Five Major Irrigated Basins in the Arid Zone Irrigation-Induced River Salinization: L. Smedema INTERNATIONAL IRRIGATION MANAGEMENT INSTITUTE Irrigation-Induced River Salinization INTERNATIONAL WATER MANAGEMENT INSTITUTE Irrigation-Induced River Salinization: Five Major Irrigated Basins in the Arid Zone L. Smedema iv The author: Lambert Smedema is Theme Manager, International Program for Technology and Research in Irrigation and Drainage, World Bank, Washington, D.C. Smedema, L. 2000. Irrigation-induced river salinization: Five major irrigated basins in the arid zone. Colombo, Sri Lanka: International Water Management Institute. / water resources development / irrigation management / river basin development / drain- age / arid lands / water quality / ISBN 92-9090-391-0 Copyright © 2000 by IWMI. All rights reserved. Responsibility for the contents of this publication rests with the author. Please send inquiries and comments to: IWMI-research-news@cgiar.org The International Irrigation Management Institute, one of sixteen centers supported by the Consultative Group on International Agricultural Research (CGIAR), was incorporated by an Act of Parliament in Sri Lanka. The Act is currently under amendment to read International Water Management Institute (IWMI). IWMI thanks the European Union for the financial support given for this study. IWMI also gratefully acknowledges the unrestricted and program support provided by the following Governments, Development Banks, Agencies, and Foundations: Australia, Belgium, Brazil, Canada, China, Denmark, Germany, India, Iran, Japan, Mexico, Netherlands, Norway, Pakistan, Peru, South Africa, Sweden, Switzerland, United Kingdom, United States of America and European Union (EU); African De- velopment Bank (AfDB), Asian Development Bank (ADB), Ford Foundation, and World Bank. The Governments of Nepal, Pakistan, and Sri Lanka provided logisti- cal support for IWMI-related activities in these countries. v Contents Acknowledgements .............................................................................................................................. vii Chapter 1 INTRODUCTION ........................................................................................................... 1 Chapter 2 SALINITY ........................................................................................................................ 3 2.1 Terminology ........................................................................................................... 3 2.2 Salt Sources ............................................................................................................ 4 2.3 Salt Regimes ........................................................................................................... 5 2.4 Safe Limits ............................................................................................................. 6 Chapter 3 IRRIGATION DEVELOPMENT ................................................................................... 9 3.1 Water Resources Development ............................................................................. 9 3.2 Irrigation-Induced Salinity ................................................................................... 11 3.3 Disposal of Saline Drainage Water ..................................................................... 13 Chapter 4 THE SELECTED BASINS............................................................................................ 15 4.1 Brief Characterizations ........................................................................................ 15 4.2 Land and Water Resources .................................................................................. 18 4.3 Institutional Arrangements .................................................................................. 19 Chapter 5 SALT BALANCES ........................................................................................................ 21 5.1 Methodology ........................................................................................................ 21 5.2 Patterns and Features .......................................................................................... 22 Chapter 6 RIVER SALINITY ......................................................................................................... 27 6.1 Natural Salinity .................................................................................................... 27 6.2 Irrigation-Induced Salinity ................................................................................... 28 Chapter 7 CONTROL MEASURES .............................................................................................. 35 7.1 Management of the Drainage Return Flows ...................................................... 35 7.2 Combating Salt Mobilization .............................................................................. 36 7.3 Basin Management ............................................................................................... 36 7.4 Non-River Disposal ............................................................................................. 37 7.5 Institutional, Financial, and Implementation Arrangements .............................. 37 7.6 Economics ............................................................................................................ 38 Chapter 8 SUMMARY AND CONCLUSIONS............................................................................ 41 8.1 Equilibrium Profile ............................................................................................... 41 8.2 Salinization Profile ............................................................................................... 41 8.3 Mobilization Profile ............................................................................................ 42 Annex A. Aral Basin ....................................................................................................................... 43 Annex B. Colorado Basin ............................................................................................................... 51 Annex C. Indus Basin ..................................................................................................................... 61 Annex D. Murray-Darling Basin .................................................................................................... 69 Annex E. Nile Basin ....................................................................................................................... 81 LITERATURE CITED ........................................................................................................................ 87 vii Acknowledgements This publication was prepared as a joint activity by the International Water Management Institute (IWMI) and the International Program for Technology and Research in Irrigation and Drainage (IPTRID). Both institutions are extensively involved in research in the sustain- able management of land and water resources for irrigated agriculture of which the control of river salinity is obviously a key element. Both are repositories of extensive knowledge on the subject and, moreover, could be relied on for well-established contacts in the selected basins. The study could also draw on a number of related studies undertaken by IWMI, the World Bank, and by some of the national basin management institutions. The author grate- fully acknowledges the ready cooperation of all the institutions and resource persons con- tacted. 1 CHAPTER 1 INTRODUCTION Rivers in irrigated basins typically serve both as the sources of the irrigation water and as the sinks for the drainage water. While most of the irrigation water is diverted from the upper reaches of the rivers, most drainage water returns to the lower reaches. These functional and spatial linkages basically explain the increase in downstream river salinity observed in almost all irrigated basins in the arid zone (figure 1). This irrigation-induced increase in river salinity generally reinforces already existing natural trends. In the humid zone where the natural salinity and the impacts of the irrigation diversion and drainage return on the river regime are proportionally much less, this phenomenon is generally not noticeable. The increase in downstream river salinity typically becomes more pronounced when the irrigated land use expands and intensifies and when municipal and industrial developments add to the salt-loading of the river water. Control of the river salinity is usually possible but generally requires the adoption of costly and painful changes in current water use practices and/or poses equally painful restrictions on further developments in the basin. However, without such control measures, the underlying processes may be expected to continue until the river salinity reaches its final equilibrium level, which may be so high that many important water- related human activities in the lower basin would become seriously impaired and many environmental values irreparably damaged. The irrigated agriculture in the basin, generally being the largest water user, would be one of the main victims. This publication reports on a study of the increase in river salinity and the implemented control programs in five major irrigated basins in the arid zone. The study is limited to salinity. Although salinity is no longer the only water quality concern in irrigated basins, in most basins it is still by far the dominant concern. The basins studied are: the Aral Sea Basin in Central Asia, the Colorado Basin in the western USA, the Indus Basin in Pakistan, the Murray-Darling Basin in southeastern Australia and the Nile Basin in Egypt. These basins cover a range of physiographical and geological conditions but are all located in the (semi) arid zone and are relatively large in size (mostly in the 0.5–1.0 million km2 range). Irrigation was the main initial water development and all five basins still support important irrigated agriculture sectors. However, other sectors with a stake in the use of the river water are rapidly gaining importance. The role of legal and institutional factors is recognized although in this publication these factors have not been dealt with as extensively as the technical factors. Together, the five basins constitute a fairly representative set from which important lessons on the control of river salinity in irrigated basins in the arid zone can be drawn. 2 F ig ur e 1. S ch em at ic o ut li ne o f th e pr ob le m . 3 CHAPTER 2 SALINITY For the benefit of the general reader, this chapter presents some essential background information on the nature and dynamics of salts in the soils and in water, their sources and global patterns, and their potential harmful impacts. For a more detailed discussion, reference is made to the standard works on soil and water salinity by Ghassemi, Jakeman, and Nix (1995); Szablocs (1989); and Tanji (1990). 