul reservoir, to conduct military exercises aimed at seizure of a ‘well guarded object’, using the armour and helicopters. Meanwhile, Kyrgyzstan, through media leak, hinted that in case the reservoir would be blown up, the resulting flood would sweep away Uzbekistan’s Ferghana and Zeravshan Valleys (Grozin, 2001).

Kyrgyzstan has tried to persuade Uzbekistan and Kazakhstan to share the maintenance and operating costs of the Toktogul reservoir but these attempts were turned down by the downstream countries. Kazakhstan claimed that it would not be able to pay the costs – between US$ 15 and 27 million per annum. Thus, the opportunity for the downstream riparians to settle the dispute was missed. However, by adopting Law on the Interstate Use of Water Objects, Water Resources and Water Management Installations on 29 June 2001, the Kyrgyz parliament left the door open to push the downstream countries into negotiations regarding the maintenance costs of the Toktogul reservoir as later Kyrgyzstan stated that in fact it demanded to pay only for the water passing through Kyrgyz reservoirs, i.e. share maintenance costs. This was welcomed by Kazakhstan who agreed to pay for the maintenance of the Kyrgyz water installations, but initially opposed by Uzbeks. However, in March 2002 Uzbekistan reached the agreement with Kyrgyzstan that it would share some costs in return for the guarantee that it would receive water for irrigation. Had more attention have been paid to the barter agreements working properly, the main step to resolving the Syr Darya dispute between Kyrgyzstan and Uzbekistan would be taken (ICG, 2002b).

Shared water storage facilities, like the Andijan reservoir located in the Uzbek part of the Ferghana Valley but supposed to re-channel some water back to Kyrgyzstan, also represent an inter-state problem (Chait, 1998).

2.6.2    The Amu Darya Basin

The Amu Darya is shared by four countries – Tajikistan as the upstream riparian, Afghanistan, Uzbekistan and Turkmenistan – and forms the border in some stretches between Tajikistan, Uzbekistan and Afghanistan, and between Turkmenistan and Uzbekistan.

The Amu Darya is much less regulated and has fewer dams and reservoirs to cause potential problems. However, there are serious tensions along the flow of the river not only between the upstream and downstream riparians, as, for example, between Tajikistan and Uzbekistan, but also between the middle and lower riparians, for example, Uzbekistan and Turkmenistan.

According to the 1992 water quota agreement, Tajikistan is entitled to 9 m³ of about 75 m³ of annual flow of the Amu Darya, or 12 percent. This is considered low by Tajikistan who needs to expand its agricultural output to supply the growing population with food. Tajikistan’s agriculture is underdeveloped since the Soviet times, and the irrigation system is derelict and in need of urgent repairs. Tajikistan sees the only way out as using more water either by increasing its water quota from the Amu Darya or by diverting the Zeravhsan river. As 95 percent of the latter are used by Uzbekistan, this would cause serious tensions with Tajikistan’s much powerful neighbour.

In contrast, increasing the Amu Darya quota seems to be quite easy, since Tajikistan has an upper hand in distributing water resources of the Amu Darya. In principle, nobody could prevent Tajikistan from taking more water than was allocated by the water quota agreement. It is very hard to monitor Tajikistan’s performance, as most equipment needed has been destroyed during the civil war in 1992-1997. But even Tajikistan were to increase its water quota moderately, this would have an immediate impact downstream.

The same water/energy complex as with Kyrgyzstan has developed between Tajikistan and Uzbekistan. Tajikistan’s central and southern parts are well provided by electricity from the Nurek hydro plant, and northern Tajikistan having no grid lines with the rest of the country relies on Uzbekistan’s intermittent supplies of electricity and gas in winter. In return, Tajikistan provides power to southern Uzbek provinces and often requests that Uzbekistan switched off electricity to northern Tajikistan to keep imports within the agreed limit not to pay higher price. This causes serious discontent as Tajikistan is forced to have electricity rationed in many provinces due to poor state of Tajikistan’s grid lines. The country desires to develop its hydropower resources to break dependence on Uzbekistan. But increasing hydro consumption would seriously affect the downstream access to seasonal water supplies and to create further discord along the Amu Darya course.