2.1 Terminology Salts commonly occurr in almost all natural substances and environments. Most salts found in soils and waters are generally naturally formed mineral salts. In solution, the salt compounds dissociate, completely or partly, into the constituent anions and cations. The major cations found in soil and water are sodium (Na+), calcium (Ca++), magnesium (Mg++), and potassium (K+) while the major anions are chloride (Cl-), sulphate (SO4 -), hydrocarbonate (HCO3 -), carbonate (CO3 -), and nitrate (NO3 -). In the solid phase, the ions are bonded and form mineral salts such as CaCO3, MgCO3, Ca2SO4, NaCl, MgCl2, Na2SO4, and others. Salinity is a general term indicating the presence of rather high levels of salts. The level of salinity is generally assessed by measuring the concentration of the salts in a water solution. This concentration expressed as the weight of the salts per unit volume of water, is called the “total dissolved solids” (TDS), which is usually given in mg/l or in its numerical equivalent, “parts per million” (the ppm unit). These units are typically used when the TDS is determined by desiccating the solution. The salt concentration may however also be determined indirectly by measuring the electrical conductivity (EC) of the solution, as there is a fairly consistent linear relationship between the salt concentration of a solution and its EC value. According to the International System of Units (IS), the unit for electrical conductivity is deci-Siemens per meter (dS/m), which is numerically equivalent to the formerly used mmho/cm unit. For the common range of river salinity, the TDS and the EC units may be mutually converted by applying the rule that 1.0 dS/m = 640 mg/l. Soil salinity is standardly measured in the “saturation extract,” which is the solution extracted from a fully saturated soil-paste. The electrical conductivity of this extract is known as the ECe value and most of the documented crop responses to soil salinity are based on this indicator. As salts differ in their solubility in water, the salt composition of a water body may change when the solution is diluted (e.g., by the inflow of freshwater) or become more concentrated (e.g., by evaporation). Most chloride salts are highly soluble as are most nitrate and sulphate 4 salts whilst alkali earth carbonates (CaCO3 and MgCO3) and gypsum (CaSO4) are less soluble. The latter salts gradually precipitate when solutions become more concentrated. These low- soluble salts are again mobilized when the solutions become diluted. These precipitation and dissolution processes commonly occurr in the soil. The river salinity, however, almost never exceeds the solubility limits of even the least-soluble salts. 2.2 Salt Sources The salts found in the soils and in the waters in a river basin generally originate from, or are classified as, one or more of the following types: 2.2.1 Primary Salts These are the salts continuously released from rocks and soils by mineral weathering and dissolution. Due to limited rainfall and leaching, most arid-zone landscapes are rich in primary salinity. The releases are, however, counteracted by salt precipitation and other forms of salt immobilization and indications are that in most irrigated basins in the arid zone, except for those basins with relatively young soils and with only moderate salinity, rates of immobilization are about the same as those of release. 2.2.2 Fossil Salts Almost all basins in the arid zone harbor large quantities of “fossil salts,” trapped in the sedimentary rock strata and aquifers, formed during previous geological periods when (part of) the basin was occupied by the sea, salt lake, salt playa, salt desert, or some other salt sink. By geological uplifting and/or erosion, some of these deeper buried saline strata may later have moved to the surface and become exposed. Release and mobilization of these fossil salts by natural or by anthropogenic processes constitute a virtually inexhaustible source of salts in many irrigated basins. 2.2.3 Atmospheric Salts Salts carried by rainfall or wind may constitute a significant source of salt in coastal areas. The atmospheric fallout of salts in coastal zones may well amount to 100–200 kg/ha/yr. while even in areas far inland it may still be about 10–20 kg/ha/yr. (Tanji 1990). Some parts of Australia are reported to be exposed to an annual atmospheric salt fallout of 10–100 kg/ha, presumably mostly originating from nearby desert areas. 5 2.2.4 Salts in Irrigation Water All irrigated land has to cope with the salt-load carried by the applied irrigation water. Without adequate leaching and drainage, these salts risk remaining in the soil/land when the water (evapo) transpires. The annual salt-loading by this source depends on the volume and salinity of the applied irrigation water but in the arid zone it may easily be about 2–3 tons/ha (refers to an annual application of 1,000 mm with a salinity of 200–300 mg/l). As most of these added salts are not main plant nutrients, the crops take up only a very small percentage. 2.2.5 Salts in Fertilizers The average salt content of a package of commonly used fertilizers may well be about 65–70 percent and even though it may generally be assumed that at least half of these salts will be taken up by the crops, the net salt-loading by the applied fertilizers may not be insignificant in basins where modern high-input farming is practiced (may well be about 200–300 kg/ha/yr., which is however still quite small compared to the salt-loading by the irrigation water). 2.2.6 Other Anthropogenic Salts Residential and industrial developments also import or mobilize salts in the basin, some of which may add to the salt-loading of the basin through fallout of polluted air and disposal of waste into the basin waters. However, even at high levels of settlement and industrial development, this usually remains a minor source. The harm generally derives less from the salt quantity than from the types of salts, some of which may be highly toxic to public health and the environment. 2.3 Salt Regimes The global occurrence of soil and water salinity has distinct zonal characteristics (Kovda 1973). While salinity commonly occurrs in the arid zone, it is generally not found in the humid zones where rainfall exceeds the evaporation and salts are readily leached from the land and diluted in the water bodies. As shown in table 1, there is a close relationship between the rainfall/ evaporation ratio and the distribution of salinity within the arid zone. The salts in arid landscapes tend to migrate, moving with the water or with the eroded material, to the geomorphological bottom areas, typically occupied by the plains that, because of their easy command, flat topography, and soil suitability, are precisely the areas favored for irrigation development (Smedema 1990). The salinity conditions in these bottom plains, however, are seldom uniform, as geological and leaching/drainage conditions vary in space and may also have varied over time. Moreover, various geochemical processes may also have influenced the salt concentration, distribution, and composition. 6 The resident salt storage in the plains of the arid zone may be enormous. A profile of 50–100 m depth in the Indus Plains may be calculated to easily store salt in the region of 1,000–2,000 tons/ha. This value compares to 250 tons deposited by 100 years of irrigation with water of 250 mg/l and an annual applied water depth of 1,000 mm. Under natural conditions, these buried fossil salts do little harm. However, these resident salts do become a major burden when they are mobilized and become part of the salt dynamics of the basins. As will be discussed in more detail later, rising water tables, induced seepage flows, tube-well pumping and other features of large-scale irrigation development have, in many instances, led to such mobilization. 2.4 Safe Limits The damage caused by high levels of river salinity has been studied in considerable detail for the Colorado River (USBR 1988). The main findings are summarized below: 2.4.1 Agricultural Use Reductions of crop yields due to the use of saline irrigation water were identified as the main form of agricultural damage. It was recognized that the use of such water also incurs other damage and costs (restrictions on crop choice, need for using more costly irrigation practices, need for improved drainage, etc.). However, many of these adjustments appeared to be at least partly motivated by external considerations and the costs directly attributable to salinity could not be generally quantified. Damage to other agricultural water use (drinking water for animals and other incidental on-farm usages) was found to be negligibly small. The yield reductions for most common crops due to salinity have been extensively researched and documented. Most crops have fairly distinct threshold levels below which the damage is minimal while above that level yields fall almost linearly with increasing levels of salinity. For sensitive crops, yields may become affected when the salinity of the applied water exceeds 500 mg/l while for tolerant crops this may not occur until it reaches levels of 3,000– 4,000 mg/l. According to FAO (1985), the following general criteria apply: no yield reduction Table 1. Climate-salinity relationships in Eurasia. Climatic Mean Annual Annual Residential Maximum Maximum Maximum Salinity in landscape annual rainfall potential salinity of river lake ground- top horizons temperature (mm) evapotrans- sedimentary water water water of solonchak (0C) piration rocks salinity salinity salinity soils (mm) (g/l) (g/l) (g/l) (%) Desert 15–18 80–100 2,000–2,500 Common 20–90 350–400 200–350 25–75 Semidesert 10–12 200–300 1,000–1,500 Frequent 10–30 300–250 100–150 5–8 Steppe 5–10 300–450 800–1,000 Rare 3–7 100–250 50–100 2–3 Forest steppe 3–5 350–500 500–800 None <1 <100 <3 <1 Source: Kovda 1973. 7 for EC < 0.7 dS/m (450 mg/l), slight yield reduction for EC = 0.7–3.0 dS/m (450–2,000 mg/l), and severe yield reduction for EC > 3.0 dS/m (2,000 mg/l). Crop growth may also be affected by high concentrations of specific salts or ions, notably by excess sodium, chloride, bicarbonate, boron, and various trace elements and heavy metals but river water seldom reaches these toxic levels (for details see FAO 1985). 2.4.2 Municipal and Industrial (M&I) Use The World Meteorological Organisation’s (WMO’s) safe limit for human drinking water is around 1,500 mg/l but this limit is of little relevance, as water becomes unpalatable to most users at much lower salinity (usually taken at 500 mg/l). Users typically respond to taste deterioration by purchasing more bottled water. Switching to alternative supplies (shifting the intake farther upstream, changing from surface water to groundwater, etc.) is usually a more feasible option than treatment. Salinity and the related hardness of the water also affect various other household usages, even at very low concentrations. Lifetimes of some sensitive household appliances and systems may be affected when salinity rises above 100 mg/l. A number of industrial systems and processes are highly vulnerable to salinity. Corrosion substantially shortens the lifetimes of metal pipe systems, especially of cooling systems, when salinity levels of the water rise above 100 mg/l. This also applies to various treatments and distribution facilities operated by water supply and wastewater utilities. Most food processing industries can tolerate salinity levels of up to 500 mg/l but some other industrial processes, e.g., the paper industry, require much higher standards (< 100 mg/l). 2.4.3 Environmental Use Damage to the environment, other than that already implicitly covered by the previous categories, was not identified in the Colorado study. Some specific environmental damage such as loss in aquatic and riverine habitat and scenery, loss of biodiversity, loss of assimilative capacity, etc., did occur but this damage was judged to be more due to the discharge reduction than to the increased salinity of the river. 8 9 CHAPTER 3 IRRIGATION DEVELOPMENT This chapter briefly describes how irrigation development impacts on the natural hydrological and salinity regimes of the basin. The land salinization processes involved and the related water management issues are also discussed. 3.1 Water Resources Development Early irrigation development in river basins was generally based on the run-of-river diversion but now the river flow in most basins is partly or fully regulated through the construction of reservoir dams and other storage facilities. These regulation works have generally drastically changed the natural flow regimes of the rivers. Depending on the installed storage capacity, previous short-duration variations in flow may hardly be noticeable, although the impacts of annual variations in weather conditions and rainfall may still be quite pronounced. The buffering impact of the reservoirs is even greater on the salinity. The general experience is that salts mix rapidly and completely in the reservoirs and depending on the size of the reservoir relative to the river flow, smooth out most of the seasonal salinity variations. Large reservoirs generally moderate the high salinity levels during the low-flow seasons as well as the lower salinity levels during high-flow seasons (figures 2a and 2b). The general experience is also that the salt precipitation in the reservoirs or elsewhere in the river system is so small that it may almost always be neglected in basin-level salt balance calculations. Water abstractions have greatly reduced the downstream river flows in many irrigated basins in the arid zone. In some of these basins, the low-flow discharges to the sea have been reduced to a trickle while in the more heavily regulated basins, this has almost become a year- round feature (Williams 1987). The Yellow River in China reportedly had no outflow to the sea during a period of 136 days in 1996. The Colorado River has had no significant flow reaching the sea for the last 10–20 years and in most years the flow in this river ceases well upstream of its mouth. The lower reach of the Jordan River has deteriorated into a minor drain. These dramatic changes in the downstream flow regime have greatly affected the ecological functioning of the river and the ecology of the channel beds and floodplains. The institutional, organizational, legal, and financial arrangements for the development and the management of the water resources in the basins and irrigation schemes are usually very country-specific. In most federally organized countries, the water rights rest with the States and water resources development and management are primarily a State responsibility. The federal governments, however, have often used their contribution in the costs to acquire some regulatory influence while the States may also have delegated some activities to specialized 10 Figure 2a. Impact of flow regulation on river salinity: Regime before regulation (1941) (Colorado River, Lee’s Ferry, USBR 1977). Figure 2b. Impact of flow regulation on river salinity: Regime after regulation (1993) (Colorado River, Lee’s Ferry, USBR 1977). 11 federal institutions. This applies specially to basins spreading over more than one State. In almost all countries, central governments have also assumed a leading role in environmental matters related to water resources development and management. 3.2 Irrigation-Induced Salinity Irrigation development in river basins in the arid zone may activate one or more of the following processes, which, as will be described later, are likely to lead to increased levels of downstream river salinity. 3.2.1 Land Salinization Irrigation development in the arid zone almost inevitably leads to increased recharge to the groundwater (not only due to deep percolation of irrigation water losses but also due to less retention/more recharge by rainfall). Where the increased groundwater recharge exceeds the natural drainage capacity of the aquifer system, water tables will rise until a new equilibrium between recharge and discharge is established. In natural imperfectly drained land, this new equilibrium may not be reached until the water table has risen into the root zone and the land has become waterlogged. When the groundwater is saline (as is typically the case in the arid zone), this waterlogging will also lead to salinization of the land. This twin problem of waterlogging and land salinization is of widespread occurrence in the arid zone, seriously affecting the productivity of the irrigated land in this zone (Ghassemi, Jakeman, and Nix 1995). The problem can be combated effectively, and already affected land can be reclaimed by the development of improved drainage (deepening and densification of the existing open drainage systems, installation of subsurface drainage systems, assuring adequate maintenance, etc.). However, since the drainage effluent will generally be quite saline, drainage development may be constrained by the lack of acceptable disposal sites (see discussion in section 3.3). The disposal problem may be reduced by judiciously matching the reclamation program with the flow regime of the receiving river. 3.2.2 Drainage Water Salinity When river water is applied to the land for irrigation, crops will take up only a very small fraction of the applied salts. As a result, the salts will accumulate in the root zone unless there is a net downward leaching/drainage flow to remove the salts from the soil and the land. In the long-term, the root zone salinity of irrigated land in the arid zone will stabilize at a level where there is an equilibrium between the salt influx (by the irrigation water) and the salt outflux (by the leaching/drainage water). This equilibrium level is largely determined by the salt concentration of the applied irrigation water and by the leaching fraction, which is the leaching/ drainage flow expressed as a fraction of the applied irrigation water. Under equilibrium salt balance conditions, there is a pronounced inverse relationship between the salinity of leaching/drainage water and the irrigation efficiency (which is related to the leaching fraction), i.e., the leaching/drainage water becomes more concentrated with 12 increasing irrigation efficiency and less concentrated with decreasing irrigation efficiency. Model calculations predict that under highly efficient irrigation with a leaching fraction of 10 percent, the leaching/drainage water may be expected to be about 10 times as concentrated as the applied irrigation water while for a leaching fraction of 20 percent, this concentration would be about 5 times (FAO 1985). In practice, the relationships between the salinity of the leaching/ drainage water and the irrigation efficiency may not be always as predicted. Equilibrium conditions may not exist or the relationships may be disturbed by the leaching/drainage water having picked up resident or other salts, being (partly) generated by rainfall or containing significant other flow components (surface irrigation waste, drainage flows from nonirrigated land, etc.). 3.2.3 Salt Mobilization Irrigation development in the arid zone introduces a new source of water, which may dramatically change the prevailing geohydrological flow regimes. Additional groundwater recharge, fed by the deep percolating irrigation losses, may build up groundwater bodies and induce previously nonexisting groundwater flows. As explained earlier, areas chosen for irrigation development in the arid zone are likely to have a rather high resident salinity. As illustrated in figure 3, these new groundwater flows may load the rivers with large quantities of previously harmless, deeply stored resident salts. Figure 3. Mobilization of primary and fossil salts by irrigation-induced groundwater flows. 13 In some cases, drainage improvements may inadvertently enhance this salt mobilization. This, for example, is often the case when vertical (tube well) type of subsurface drainage is installed. As groundwater salinity typically increases with depth, skimming types of drainage technologies (like horizontal pipe drainage), which draw most of its water from the upper groundwater zones, generally mobilize less-resident salts. In areas naturally subject to upward saline seepage (such as the northern part of the Nile Delta), improved drainage, which lowers the phreatic drainage base, will increase this seepage and the salinity of the drainage water. 3.3 Disposal of Saline Drainage Water As discussed above, drainage water from irrigated land in the arid zone is likely to have a fairly high salinity that may pose problems for its disposal, especially when there is no ready access to natural or otherwise acceptable salt sinks (seas, salt lakes, etc.). For such schemes, one of the following modes of disposal or disposal management options may be considered. 3.3.1 Down the River This is the natural mode of disposal, which was traditionally used in almost all basins until the limits of the downstream salinity were reached. In some cases, reaching these limits can be prevented by enhancing the river flow during critical low flow periods, e.g., by changing reservoir operation rules, by limiting the upstream water diversions, by reducing the saline water disposal during low-flow periods or by otherwise adjusting the salinity disposal to the dilution capacity of the receiving river. 3.3.2 Evaporation Ponds Evaporation ponds are widely applied throughout the arid zone. Typically, these are natural depressions in the landscape towards which the drainage water can be easily directed. They are usually located in desert areas outside the irrigated perimeters, either sideways or at the lower end of the irrigation systems. However, small constructed evaporation ponds, e.g., serving individual farms, may also be found within irrigated schemes. 3.3.3 Limiting the Saline Effluent Generation Although salt balances need to be maintained, some options for limiting the disposal flow are generally available. Reuse of drainage water is such an option, although it is clearly not a long-term solution as salts are being stored somewhere in the basin and limits will eventually be reached. The same applies to not meeting the leaching/drainage needs. Improving irrigation efficiencies helps by leaving more water in the river and also by reducing the drainage volumes but as the latter will have a higher salinity (see discussion in section 3.2), this will generally offer only limited relief to the downstream salinity. 14 3.3.4 Limiting the Salt Mobilization This is a highly effective and desirable measure with no negative side effects. Ideally, the fossil and other resident salts stored in the basin should not be mobilized, but as irrigation development almost inevitably leads to rises in the water table and generation of new piezometric gradients and groundwater flows, this mobilization cannot be always avoided. Judicious irrigation development planning can help. In some cases, the mobilized salts may also be prevented from reaching the river by the installation of interception drains. 