The most dramatic conflict over the Amu Darya water resources is between downstream nations of Uzbekistan and Turkmenistan. Both equally depend on their cotton production and irrigation agriculture and both claim that each of them exceed their water quotas. Due to the very poor state of Turkmenistan’s water infrastructure, most water received by Turkmenistan is wasted. The country does not want to spend huge funds for the expensive rehabilitation of crippling Turkmen canals and draws off more water from the Amu Darya instead. The relations between two countries dramatically worsened in the late 2002 when the Uzbek ambassador has been declared persona non grata in Turkmenistan on accusation of participating in the conspiracy to oust and kill President Niyazov. Uzbek-Turkmen relations over water can grow even worse, given Turkmenistan’s ambitious plan to complete a huge reservoir in the Karakum desert, called the Golden Century Lake. Another point of contention is the Tyuyamuyun reservoir in the delta of the Amu Darya divided between Uzbekistan and Turkmenistan. Both sides feel displeased with the wasteful use of water, and this led to an outbreak of violence in 1992 over the re-direction of drainage waters and raids by both sides to cut off pipes and irrigation canals (Smith, 1995). Today, the Tyuyamuyun remains one of the several disputed areas in continuing water dispute with Uzbekistan.

Throughout the independence period, rumours have circulated of a small-scale secret war between the two states over the river resources, Uzbekistan troops taking control of water installations on the Turkmen bank of the Amu Darya, and even of a massacre of a large number of Uzbekistan troops in Turkmenistan in 2001. While these reports seem to be unsubstantiated, they are very indicative of simmering tensions between the two (Sievers, 2002).

2.7       Potential water disputes

The complicated water situation in Central Asia forced the governments of the five states to consider alternative plans for developing water infrastructure to gain better control over water resources. Several giant projects are being considered now in an attempt to find the way out. With little exception, all of them date back to the Soviet planning system, and several projects have been frozen in the 1970s-1980s due to the lack of funds. Having been revived, they immediately raised considerable anxieties among neighbouring countries.

Among those projects are: the Rogun reservoir able to give Tajikistan full control over the Amu Darya, Golden Century Lake in the middle of Turkmenistan’s Karakum desert, the project of diverting Siberian rivers of Ob and Irtysh to help replenish the Aral Sea, etc. The former two are discussed in Chapter 4.

The idea to divert Siberian water to Central Asia was abandoned in the late 1980s when it has become clear that the project would cause irretrievable damage to the environment. In recent years, however, the project has been revived by Uzbekistan. The main argument of the project’s proponents, among whom is the mayor of Moscow Yury Luzhkov, is that the Siberian rivers frequently flood and hence have excess of water to share with water-deficient Central Asian countries. Yet in reality floods are the integral part of a river system and there is no any “excess” water. In projecting a huge canal 2,500 kilometres long, 200 metres wide and 15 metres deep from Western Siberia to Central Asia, it was estimated that 23 cubic kilometres of water would be diverted, of which 2 cubic kilometres would be lost due to the filtration. Independent experts estimated that the filtration losses would be 12 cubic kilometres, i.e. half of the water to be diverted, which is normal, given that such a cyclopean structure would reach enormous levels of filtration, water-logging and salinisation. One expert put the idea of a miscalculating the project like this, “In the Soviet times, we would have been given the task of calculating a project cost of a bridge to the Moon, and we would have calculated that. But nobody would care about the consequences” (Radio Liberty, 2002). Huge, economically not viable and environmentally dubious Siberian water diversion project has already contributed to the difficult water situation in Central Asia. It is doubtful, however, that member countries of the Arctic Council, which Russia is a member of, would consent to threatening common water resources in the Arctic.

2.8       Multidimensional nature of Central Asia’s water disputes

Water disputes contribute to the broader complex of problems across the region, including border disputes, Islamic extremism, high population growth, ethnic tensions, clan competition, human rights and political instability. The desiccation of the Aral Sea has been the important factor to the worsening socio-economic conditions in the area, fuelling nationalist ideas among the population of Karakalpakstan, the Uzbek autonomous republic adjacent to the disaster zone, and aggravating water situation in the region. Lack of public participation, particularly in authoritarian Turkmenistan and Uzbekistan, and attempts of those governments to find military decisions in already difficult relationships between those countries and upstream states makes the whole situation white-hot.

Ethnic dimension is extremely important for ethnically diverse Kyrgyzstan and Uzbekistan. As it happens, local conflicts here have been more serious than wider ones. Disputes over land and water resources provoking wider ethnic conflict have led to hundreds of victims in Kyrgyzstan in 1990. Poverty, rising costs and crumbling water infrastructure are adding to strains in local water system. Water affects the poor who end up paying the large proportion of their income for the resource.