3.3.5 Land Use Planning Land use planning in the basin can help in various ways to limit river water abstraction, generation of saline drainage flows, and salt mobilization or otherwise help control downstream river salinity. Less-water-demanding crops may be grown, irrigated land with uncontrollably high saline drainage rates may be retired or converted to rain-fed cropping, land with a high salinity may be left unreclaimed, and unproductive depressions may be designated as salt sinks (the “dry drainage” solution). 3.3.6 Outfall Drains The construction of special drains for the collection and transport of saline effluent to a natural salt sink (usually the sea) may ultimately be required for basins that have no ready access to such sinks. Temporary solutions may be appropriate as intermediate steps but, as indicated above, most of these solutions have a limited capacity and do not fully maintain or restore the salt balance and, therefore, do not assure the long-term sustainability of irrigated agriculture in the basin. 3.3.7 Other Options Disposal by means of bores to shallow saline aquifers is reportedly practiced in the southern part of the Murray Basin (Ghassemi, Jakeman, and Nix 1995). The salinity of the effluent is typically 1,000–2,000 mg/l and that of the receiving aquifer 25,000 mg/l and higher. Injection into deep aquifers is under investigation but is generally judged to have limited opportunities (FAO 1997). Desalinization is applied in a special case in the Colorado Basin (see annex B) but wider application will probably become only a serious option when the costs of disposal approach the costs of desalinization. In most basins this point has not yet been reached. 15 CHAPTER 4 THE SELECTED BASINS The global and climatic setting of the selected basins is shown in figure 4. All main continents are represented. Some pertinent general characteristics and special features of the selected basin are given in this chapter. More complete and detailed information on each basin is provided in the attached annexes. 4.1 Brief Characterizations These short descriptions are presented to familiarize the reader, without reading the annexes, with the prevailing salinity conditions, problems, and some other related features of the selected five basins. 4.1.1 Aral Sea Basin The rapid expansion of irrigated area and the related increased abstraction of the river water after the Second World War have greatly disturbed the natural hydrological and salinity regimes and have placed the socioeconomic future of the basin in considerable jeopardy. The greatly reduced river discharges to the Aral Sea have cut its size by half, increased its salinity threefold and have resulted in large-scale environmental degradation of the sea and its surrounding areas. The downstream flow and salinity conditions in the rivers have deteriorated to the extent that basic community and environmental functions are under severe stress and no further non- compensated abstraction and disposal of saline drainage can be allowed. The challenge is to halt further degradation of the Aral Sea and of the river functions, and, if possible, achieve some limited restoration of the past damage, without jeopardizing further socioeconomic development of the basin. 4.1.2 Colorado Basin The experiences with salinity control in this basin offer, in many respects, a “window to the future” for many other basins. The basin features a highly intensive, diversified, and competitive use of the river water, sophisticated institutional and legal arrangements, highly developed stakeholder participation as well as an international dimension as, although most of the basin is in the USA, the river runs to the sea through Mexico. Some of the river salinity control measures instituted are, in fact, motivated by the Colorado water sharing agreement 16 F ig ur e 4. M aj or r iv er b as in s of t he a ri d/ se m ia ri d zo ne ( ad ap te d fr om U N E P 1 99 2) . 17 that stipulates the delivery of a specified volume of water of a specified salinity to Mexico. A unique feature of this basin is also the considerable export of water, for irrigation and M&I use, from the basin to adjoining water-scarce areas. The current export already amounts to some 30 percent of the river flow and is bound to increase through trading with financially strong users outside the basin. 4.1.3 Indus Basin The Indus Basin is such a dominant feature of Pakistan that the basin and the country are almost synonymous. The basin harbors almost all of the country’s important irrigated agriculture sector and most of its major urban and industrial centers. The river water is currently almost exclusively used for irrigation as other water demands are still limited and are mostly met from the groundwater resources. The basin is subject to a massive problem of irrigation-induced waterlogging and land salinization, seriously affecting some 35 percent of the irrigated lands. The import of salt with the irrigation water far exceeds the disposal while the mobilization of fossil salts from deeper strata further adds to land salinization problems. Increased salt disposal from the basin is of strategic importance to the survival of irrigated agriculture in the basin. Steps are being taken to achieve this by investing in improved land drainage, by constructing outfall drains to the sea, and by instituting reforms meant both to promote more efficient water use and to assure better maintenance of the drainage systems. 4.1.4 Murray-Darling Basin Current salinity conditions and driving forces in this basin are broadly comparable to those in the Colorado Basin. One similarity is that both basins spread over several States and are located in countries with strongly developed federated governmental structures under which water affairs are primarily handled by the States. Although the current river salinity is still mostly at an acceptable level, the noticeably increasing trends have raised acute concerns amongst the water users as well as in the civil society at large. Another striking similarity with the Colorado Basin is in the technical measures taken to control downstream river salinity (land use planning, interception of saline seepage flows, restrictions on the down-the-river disposal of saline drainage water, etc.) but there are also significant differences in the political approaches, in the implementation arrangements, and in the sharing of the costs. 4.1.5 Nile Basin Egypt has long since embarked on improved drainage of its irrigated land, which has, by and large, brought land salinization in the basin under control. River salinity would not be a problem if the drainage water could be allowed, as done in the past, to discharge to the sea. Available fresh river flows are becoming increasingly insufficient to meet the water requirements of the ever-expanding area of irrigated land in the adjoining deserts and of the growing urban and industrial sectors; and the new challenge is to reuse drainage water to the maximum extent possible without harmfully salinizing the land and water resources. Although salinity problems in this basin are essentially confined to the Egyptian part of the basin, since salinity cannot 18 be separated from water availability, the water requirements of the upstream countries from where the river flow originates should also be considered. 4.2 Land and Water Resources In all five basins, only the bottom parts of the basin (the “plains” in which most of the irrigated areas are sited) are properly located in the arid/semiarid zone while the main catchments are mostly outside this zone, typically in areas with higher rainfall and many mountains. The annual rainfall in the plains varies from <100 mm to 500 mm in most basins and is somewhat higher in the upper basin than in the lower basin. Most basins, at least in the upper catchments, also have defined rainy and dry seasons. The five basins studied share these features with almost all other major irrigated basins in the arid zone. Some other items of information on the land and water resources of the studied basins are presented in table 2. The “size” data in this table refer to the total watershed of the basin while the river flow data refer to long-term average flows passing through the “rim stations” and the final outfall points to the sea. The rim stations are the points where the main river/ tributaries leave the catchment and enter the plains; most basins have such stations, typically coinciding with existing or potential dam sites. Table 2. Land and water resources of the selected basins. Basin Annual Flow Size At the At the Reservoir Irrigated Water rim outlet to capacity area use for stations the sea irrigation (M ha) (km 3 ) (km 3 ) (km 3 ) (M ha) (%) Aral Sea (Amu Darya and Syr Darya) 180 116.0 11.5 55 7.5 90 Colorado 63 18.5 Nil 76 2.0* 80 Indus 94 181.0 39.0 19 16.2 95 Murray-Darling 106 24.3 5.5 30 1.5 95 Nile** 296 55.5 13.3 130 3.1 85 *Including 0.6 million hectares (M ha) outside the basin and 0.2 M ha in Mexico. **Value for the size is for the entire basin but all oth er values are for Egypt only. The reservoir capacity in the table refers to the total storage. The Colorado River is clearly a highly regulated river with a constructed storage capacity of more than four times the annual river flow. The Nile and Murray-Darling basins also already have considerable storage capacities while the Aral Sea Basin and, most notably the Indus Basin, are relatively less-developed. As mentioned earlier, in the Lower Colorado River, the seasonal variations in salinity (lower during the high-flow season and higher during the low-flow season) have largely disappeared. However, this is far less the case in the Indus and Aral Sea basins, which have considerably less reservoir storage. However, the storage in all the basins is sufficient both to enhance river flows during periods with extreme low inflows and to dilute associated high- salinity levels. 19 The differences in the use of the river flow for irrigation (last column of table 2) reflect the relative importance of the irrigation versus the municipal and industrial (M&I) and other uses of the river water. The M&I sector is of considerable importance in the Colorado and of growing significance in the Nile Basin while in the other three basins, the irrigation sector is still using almost all the water. However, even in the Colorado, some 80 percent of the water goes to irrigation indicating that even in highly developed basins, the quantitative demands of these other users remain limited. The “flow to the sea” figures confirm the previously mentioned high rates of river flow depletion. Abstraction varies from 100 percent for the Colorado to about 90 percent for the Aral Sea, about 80 percent for the Indus, and some 75 percent for the Nile and the Murray- Darling. The absolute outflows to the sea are quite small in all the basins (nil for the Colorado and quite minimal in most years for the Syr Darya), except for the Indus River where, due to limited storage, there is still a large discharge to the sea during the rainy season. For navigation purposes and prevention of salt intrusion, a fairly high outflow to the sea is also maintained in the Nile River. The discharges of the Amu Darya and Syr Darya rivers to the Aral Sea have fallen well below the minimal requirements for the physical survival of the Aral Sea, as well as for the ecological well-being of the lower reaches of the rivers. The latter applies even more to the lower reach of the Colorado River. 4.3 Institutional Arrangements The institutional arrangements for the development and management of the water resources in the five basins are all quite different. In four of the basins (Aral Sea, Colorado, Indus, and Nile), the arrangements are complicated because these rivers and the sea cross international boundaries while three basins (Colorado, Indus, and Murray-Darling) face comparable problems within the country as they serve the interests of more than one province/State. In three of the international basins (Aral Sea, Indus, and Nile), most of the headwaters are outside the countries in which most of the water is used. Although some agreements have already been made, conflicts of interests remain. As the water in the headwaters is, in all cases, of good quality, cross boundary salt fluxes are generally not a major issue. The existing flow- sharing agreements, however, have an important bearing on the downstream salinity as these agreements determine the availability of upstream dilution of water for downstream river salinity control. In the fourth international basin (Colorado), the headwaters and most of the uses are in one country (USA) but the tail end is in another country (Mexico). Here the current updated agreement covers both the quantity and the salinity of Mexico’s entitlement. Most of the involved countries have a pronounced federated structure as far as water resources development and management are concerned. In the USA (Colorado), Pakistan (Indus), and Australia (Murray-Darling), the authority over the water resources rests with the provinces/States and the federal governments have, principally, only a regulatory role. For pursuing common interests, solving entitlement conflicts, and implementing cross-boundary projects, various types of consultative, cooperative, and executive structures have emerged, often facilitated by the federal government. In most basins, the cross boundary planning and implementation tasks have also been delegated to inter-provincial/inter-State or federal bodies and programs. Some of these bodies and programs cover both water quantity and quality issues but with others these are treated 20 separately. Stakeholders’ participation may have been treated as an integrated part of the water resources development and management processes or function as parallel processes. The arrangements are generally quite specific for each basin and further details are given in the annexes. 21 CHAPTER 5 SALT BALANCES For the five basins, reasonably reliable data are available for discharge and salinity of inflows at the rim stations and also for outflows to the sea or other disposal sites. Using these data and calculating the annual salt fluxes as “annual discharge volume x weighted average salinity,” annual salt balances for the basins have been calculated following the methodology presented schematically in figure 5. As shown, four modes of salt disposal have been recognized: a) down the river to the lower basin (applies to upper basins only); b) to the terminal sea or other natural disposal site; c) to evaporation ponds (E-ponds); and d) export of salt outside the basin. Where the required data were available, the basin balances have also been broken down into separate balances for the upper and lower basins. 5.1 Methodology In the applied calculation methodology, the resident salt storage covers, in principle, the storage in both the river system and the land system (soil and underlying groundwater). However, it has generally been assumed that the salt storage of the river system is negligible compared to that of the land and that all of the calculated storage changes are changes in the land storage. This assumption needs further checking as it may not be valid for all basins (more detailed salt balance calculations for the Amu Darya and the Syr Darya basins show considerable differences between the summation of the salt balances of the separate irrigation schemes and the salt balance of the entire basin, and these differences are thought to be due to salt generation/removal processes in the river system). In interpreting the calculated salinity balances, it should be clearly understood that the basin is treated as a “black box” and that the balances give the net change in resident salt storage in the (sub) basin. A gain in storage only means that, over the considered area and time period, on balance salt is being accumulated. However, there may be salt losses in part of the area or for part of the time. The balance calculations also do not allow differentiating between the various internal salt mobilization and immobilization processes. Very little is known about the quantitative dimensions of these internal processes. As discussed in section 2.2, the balance between mineral weathering and precipitation is likely to be negative while that between fertilizer application and crop uptake is likely to be neutral or slightly positive. Following common practice in salt balance calculations, it is assumed that these negative and positive balances cancel one another (Rhoades 1974). Even when this assumption is not entirely correct, the error on the basin salt balance would be quite small (best estimates suggest the annual net balance of these four processes, positive or negative, to be about some hundreds 22 of kilograms of salt per hectare, which is, as will be shown in the next section, at least ten times less than the recognized salt fluxes). Accepting the above cancel-out assumption and ignoring the other minor salt sources identified in section 2.2, a positive basin salt balance may be interpreted as indicating that salts imported with the irrigation water are accumulating in the basin. Similarly, a negative balance may be interpreted as indicating that resident salts are being mobilized and disposed of to the river system. The calculation method does not allow for differentiating between the origin of the latter salts, which may be any combination of originally imported salts and fossil, and primary or other types of resident salts. A negative balance indicates only that more salts are being disposed of to the river than are being diverted from the river, meaning that some mobilization and removal of resident salt are taking place. 5.2 Patterns and Features The calculated salt balances (table 3) are obviously all quite different, which is not surprising in view of the different geological and physical characteristics and irrigation development histories of the involved basins. In four of the basins, the salt balance is negative meaning that these basins are desalinizing. In the Aral Sea, Colorado, and Murray-Darling, the salts being removed are mostly fossil and primary salts mobilized by irrigation developments in the middle parts of the basins. The geological age and origin of the salts and the prevailing mobilization mechanisms are however quite different for each of these three basins (see the detailed descriptions given in the annexes). In the case of the Nile, the mobilized salts are almost all marine salts of quartenary age from the deltaic lower part of the basin while the mobilization is due to improved land drainage rather than to irrigation development. Figure 5. Salt balance components. 23 The Indus is the most distinctly different case in that a large part of the salts carried by the irrigation water remains in the land, especially in the upper part of the basin. Both the Nile and the Indus basins are extensive alluvial plains; however, the drainage conditions in terms of both natural and constructed drainage are much more favorable in the Nile Basin and this would seem to explain most of the contrastingly different salt regimes. Some specific patterns and features of the salt balance of each of the five basins are highlighted below: 5.2.1 Aral Sea In both the Amu Darya and the Syr Darya basins, there is considerable irrigation-induced mobilization of resident salts. A more detailed analysis, breaking down the basin in the upper, middle, and lower parts, shows that this mobilization occurs particularly in the middle part of the basin where, in the seventies, several pumped irrigation schemes were developed on an elevated plateau and these schemes generate highly saline seepage flows. The salt mobilization of some of these schemes is as high as 10 tons of salt/ha/yr. The high salt influxes at the rim stations indicate that considerable salt-loading of the rivers already occurs in the catchments above the rim stations. As explained in annex A, most of the headwaters originating in the high mountain ranges are of low salinity, so considerable salt pickup must occur between the headwaters and the rim stations. This may be due to the tributaries passing through areas with saline geology. Table 3. Salt balances in million tons per year (Mt/yr.). Disposal Resident storage (below rim stations) Basin Input Lower Sea E-pond Exported + - (Increase) (Decrease) Aral Sea - Amu Darya 36.4 8.4 25.3 9.1 6.4 - Syr Darya 17.4 4.6 18.7 5.9 - Total 53.8 13.0 44.0 9.1 12.3 Colorado - USA 3.7 1.8 7.4* 5.5* - Mexico 1.8 1.8 - Total 3.7 7.4* 1.8 5.5* Indus - Upper basin (Punjab) 33.0 19.0 2.2 11.8 - Lower basin (Sindh) 19.0 16.4 2.6 - Total 33.0 16.4 2.2 14.4 Murray-Darling - Upper basin (Riverine Plain) 1.1 3.2 0.2 2.3 - Lower basin (Mallee) 3.9** 5.5 0.1 1.7 - Total 1.8** 5.5 0.3 4.0 Nile - Upper basin (Valley) 10.9 12.6 0.7 2.4 - Lower basin (Delta) 12.6 34.1 21.5 - Total 10.9 34.1 0.7 23.9 *Calculated on the basis of the given total river-loading of 9.2 Mt/yr. (see annex B). **Including 0.7 Mt/yr. brought in by the Darling River (see annex D). 24 Another remarkable feature of this basin is the high proportion of drainage water disposal to evaporation ponds (E-ponds). As many of the irrigation schemes are distant from the river, disposal to E-ponds rather than to the river makes sense as it is generally less costly and does not burden the river with salts. Suitable nearby depression sites were found in the desert adjoining the irrigation schemes. Although in the planning and management of these ponds little consideration has been given to environmental factors, to date no significant adverse environmental impacts in and around these ponds have been reported. A concern is, however, that many of the E-ponds used have reached their maximum capacity (see annex A). As few new sites are available, studies are now undertaken on how the drainage discharges to these ponds can be reduced, e.g., by using the drainage water to irrigate salt-tolerant forestry schemes that concentrate the drainage water and reduce the volumes to be disposed of. 5.2.2 Colorado The current total salt-loading of the river is reported to be about 9.2 Mt/yr. of which, as shown in table 3, some 3.7 Mt/yr. occur above and 5.5 Mt/yr. below the rim stations. Of this total loading, about 5.2 Mt/yr. are due to natural mechanisms and 3.4 Mt/yr. are irrigation-induced (see annex B). A more detailed balance indicates that all the net mobilization of resident salts occurs in the upper basin. Most of the salt mobilization is by seepage flow passing though saline-shale strata, which underlie much of the land in the upper basin. Some of these flows are of natural origin but many are irrigation-induced. These saline seepage flows are the main cause of the salt-loading of the Colorado River and various control measures are being taken to reduce this loading (canal lining, improvement of on-farm water management, land use planning, interception drainage, etc.). According to the detailed balance, more salt enters the lower basin than is disposed of by the four modes of disposal identified. Salt is being exported to Southern California but this disposal is more or less proportional to the water exported. All of the water entering Mexico is evaporated or otherwise consumed (there is no outflow to the sea). Some salt accumulation seems to occur in this part of the basin but the available information does not allow further pinpointing. 