Especially vulnerable to violent eruptions over water and ethnicity is the Ferghana Valley shared by Tajikistan, Kyrgyzstan and Uzbekistan which has already seen the outbreaks of violence, like it was in 1990, when bloody clashes between inhabitants of the Kyrgyz town of Osh claimed over 300 lives, or earlier, in 1989, when hundreds of the Meskhetian Turks, who had been deported to Central Asia by Stalin in the 1940s, were killed in the Uzbek town of Ferghana in what was called one of the most dramatic episodes of inter-ethnic relations in the Soviet Union.

The roots of any future ethnic strife in Central Asia lie in the unresolved social and economic problems, competition for scarce water and grazing resources and contentions over discriminatory land allocations (Elhance, 1997). Potentially explosive ethnic cleavages tear apart many countries in the region. Kyrgyzstan is divided between northern and southern part, with the latter gravitating towards Uzbekistan and inhabited by large proportion of ethnic Uzbeks who have repeatedly demanded to give the area more autonomy. Uzbekistan’s headache is the large Tajik population living in Samarkand and Bukhara. Over the last years, Uzbekistan’s policy was to suppress Tajik ethnic identity by not allowing schools to learn in Tajik language and forcing many Tajiks to call themselves ethnic Uzbeks in their Uzbek passports. The same situation is in Tajikistan where a lot of ethnic Uzbeks live. In the late 1990s, after the relations between Uzbekistan and Tajikistan have worsened, the Tajik government expelled many Uzbeks to neighbouring Kyrgyzstan and Afghanistan in an attempt to get rid of “extremist elements”. Ethnic minorities have often been viewed by authoritarian governments as potential provocateurs, separatists and extremists (ICG, 2002a).

Historic competition between peoples of Central Asia is fuelled by the fact that their leaders do not seem to like one another. There is a great personal competition between the three former Communist leaders, Uzbekistan’s Islam Karimov, Kazakhstan’s Nursultan Nazarbaev and Turkmenistan’s Saparmurad Niyazov. Each wants to show that his country is the region’s most powerful and he therefore should be viewed as the most prominent political figure in Central Asia. As a consequence, the leaders of the region does not see their countries to be the part of a functioning regional subsystem and are rather isolationist in their policies. For all this, they regularly meet to discuss their common problems (Olcott, 2001).

Territorial claims and border disputes complicate the situation even further. Given complex ethnic mosaic in Central Asia, Soviet planners did not build administrative units along ethnic lines and took great care not to construct republics with strong ethnic identity which would allow them to eventually secede from the Soviet Union. As a result, thorny disputes as to whose territory was initially whose have occurred, burdened by the territorial exchanges. For example, Karakalpakstan began life in 1924 as part of Kazakhstan but in 1938 had been given away to Uzbekistan. Moreover, Uzbekistan, Kyrgyzstan and Tajikistan all have small enclaves on each other’s territory which are nominally a part of their country but are geographically isolated. Only Kyrgyzstan has two Tajik enclaves, with population of some 30,000 people, and five Uzbek ones, with population of about 50,000 people. Issues related to them are highly divisive, and solving this problem appeared to be very difficult (ICG, 2002a). However, the process of delimitation of the borders has already begun. In 2000-2003, Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan signed a number of border delimitation agreements and started works on delimiting the borders. By 2001, Kazakhstan and Uzbekistan have finally settled the border problems and exchanged disputed territories. Yet many issues, like the problem of the Uzbek enclaves in Kyrgyzstan, still remain unresolved.

In this context, the growing need to take into account all factors surrounding water disputes is apparent. Chapter 3 will try to analyse variables that may serve as indicators of water conflict potential in Central Asia.

Back to Table of Contents

3  Indicators of Water Conflict

3.1       Measuring water conflict potential

Water is the most politicised of all natural resources and it is more likely to become a source of armed conflict. It is no wonder therefore that such possibility has been widely discussed (see, for example, Gleick, 1993; Gleditsch, 1997; Just and Netanyahu, 1998; Wolf, Yoffe and Giordano, 2003). Despite the growing literature on water and conflict, little work has been done to provide arguments for the common thesis that “growing conflict over water looms ahead” (Samson and Charrier, 1997). Generally, the Jordan and Nile basins are cited to give an example of international conflict prone basins. On the other hand, Wolf (1998) gave a historic evidence of co-operation between riparians and stated that the only recorded war over water was fought 4500 years ago between two Mesopotamian states, Lagash and Umma, in what is now southern Iraq. The same author pointed out at the “loose definitions” in the terminology of the literature about water and conflict where “terms such as conflict, dispute, tensions, and war are regularly used interchangeably” (Wolf, Yoffe and Giordano, 2003).