5.2.3 Indus For the Indus, the balance is strongly positive with 16.6 Mt of more salt entering the rim stations annually than the amount being disposed of to the sea. Of this surplus, some 2.2 Mt/yr. are disposed of into E-ponds while the remaining 14.4 Mt/yr. are being added to the salt stored in the land (partly in the soil and partly in the underlying groundwater). This amounts to an annual increase of salt storage of about 900 kilograms of salt per irrigated hectare. This does not include the mobilization of salt by tube-well irrigation as this is an internal shift in the salt storage of the land, which is not registered in the applied balance calculations (shift of salts from deeper down to the surface layers). However, this internal mobilization, which is estimated to be about the same as the calculated external salt-loading of the land, contributes considerably to the salinization of the land as, in both cases, the salts mostly accumulate in the upper soil and shallow groundwater layers. 25 As mentioned earlier, most of this salt accumulation is thought to be due to underdevelopment of the drainage system in the Indus Plain. This is a vast plain in the arid zone, which did not need and did not have a well-developed arterial drainage system in the natural conditions. When irrigation was introduced, only a few additional drains were constructed and the present drainage density in most of the plain does not exceed 5–10 m/ha (compared to a drainage density of at least 50 m/ha in the Nile Plain in Egypt). Moreover, injudiciously constructed roads block many of the natural drainage ways and other infrastructure while whatever little drainage exists is seldom properly maintained. Such a system does provide for sufficient drainage to balance the incoming salts. 5.2.4 Murray Here both the upper and the lower basins are disposing of more salt than are entering and in both parts of the basin the salt loss seems to be a mixture of salt generated by improved drainage and the reclamation of salinized irrigated land and fossil salts from the deeper strata mobilized by natural or irrigation-induced seepage flows. The quantities involved are much lower than in most of the other basins but still cause serious salinization of the downstream river flows. If the rising trends are allowed to continue, the water supply of some high-value irrigated agriculture and eventually even the water supply of the city of Adelaide may become affected. The parties involved have recently adopted an elaborate program for the control of the river salinity at its current levels (see annex D). 5.2.5 Nile The salt balance and underlying salt regime for the Nile Basin is diametrically opposite to that in the Indus Basin. In the Nile Basin, each year some 23.9 Mt of more salt are removed from the basin than are entering and as shown in table 3, very little of this salt originates from the upper basin (the Nile Valley). The alluvial deposits of the valley have been thoroughly desalinized down to a great depth by the long history of flooding and leaching and the salt removal that still occurs is thought to be mostly made up of leached remnants of high fertilizer application and by the salts brought in by the saline seepage from the newly developed irrigation schemes on the elevated desert land along the edges of the Valley. The desalinization of the marine deposits in the lower basin (the Delta) is far less advanced. Most of these deposits are of much younger origin and may well have been deposited in a much more marine environment than the valley deposits. While in the south, near Cairo, a fairly deep desalinized upper layer seems to have been already formed, most land in the northern zone along the Mediterranean Sea remains heavily saline up to just below the root zone (see annex E). These marine salts are being mobilized continuously by the subsurface drainage flows created by the recently installed land drainage systems and by the natural upward seepage flows generated by the prevailing piezometric head differences with the sea level. Ironically, this upward saline flow is enhanced by the improved land drainage. As most of these salts end up in the drainage systems, most of the drainage water in the northern delta zone is highly saline, with salinity values often about 2,000–3,000 mg/l. 26 27 CHAPTER 6 RIVER SALINITY As is to be expected, in view of the inherent relationships between the land and the river systems in a basin, the salt regimes of the rivers reflect, in many respects, those of the land. This is especially true under natural conditions where the river systems are the principal recipients of drainage flows generated from the land in the basin. Irrigation and other developments have severed some of these natural relationships but in most basins they are still sufficiently strong for the rivers to remain a key indicator of the prevailing salinity conditions of the land. These relationships are further explored in this chapter and measures, which can be taken to control the rise of the river salinity, are discussed. 6.1 Natural Salinity Natural river flows are generated by the natural drainage flows in the basin. These flows would normally pass over various land surfaces and through various subsurface formations and, in the process, would pick up salts. The degree of mineralization of the water depends on the pattern of the flow paths and on the geological properties of the formations traversed (Drewer 1988). Although, under normal conditions, the salinity of the headwater streams would generally not exceed 30 mg/l, considerably higher levels may occur in some streams arising from geologically more saline sub-catchments. Most of this variation averages out as the streams combine and the main headwater tributaries may normally be expected to have a weighted average salinity about 40–50 mg/l. The salinity of the diverted irrigation water may be considerably higher than that of the headwaters, depending on the climate and geology of the areas and the distance traversed between the headwaters and the intake points. This distance may be relatively short (< 100– 200 km for the upper catchment irrigation in the Murray and the Colorado Basin) or extremely long (4,000–5,000 km for Nile irrigation in Egypt). Over this reach, the river salinity may increase due to natural pickup, in-stream evaporation, and salt-loading by return flows from upstream- irrigated areas. In the five basins studied, the salinity of water in reservoirs at the rim stations varies about 50–150 mg/l for the Murray Basin (upper catchment reservoirs), 140–150 mg/l for the Indus Basin (Tarbela Reservoir), 150–200 mg/l for the Colorado Basin (upper catchment reservoirs), about 200 mg/l for the Egyptian Nile Basin (Aswan Reservoir) and about 300 mg/ l for the Aral Sea Basin (upper catchment reservoirs). These are the values that typically apply to the irrigation water diverted at the most upstream intakes. For the more downstream intakes, it would generally be higher. 28 6.2 Irrigation-Induced Salinity In all five basins, the development of irrigation has led to an increase in the downstream river salinity. The causes and mechanisms for this increase, which have been discussed earlier, are summarized below: ? Reduction of the dilution capacity of the river system due to abstraction and consumptive use of the river water for irrigation. ? Use of the river system as a means of disposal for saline drainage water generated by regular leaching and drainage of the irrigated lands; this salt-loading by the drainage return flow is extra high when the drainage water is generated by vertical drainage or when salinized land is being reclaimed. ? Loading of the river system by primary, fossil, and other types of resident salts mobilized by irrigation-induced seepage flows, tube-well irrigation, or by some other irrigation- induced mechanisms. In most basins, the observed increases in downstream river salinity are not due to irrigation development alone. Flows are also reduced by abstractions for other water uses although in the five basins these abstractions are much less than the amount used for irrigation. As shown in the relevant annexes, changes in land development, other than for irrigation, have also added to the salt-loading of some rivers. A considerable part of the river salinity in the Murray-Darling River system originates from “dryland salinity” while in the Colorado Basin this applies to runoff and erosion of saline grazing land. In the latter basin, waste disposal from industry and mining also adds to the river salinity. However, although these other contributions should not be ignored, irrigation development is clearly the main factor causing the increased downstream river salinity in all the five basins studied. 6.2.1 Observed River Salinity Profiles On the basis of the most recent data available, normalized longitudinal salinity profiles for the main stems of the river systems in the five basins have been established as shown in figure 6. The upstream ends of these profiles are the rim stations while the downstream ends are the final outfall points (in case of outfalls to a sea, the end point is taken upstream of the salt intrusion reach). All river lengths were set at unit length and stations were located according to their proportional distance from the rim stations. The profiles depict annual average conditions. For highly regulated rivers like the Colorado and the Nile, these average profiles are quite representative but for other basins the salinity values may be considerably lower during the wet season and higher during the dry season (see the seasonal profiles for the Murray River presented in figure D2, annex D). In analyzing these salinity profiles, it should also be realized that these arid zone rivers naturally had quite pronounced salinity profiles (see figure B2, annex B). 29 The salinity profiles presented are largely in conformity with the basin-level salt balances discussed earlier and summarized in table 3. These balances show the change in the resident salt storage in the basin, much of which directly relates to the salt-loading of the river. A decrease in resident storage will generally imply more salt-loading while an increase in the resident salt storage indicates the opposite trend. However, the relationship between change in resident salt storage and river-loading is not necessarily fully reciprocal as some of the lost resident salts may be stored in an E-pond, exported, or otherwise never reach the river. The relationship between the river salinity profile and the resident salt balance is most apparent for the Indus Basin where the minimal increase in the downstream river salinity (from 150 mg/l at the rim stations to 200 mg/l at the Kotri barrage, an increase of only some 50 mg/l) Figure 6. Normalized salinity profiles of the five selected rivers. 