Nevertheless, in order to provide indication of potential water conflict, combinations of variables, or indicators, have to be developed. Undoubtedly, the practice of establishing various sets of indicators, including those for sustainable development, over the past decade has greatly influenced this idea. However, unlike indicators of sustainable development designed to communicate with the public at large to provide clear picture of a country’s progress towards sustainable development, indicators of water conflict proneness are meant to serve as a ‘yardstick’ for decision makers who are involved in negotiations regarding specific river basin management. Yet such ‘limited’ value does not impede the indicators of being used by international scientific and political community able to translate them into a pressure on the governments of the riparians involved in a dispute well before this would turn into a violent conflict. The use of Geographic Information System (GIS) makes such analysis even more effective by bringing together spatial and non-spatial variables, thus facilitating identification and interpretation of potential indicators.

One of the frequently cited instruments which could serve as a model for developing a set of water conflict indicators is the Index of Human Insecurity (IHI) developed to facilitate identification of vulnerable or insecure regions. It is considered an “aggregate measure of human welfare that integrates social, economic, and political exposures to and capacity to cope with a range of potentially harmful perturbations” (Lonergan, Gustavson and Carter, 1999). The IHI identifies four “key system components” – the environment, the economy, society, and institutions. Within each of these four indicator categories are four variables, each of which measure either a key structural relationship (e.g., linkages, defining characteristics) or a key functional relationship (e.g., processes, flows) of the system. Where data in a time series are missing, IHI developers utilize statistical techniques to establish a complete time series for all indicators and all countries, where there is sufficient initial data. However, the index for each year is specific to that year, making it difficult to compare changes in a country’s IHI from across years (Yoffe and Ward, 1999).

Another water-specific indicator is water stress measured by Falkenmark’s (1992) Water Stress Index (WSI) which divides the volume of a country’s available water resources by its population. This measure, however, does not account for a country’s ability to adapt to water stress, such as with more efficient irrigation technology. Ohlsson (1999) has developed a Social Water Stress Index (SWSI) to incorporate a measure of a country’s adaptability. The SWSI is a water stress index (freshwater availability per capita) divided by UNDP's Human Development Index and then divided by 2 (rounded to nearest wholes). Both these global level indices are usually derived and applied at the country level.

Several other potential indicators were also mentioned in the literature, including overall population growth rates within a country, population density within and outside a basin, relative power and riparian position of countries within a basin (Wolf, 1999), the degree of democratisation of countries sharing a river basin and so on.

3.2       Testing indicators of water conflict

The methodology for establishing indicators of international freshwater conflict and co-operation was developed under Basins At Risk (BAR) project at Oregon State University, USA (Wolf, Yoffe and Giordano, 2003). BAR which has spanned for 4 years effectively developed and created legal and spatial framework to further evaluate international river basins at potential risk for future water conflict. All 263 international river basins were delineated and a database documenting historical incidents of international water conflict and co-operation between 1948 and 1999 was set up. Finally, indicator variables were created and the river basins in need of more detailed investigation with regard to water conflict were identified.

Indicators have been selected in accordance with the following criteria: relevance to the selected framework; general availability of the data; existence of a theoretical or empirical link with security issue; and an adequacy of spatial and temporal coverage allowing for effective representation and modelling (Lonergan, Gustavson and Carter, 1999). Emphasis was made on the regional and basin scale indicators rather than on indicators of potential water conflict at global scale. Internationalisation of a basin was at special focus. Assuming that there is a causal link between the internationalisation of a basin and incidents of conflict among the states that now share that basin, the presence of ethnic minorities with nationalistic aspirations becomes a potential indicator (Wolf, 1999).

A key question regarding the above variables is whether they are relevant to indicating water conflict. Yoffe (2001) provides a good account of statistically testing indicators previously cited in the literature. As main statistical tool, linear regression was used to assess the relative strength of various independent variables in explaining the variability of the event data. Also, other statistical methods such as two-sample t-test were employed. Linear regression has been chosen because “it offered a concise summary of the mean of the response variable as a function of an explanatory variable” (Yoffe, 2001).