30 is clearly due to diverted salts being stored in the land rather than being returned to the river. In the other four basins, increasing downstream salinity reasonably agrees with the calculated salt balances. The high rates of salt mobilization in the middle parts of the Amu Darya and Syr Darya rivers and in the Nile Delta reasonably well match the relatively steep profiles in these parts of the rivers. Differences in salt-loading, however, can only partly explain the observed river salinity profiles as these profiles reflect the combination of salt-loading and flow reduction. In the case of the Colorado River, the pronounced downstream increase in river salinity is probably more due to the high rates of water abstraction than to salt-loading, which is in fact rather moderate (even nil in the lower reach). The differences in the degree of abstraction (100% for the Colorado and 75% for the Murray, see table 2) constitute a plausible explanation for the much more moderate increase in downstream salinity in the latter river. Differences in the degree of abstraction may also largely explain the higher salinity of the Syr Darya River as compared to the Amu Darya River. 6.2.2 Generic Profiles As river salinity profiles reflect the combined impacts of water abstraction and salt-loading and both these influencing factors may vary considerably, it is clear that a wide range of profiles may occur. The profiles given in figure 6 are just a small sample of the expected range. However, it is suggested that the observed variation can be captured, to a large extent, by distinguishing three generic type of profiles, with each type being related to a particular status of the salt balance. These three types are shown in figure 7. Other generic profile types may be conceived (e.g., based on the degrees of water abstraction, on the irrigation efficiencies, and on the percentage of drainage water reuse). However, the distinguished types are considered to be of the greatest conceptual and diagnostic value. Equilibrium Profile. This type of profile is expected to form when the salt balance of the basin is in equilibrium (Input = Output, see figure 5). This would be the case when the following two conditions are fulfilled: ? all the salts diverted from the river with the irrigation water return to the river with the drainage flows ? there is no salt-loading of the river by primary, fossil, or other resident salts already present in the basin Actually, these two conditions do not need to be strictly fulfilled. Since the basin is treated as a ‘black box,’ there could be some influx of primary or fossil salts into the river as long as this is compensated for by an equal amount of diverted salts, which are not returned. Since in the equilibrium case the salt-load of the river system remains constant along the length of the river, the slope of the river salinity profile is determined primarily by the remaining river flow. When the rates of both abstraction and drainage return flow are constant along the length of the river and the return flow is a fixed percentage of the abstraction, the profile would theoretically be of an exponential form (see figure 8). As, in most basins, most of 31 the abstraction occurs in the upper reaches of the river while drainage return flows occur in the lower reaches, the coefficient of exponentiality may be generally expected not to be constant along the length of the river but expected to increase in the downstream direction. Salinization Profile. This type of profile applies to cases where not all of the diverted salts return to the river but partly remain in the irrigated lands. The salt balance in this case is positive (Input > Output) and the salt-load carried by the river decreases in the downstream direction. As shown in figure 7, this profile has not only a smaller slope than the equilibrium profile but also lower river salinity values all along the length of the river. When there is no return of the diverted salts and no influx of any other type of salts, the profile would be a horizontal line with a salinity equal to the river salinity at the rim station. The full (100%) return of salts is the equilibrium profile. There can be various in-between profiles depending on the ratio of abstraction to drainage return flow and on the salinity ratio of the abstracted and drainage returns. Mobilization Profile. This type of profile represents the negative salt balance (Input < Output), which occurs when fossil, primary, or other resident salts in the basin are being mobilized and are entering the river system. As the current study confirms, this is a very common case for irrigated basins in the arid zone, making the mobilization profile the most common of the three generic profiles. The factors that determine the shape of the mobilization profile are partly the same as for the other two profiles, i.e., the abstraction/return flow ratio and the irrigation/drainage water salinity ratio. The specific factor, in this case, is the degree of surplus loading of the river system. This surplus causes the river salinity of the mobilization profile to be higher than that of the equilibrium profile and the total salt-load of the river to increase in the downstream direction. Reaches with large influx of mobilized salts as the middle reaches of the Amu Darya Figure 7. Generic river salinity profiles. 32 Figure 8. Computed river salinity profiles. 33 and Syr Darya rivers and the Delta reach of the Nile, show up as steeply rising sections of the profile (see figure 6). 6.2.3 Computed Profiles River salinity profiles may also be computed when the involved systems and mechanisms are somewhat schematized. An example is presented in figure 8 where both the rate of abstraction and drainage return flow have been assumed to be constant along the length of the river. Under these conditions, an analytical formula for the equilibrium profile can be worked out, which indeed confirms, as suggested earlier, this type of profile to be of the exponential form. The shape of the profile is defined by the two parameters: ? ?= the return ratio (drainage return as a ratio of the abstraction) and ? = the abstraction ratio (abstraction as a ratio of the flow at the head of the river). The computed increase of the river salinity (concentration/concentration at river [C/Co]) over the leng th of the river (l=l thus l/l=1) for a range of abstraction and return ratios is given in table 4. As is to be expected, the downstream salinity rises sharply with increasing abstraction and decreasing return ratios. For a typical flow abstraction of 60–70 percent (? = 0.6–0.7) and 20–30 percent drainage return (? = 0.2–03), the salinity at the tail end of a river is predicted to be about twice as high as it was at the head. So, for salinity at the rim stations of about 150– 200 mg/l, the tail-end salinity may be expected to be about 300–400 mg/l. The profiles in figure 6 are either mobilization or salinization profiles, thus not representing the equilibrium case for which the computations apply. However, since the increase in salinity over the length of the Table 4. Calculated river salinity increase for equilibrium profiles (C/Co). ? =0.5 ? =0.4 ? =0.3 ? =0.2 ? =0. ?=0.5 1.3 1.4 1.5 1.7 1.8 ?=0.6 1.4 1.6 1.7 1.9 2.2 ?=0.7 1.5 1.7 2.0 2.3 2.7 ?=0.8 1.7 1.9 2.3 2.8 3.6 ?=0.9 1.8 2.2 2.7 3.6 5.3 ? = abstraction ratio; ? = return ratio. river in the former profiles is generally more while in the latter it is less than twofold, these calculations would seem to be of the right order of magnitude. The analytical formula developed, being based on the maintenance of the salt balance, takes into account that the salinity of the drainage return flow increases when the return ratio decreases. This explains why the values in table 4 increase horizontally, i.e., when the abstraction ratio is kept constant and the return ratio decreases. As the irrigation efficiency is broadly inversely related to the return ratio (return ratios generally decrease as irrigation efficiencies increase), the formula implicitly also expresses the impact of irrigation efficiency on the river salinity. However, efficient irrigation generally not only reduces drainage return ratios but also 34 lowers the abstraction ratios (more water is left in the river, as less water is needed to irrigate the same area). This water-saving impact of efficient irrigation is not accounted for in the formula but may be recognized by reading table 4 for decreasing return ratios (= increasing efficiencies), not horizontally but diagonally upwards at some arbitrary angle (reading the table this way, abstraction ratios decrease with decreasing return ratios/increasing irrigation efficiencies and the increase in river salinity becomes less than when reading horizontally). 35 CHAPTER 7 CONTROL MEASURES In all five basins, increasing levels of downstream river salinity are of considerable concern and various measures are being taken to halt or reverse these trends. The adopted programs cover a wide range but are mostly a combination of the measures described in section 3.3. The institutional frameworks under which these measures are planned, implemented, and financed are generally unique to each of the basins. For a full discussion of the adopted control programs and their organizational arrangement, reference is made to the case material in the annexes; the discussion here will be limited to pointing out general features, highlighting some innovative approaches. 7.1 Management of the Drainage Return Flows Ideally, drainage return flows from irrigated land should only return the salts originally diverted from the river. Since, under salt balance conditions, the salinity of the drainage water is inversely proportional to the flow rate (see discussion in section 3.2), a high drainage rate would generally be expected to lower the increase in river salinity. In practice, however, this may not always be true as high return flows may load the river unnecessarily with mobilized resident salts. High return flows also require extra investments in drainage systems: outfall drains, E-ponds, and the other drainage facilities used to collect and dispose of the drainage water. Improved irrigation water management helps control drainage returns flows as it saves on the required river abstraction and reduces the disposal volumes. Specific measures to be considered include lining of canals and improving on-farm water management. These measures are, however, generally only attractive to farmers when there are adequate premiums on saving water and reducing drainage rates (see Colorado and Murray-Darling experiences in annexes B and D, respectively). Low-salinity drainage water may be reused provided that the salt balance of the land is maintained. In most irrigation schemes, reuse is, in fact, the rule rather than the exception as many tail-end farmers depend heavily on the drainage water coming from headlands and downstream river diversions almost always include a proportion of drainage water. Even drainage water with a fairly high salinity may be reused for the irrigation of salt-tolerant tree plantations. This latter reuse is particularly advantageous when the drainage water is eventually being disposed of through an outfall drain or to E-ponds as it reduces the disposal volume. 