Sixteen indicators, including GDP per capita, population density, number of dams, water stress, HDI, hydropower have been tested using linear regression, three indicators such as freshwater treaties, adjacency and riverine contiguity have been analysed using two-sample t-test, and four indicators such as dam density, freshwater treaties, climate and precipitation had no statistical tests conducted on them due to structure of data.

Yoffe’s most important finding was that most of the commonly cited water conflict indicators proved to be unsupported by the data. Neither government type, climate, IHI, water stress or number of dams, nor agricultural dependence on water resources and energy needs showed a significant relevance with water conflict. Based on the assessment, river basins are at potential risk of freshwater conflict if:

• population density exceeds 100 people per 1 km²
• per capita GDP is less than $765
• overall unfriendly relations (<-1 at water event intensity scale)
• there are politically active minority groups that may lead to the internationalisation of a basin
• large dams or other water development projects are planned
• no or limited number of water treaties is available (Yoffe, 2001).

This study will employ these indicator variables as the research framework.

3.3       Role of GIS

A number of the attribute data mentioned above can be effectively represented within a Geographic Information System (GIS), such as ESRI ArcView, designed to capture, store, manipulate, analyse and visualize disparate sets of geographically referenced data. Moreover, some spatial information such as runoff and slope would be extremely difficult to analyse without the use of a GIS.

However, GIS has its limitations inherent to the technology, including problems of multiple scales and data interpretation. The examination of indicators at an international, national and sub-national level introduces questions of scale that will need to be further addressed. There is also a lack of distinct frameworks to analyse the interaction of human and environmental systems (Yoffe and Ward, 1999). Linked to the problem of frameworks is the problem with interpreting non-spatial matters using spatial analysis. For instance, the degree of democratisation of a country might be difficult to represent if one assumes that different areas have different levels of democratisation. Special problem regarding Central Asia’s GIS is the lack of data; several important databases located in Tashkent, Uzbekistan, contain a lot of vital information but this information is virtually closed for external use.

3.4       Attempts at applying water conflict indicators to Central Asia

Previous attempt to apply potential indicators of water conflict to Central Asia’s water disputes has been made by Myers (1993) who created a water scarcity index using population and per capita water availability data. Assuming that water stress is to occur when more than 2,000 people share every unit of one million cubic meters of available water, the author calculated water indices for each of Central Asia’s countries by comparing 1991 population data with all available water and the amount of water available solely from indigenous sources (i.e. that available from precipitation within the borders of a country). In 1993, all the countries appeared to have sufficient water resources per capita. Both downstream countries of Turkmenistan and Uzbekistan relied on water flowing in from outside and indigenous water resources stress was high (2,180 people per one million m³ of indigenous water in Uzbekistan and 3,287 people per one million m³ of indigenous water in Turkmenistan).

Klötzli (1994) used a number of indicators for vulnerability to water scarcity developed by Gleick (1993), including the ratio of water demand to supply, the per capita availability of water, the dependence on imported surface water and the importance of hydroelectric production. Having calculated these, the author came to a somewhat obvious conclusion that “these indicators… show a distinct difference between the water-rich republics of Kyrgyzstan and Tajikistan, and the republics not controlling the sources of water courses like Uzbekistan and… Turkmenistan… The vulnerability indices not only express different access and pollution control opportunities, but also different, often contradictory modes of water utilisation (hydropower vs. agriculture)” (Klötzli, 1994).

Similar method based on Gleick’s findings was used by Smith (1995) to determine the relative degree of dependence of particular regions on water resources and to identify those regions that may become potential focal points for future intra- and interstate conflicts. It is argued that high population growth can be more damaging to the environment than high population density and southern Uzbekistani provinces as Surkhandarya and Kashakdarya and several regions in the Ferghana Valley that experienced some of the highest population growth rates in the region between 1981 and 1991 are most likely to become a scene of a resource-related conflict. The author points out that the areas combining high population growth and no indigenous water supplies (Andijan and Ferghana in eastern Uzbekistan, Karakalpakstan in the west) are of prime concern. As Turkmenistan’s Dashhovuz region and Uzbekistan’s Karakalpakstan share the water of the lower Amu Darya, future water-induced conflicts may occur in this region. Each region was ranked according to the basic characteristics believed to affect the water conflict potential (population growth rate/density, per capita water availability from total and indigenous sources, share of water derived from external sources, minority population share of total population). Eight of ten regions having the highest rankings for water-resource vulnerability are located in Uzbekistan, with top of the top four located in the Ferghana Valley. Thus, the regions constituting the Ferghana Valley and those occupying the lower Amu Darya appear to be the most sensitive to potential water-induced conflicts.

Carney and Moran (2000) examined the interstate affect in Central Asia toward Russia, i.e. “the current discernable feeling of a government toward another government” which is expressed “…in all behaviour whether it takes the form of word or deeds”, and developed “scale of affectivity” ranging arbitrarily from -2 to +2. Although different in goals, this study employed the similar technique in scoring and coding news reports from Foreign Broadcast Information Service (FBIS) as discussed below.

Also mentioning the Aral Sea Basin’s famous conflict potential, Yoffe (2001) and Wolf, Yoffe and Giordano (2003) argue that the basin countries tend to negotiate current conflicts. However, none of the treaties signed recently include all the basin riparians, and therefore potential for continued disputes, at least in the nearest future, is likely.

Given global scale of BAR project, considering and spatially interpreting all physical, social, economic and political variables for Central Asia becomes even more vital and can provide solutions based on a more holistic approach to natural resources, while recognising the historical, geopolitical and natural characteristics of the region.

3.5       Water Event Intensity Scale

Wolf, Yoffe and Giordano (2003) recount of compiling a systematic database for water conflict/co-operation worldwide, in an attempt to gather every reported interaction between two or more nations, where water was the driver of the event. Using various international news databases, like FBIS and its online version, World News Connection (WNC), the authors developed ‘a friendship-hostility scale’ – a ratio of all co-operative to conflictive events between nations over water, having modified a scale created by Conflict and Peace Databank (COPDAB) project directed by Professor Edward E. Azar.

COPDAB’s scale ranges from level 1, representing the most co-operative events, to level 15, representing the most conflictive events, with level 8 representing neutral events. To make it water-specific, the authors inverted it so that neutral events were centred on zero, to make it ranging from -7 to +7, with -7 denoting the most conflictive events, 0 denoting neutral events, and +7 denoting the most cooperative events. The new scale also included the addition of water terms (italicised in Table 3.1), and a new category, "formal declaration of war." To accommodate this category, which is not part of the original COPDAB Scale, category 13 (Small scale military acts) and 14 (Limited war acts) were merged into one category, number 13. Category 14 was given the heading and description of category 15 (Extensive war acts causing deaths, dislocation or high strategic costs), and Category 15 was changed to indicate a formal declaration of war (Yoffe, 2001).

Table 3.1: Water Event Intensity Scale

For the purposes of this study, BAR global Water Event Intensity Scale is adapted to the Amu Darya basin countries in Central Asia to determine overall relations and identify focal points for the future conflicts while concurrently applying GIS-based variables of water conflict.

3.6       Aggregating data for the Amu Darya Basin

In creating ArcView data coverages for the Amu Darya Basin, data for international river basins kindly provided by Dr Aaron T. Wolf was used which was then matched to the US Geological Service Hydro1k (USGS, 2000) dataset, a global coverage of streams and drainage basins derived from digital elevation data. The data collected fell into roughly two categories: biophysical (e.g. runoff, population, dams) attributes suitable for the spatial analysis and economic/political (e.g. GDP per capita, overall relations, ethnic minorities) attributes. Although some of the variables used are quantitative and can be measured, the indicators comprising political category are qualitative in nature (internationalisation of a basin, future water infrastructure, water treaties). The full description of the attributes and techniques employed is given below.

3.6.1    Population

In pulling various water-related attributes together in order to draw a picture of a basin’s specific conflict proneness, population growth is among key factors for assessing water scarcity. The relation of water access to population distribution should therefore be assessed. As a river basin rarely follows political borders, per basin population density instead of per region population density was evaluated. In doing so, the Landscan gridded population of the world (Landscan, 2001) was used to approximate population distribution. Building on the Landscan data proved very effective as the Landscan team used remotely sensed slope, land cover, road proximity and night time lights to refine the GIS-based gridded population cell values, providing the best accurate dataset. Gridcell values for the Aral Sea Basin were summed to combine this table with the Amu Darya Basin, thus calculating a population density for that area, as described by Yoffe (2001) (Figure 4.1).

3.6.2    Runoff

Runoff is considered the total amount of surface flow in a given area. Any assessment of a water resource related issue would be incomplete without some approximation of water availability within the study area (Yoffe, 2001). Data for the Amu Darya runoff were obtained through a spatial resolution grid of composite runoff fields based on observed river discharge and simulated water balances (Fekete, Vörösmarty and Grabs, 2000). The project was the outcome of a joint effort of the Complex Systems Research Centre at the University of New Hampshire and the Global Runoff Data Centre in Koblenz, Germany. The project combined “…two sources of information (observed discharge and simulated runoff) to estimate continental runoff” which was “the most reliable assessment at present” (Fekete, Vörösmarty and Grabs, 2000). To estimate the composite runoff for the Amu Darya, runoff data for the eight gauging stations located in the Amu Darya basin were used. The cell values in mm/yr were multiplied by the area of the associated grid cell in sq. km to produce a runoff volume grid (mm*km²/yr) (Figure 4.3).

3.6.3    Dams

The third key factor in assessing water scarcity is the number and density of dams in a study area. According to Yoffe (2001) and Wolf, Yoffe and Giordano (2003), when statistically testing potential water conflict indicators, dams themselves did not appear to be a potential indicator for water conflict, yet in basins without water treaties lower dam density basins tended to exhibit slightly less conflict. With negative overall relations between countries and absence of a transboundary institution, unilaterally setting a large dam or diversion project can provide a context for the conflict over water, while positive relations and presence of a transboundary institution can mitigate the situation. To determine the number of dams and dam density for the Amu Darya Basin, global dataset provided by Wolf was used, based on the Digital Chart of the World data and International Commission on Large Dams (ICOLD) World Register of Dams (Figure 4.4). More importantly, large proposed water infrastructure projects that proved to have significant influence on the whole water situation in the basin were identified (e.g. the Rogun Dam in Tajikistan, etc.).

3.6.4    Minority groups

Internationalising a basin provides an important setting for future water conflict (Wolf, Yoffe and Giordano, 2003). Those basins whose management institutions were developed under a single jurisdiction and then became divided among two or more nations when that jurisdiction suddenly collapsed, showed much higher levels of conflict (Yoffe, 2001). In order to check for the active nationalist movements who might cause future internationalisation of the Amu Darya Basin, data from two major sources have been analysed: Minorities At Risk Project, at the University of Maryland’s Centre for International Development and Conflict Management (Gurr, 2000) and Unrepresented Nations and Peoples Organisation (UNPO, 2002).

3.6.5    GDP

Low per capita GDP (< $765/person according to the World Bank lowest income country definition) was identified to be one of the indicators of conflict over water (Yoffe, 2001). As Wolf, Yoffe and Giordano (2003) put it, “the higher the per capita GDP, or the lower population density, the greater the co-operation – barely”. To establish macro-economic data for the Amu Darya Basin countries, the World Bank data for 2000 were used.

3.6.6    Overall relations

It is pointed out (Yoffe, 2001) that countries that co-operate in general also co-operate over water, and countries with overall unfriendly relations are also unfriendly over water issues. To construct a ‘friendship-hostility scale’ as appeared in Wolf, Yoffe and Giordano (2003), a rough water event intensity scale was built employing the technique described in Section 3.5. In gathering country-to-country events and interactions, two major sources were used: the WNC and Transboundary Freshwater Dispute Database. Both build on the reports about Central Asia by the two leading Russian news agencies – ITAR-TASS and Interfax, as well as on the news coverages by Eurasianet and the Mashhad Voice of the Islamic Republic of Iran (Tajikistan news). Notably, the water event data are based on public reports and therefore lack reports on non-public co-operation (“picnic table talks”, non-official visits, etc.)

3.6.7    Freshwater treaties

Generally, freshwater treaties mitigate conflict, and no or limited freshwater treaties for a basin increase the likeliness of conflict over water. To analyse existing freshwater treaties for the Amu Darya Basin, the International Freshwater Treaties Database at Oregon State University (TFDD, 2002), containing a full text of over 400 international freshwater related agreements, was used.

Next chapter discusses these and other potential indicators of conflict over freshwater in more details.

Back to Table of Contents

4  Water Conflict Potential in the Amu Darya Basin

4.1       Population density/runoff

The Amu Darya’s population is extremely unevenly distributed. Only in Uzbekistan which has the region’s highest population density (average >50 people per km², with total population of about 25 million in 2002), 80 percent of land is desert. In 1990, the rural population occupied a total of 4.5 million hectares of arable land, or about 16 percent of all available agricultural land in the country, and thus the effective density of rural population was 2.7 people per hectare of arable land. Given the rate of natural population growth in Uzbekistan being one of the highest among the former Soviet republics, rural population steadily increased from 1984 to 1994 to 61 percent, while the arable land and cultivated area remained practically unchanged (Lerman, Garcia-Garcia and Wichelns, 1996).

Population distribution in the Amu Darya Basin varies accordingly, from 20-30 people per km² in the downstream sections of the river (except for the river stretch just before flowing into the Aral Sea that runs across the Uzbek territory and where population density is about 40-50 people/km²) to 25-40 people/km² in the middle course and to 20-35 people/km² in the upstream sections of the Amu Darya (Figure 4.1).

Figure 4.1: Population density in the Amu Darya Basin (1995)

The population pattern in Central Asia has remained the same over thousands of years. People tend to inhabit fertile oases and valleys rich in water. Many human habitats can be found located along the rivers (Figure 4.2), and almost all of them date from the pre-Islamic times.

Figure 4.2: Population density in the Aral Sea Basin
Source: Akmansoy, 1998

In terms of runoff (Figure 4.3), the Amu Darya can be divided into three sections: an upstream section, a middle course section, and a downstream section. The upstream section extends from the point of confluence of the Vakhsh and Pyandj rivers up to the town of Kerki, the middle course is between Kerki and the Tyuyamuyun Gorge, and the downstream section is confined to the mouth of the river. In the upstream section, the flow of the Amu Darya steadily increases due to the Kunduz, Kafirnigan, Surkhandarya, and Sherabad tributaries.

Figure 4.3: Estimated Annual Runoff in the Amu Darya Basin

The reverse situation can be observed in the middle course and in the downstream section: the flow gradually declines due to natural losses of runoff and huge water diversions for agricultural needs. It is quite difficult to evaluate those natural and anthropogenic losses as the data from many gauging stations for the last two decades are not available. Average annual runoff at the town of Kerki (upstream section) between 1957-1987 was 1903 m³/s, together with water diverted to Karakum and Karshi canals, 1351 m³/s in the Tyuyamuyun Gorge (middle course) in the same period and 656 m³/s in the downstream section (Ivanov and Izmaiylov, 1995). Figure 4.4 represents total number of dams per Amu Darya riparian.

Figure 4.4: Number of dams per country in the Amu Darya Basin

4.2       Planned water infrastructure

Unilateral decisions to embark on new water projects are not uncommon in Central Asia. While providing for national solutions, such decisions and further actions are among key drivers of potential water conflict. Trying to find a way out of the existing situation, some Central Asian governments were forced to develop plans for building more infrastructure to get control over water resources. Two projects which are the legacy of the Soviet-style gigantic undertakings and which caused extremely negative reaction among neighbours are discussed below.

4.2.1    The Rogun Dam

Started in 1976, the Rogun Dam, on the Vakhsh river, was projected to be 335 metres high – the highest in the world - with the capacity to produce 3,600 MW of energy. The dam was expected to begin operation in 1993. However, in 1990 the construction was halted due to the escalating political situation which turned into a 5-year civil war in Tajikistan. In 1993, a massive flood destroyed most of what has been already built. By that time, $802 million was invested into the project, with the total cost of $2.3 billion (Dyuzheva, 2002; Yerofeeva, 2002).

Figure 4.5: Power stations in the upstream sections of the Amu Darya
Source: Dyuzheva, 2002

Tajikistan already controls 40 percent of the flow of the Amu Darya through the Nurek reservoir. The Rogun would put it firmly in control of the river, allowing to control the flow into Uzbekistan’s Surkhandarya and Kashkadarya Provinces that already experience sharp water shortages. Knowing Uzbekistan’s opposition to the project, Tajikistan continues to press with the completion of the Rogun Dam that would need $700 million to $1 billion to complete.

Most international finance institutions are reluctant to put any money into the project, claiming that it would cost more than any benefit it might offer. They are also aware that the project would seriously strain relations between Tajikistan and Uzbekistan. However, getting Tajik government concentrate on more feasible, realistic and low-cost project yielded no fruits. On 29 October 2002, the Baltic Construction Company, one of the biggest Russian investment/construction groups, became a contractor and investor for completion of the Rogun Dam project (Dyuzheva, 2002; ICG, 2002b).

4.2.2