36 7.2 Combating Salt Mobilization An analysis of salt balances (chapter 5) indicated that irrigation-induced mobilization of resident salts is a major cause of salt-loading of the rivers in four out of the five studied basins (even in the fifth basin, the Indus, there is considerable internal mobilization of resident salts by the tube-well pumping that, as explained earlier, does not show up in the salt balance). Much of the salt mobilization and the resulting river-loading seems to be an inherent feature of irrigation development in arid-zone basins. These basins, by the nature of their geomorphology and climatic setting, tend to be rich in resident salts. The introduction of irrigation, almost inevitably leads to a buildup of groundwater and a rise of the water table under the irrigated land as maintenance of the salt balance in the upper soil layers and the economics of water distribution/application both justify the allowance of some deep percolation, which will often exceed the natural drainage capacity of the land. In the process, resident salts in the deeper layers may be mobilized and become part of the (geo) hydrological and geochemical regimes of the basin. Efficient irrigation may be generally expected to reduce such mobilization as equilibrium may be established with less groundwater buildup and lower seepage flows but in the long run it should not be expected to fully prevent mobilization. If the resident salt layers and the drainage flow paths can be mapped, measures that reduce the irrigation losses, which feed saline seepage flows, can be very effective. This also applies to the judicious construction of interceptor drains/wells that prevent saline seepage flows from reaching the drainage and river systems. Another effective measure is the replacement of tube wells pumping saline water with shallower drainage systems. The best opportunity for reducing the mobilization of resident salts (the time for which, unfortunately in most cases, has already passed), generally lies in the project planning stage. Planning of irrigated areas and the system layouts, on the basis of careful mapping and modeling of the geohydrological conditions of the area, could have prevented much of the salt mobilization and salt-loading of the rivers currently taking place in many irrigated basins in the arid zone. 7.3 Basin Management In some basins, there is still considerable scope for river disposal, especially when extra dilution of water can be made available during certain critical periods. The downstream salinity of the Indus River is still exceptionally low during most of the year, rising to critical levels only during a short pre-monsoonal period (see annex C). In the future, when more storage may be available, it may well be possible to mitigate this critical period and allow more year-round river disposal (particularly combined with judicious management of drainage return flows, including temporary storage, and full use of alternative disposal options). In some of the other basins, there also seem to be opportunities for creating more storage or enhancing flow during critical periods by adapting reservoir operations. Improved irrigation water management resulting in reduced abstraction of river water also helps, especially when realized by schemes in the upper reaches of the rivers. Alternative land use and/or cropping patterns (possibly including retirement of irrigated land) may be considered for schemes in the upper part of the basin having high water requirements, generating much salt-loading while yielding only low economic returns. For schemes in the threatened downstream zone, the feasibility of shifting intakes upstream may be investigated. 37 Extension services may also promote a general adjustment of the salt tolerance of the cropping pattern to the salinity of the irrigation water. Land use planning should also be extended to all nonirrigated land that, one way or another, contributes to the salt-loading of the river. 7.4 Non-River Disposal In almost all the five basins, some forms of non-river disposal are already practiced. Disposal to E-ponds is the main mode of salt disposal in the Aral Sea Basin but it is also used in most of the other basins on a smaller scale. In one basin (Indus, see annex C)), there is already an operational outfall drain while this option is under consideration for the Amu Darya Basin. Deep-well injection is not yet practiced at an operational scale in any of the basins; nor is desalinization. The Yuma desalinization plant in the lower Colorado Basin was constructed as a safeguard to assure that the salinity of the river water allocated to Mexico would not exceed the agreed limit but so far the use of the plant has not been necessary (see annex B). Many of the E-ponds currently used are located in desert areas adjoining the irrigated areas. The drainage catchments served by these ponds generally do not have a ready outlet to the main river system and the choice of this disposal option seems to be motivated as much by hydrological and cost considerations as by the desire to reduce the salt-loading of the river. Generally, the ponds seem to perform their function quite well. The lessons from the Aral Sea confirm, as predicted by evaporation fundamentals, that there is a limit to the receiving capacity of these ponds. To optimize the available E-pond capacity, drainage water that is still of usable quality should, as much as possible, be reused and further concentrated before being disposed of. Most of the constructed ponds seem to have caused no significant environmental damage. These risks have been widely publicized after the selenium poisoning observed in some E-ponds in western USA. However, in view of the experiences in other countries it would seem that these risks are rather limited and that E-ponds when well-designed and -managed, constitute a feasible and often effective alternative to river disposal. Worldwide, there are only a few large outfall drains in operation and the available experiences, almost exclusively based on the Left Bank Outfall Drain (LBOD) in the lower Indus Basin, are still far from complete and conclusive. Many of the problems encountered during the LBOD construction are tainted by the prevailing political unrest in the area. Experiences now being gained with this drain on the nature of flow regime, on the maintenance requirements, and on the environmental impacts may well be decisive for the future prospects of this disposal option. 7.5 Institutional, Financial, and Implementation Arrangements In two of the basins (Colorado and Murray-Darling) special institutional arrangements and programs have been set up for the control of river salinity. Such a dedicated approach is also under consideration for the Aral Sea but for the Indus and the Nile rivers, salinity remains in the control of the existing water management institutions and programs. These arrangements and programs have very little in common and the specifics may best be understood from the case-by-case descriptions given in the annexes. Here only a few key aspects will be discussed. 38 7.5.1 Stakeholders The control of the river salinity can generally only be achieved by placing severe restrictions on the present or future disposal of saline water into the river and on the present or future abstraction of river water. Under these conditions, one can only expect to succeed in having the required control measures accepted and implemented when all the stakeholders have formally agreed, for example in the form of a compact with the executive agency on which measures need be taken and how the costs and benefits would be shared. 7.5.2 Baseline Conditions The control programs need to be based on well-defined baseline conditions against which the planned salinity control can be monitored and the impacts of the implemented measures can be evaluated. Current conditions may serve as baseline conditions. It is important that the baseline conditions be established as accurately and unambiguously as possible. Normally, there will also be agreed targets indicating to which level and within which period the current salinity needs to be reduced or the future salinity will not be allowed to exceed. The baseline conditions and targets may relate to average conditions but may also cover some critical periods in the annual cycle. 7.5.3 Modeling Simulation models that can predict and help visualize the river salinity conditions to be expected under various control scenarios have proven very useful in presenting the planned program to various audiences and in reaching agreement on the need for control and on the nature of the required measures. These and other more-refined models may also be used for the design of control programs. In the Colorado Basin, the simulation model has been even given a limited legal status as a tool for assessing benefits and contributions and for arbitrating conflicts (see annex B). 7.5.4 Cost Sharing Since the causes of the current high river-salinity conditions generally have a complex and long history and usually cannot be precisely identified and quantified, experience shows that it is generally not productive to try to assign responsibilities for past river salinization. The best approach is to accept the costs of combating past salt-loading as common costs and to apply the “polluter pays” principle for measures required to keep the salinity within the agreed targets. 7.6 Economics For the Colorado and the Murray basins, some indicative data on the economics of the river salinity control measures instituted are available (for more details, see annexes B and D). The 39 given costs generally refer to the conditions obtained in the early/mid-nineties. For the Colorado, it is estimated that the total costs of salt-loading reduction average about US$70/ton of salt while the corresponding benefits are calculated to be about US$340. Costs vary considerably for the different control measures (from US$5 to US$138/ton of salt reduction). Best benefit/ cost ratios are generally obtained for measures that reduce irrigation losses (improved on- farm water management and canal lining) and that control erosion of saline land. Plugging of flowing brine wells was also found to be highly cost-effective. The benefits derive mostly from the agriculture sector but damage reduction to industrial and household installations also contributes to the benefits. For the Murray Basin, it is estimated that the annual damage of a further 1.0 mg/l increase of the river salinity (measured at Morgan in the central part of the lower basin) is about A$130,000 (US$100,000). It is also estimated that the investment costs of the initial salt interception program that is designed to reduce the river salinity at Morgan by some 80 EC units (50 mg/l) will come to some A$27 million (US$21 million) and have annual O&M costs of some A$1.7 million (US$1.3 million). Taking the total annual costs (capital and O&M) to be about US$2.5–3.5 million, this amounts to some US$50,000–70,000/mg/l river salinity reduction, which is equivalent to US$8–12/ton of river salt reduction, taking the average annual flow at Morgan at 6.1 km3. Most of the benefits are said to derive from the M&I use of the water. 40 41 CHAPTER 8 SUMMARY AND CONCLUSIONS In irrigated river basins, river flows are systematically depleted by the diversion of irrigation water while, on the other hand, the remaining flow is increasingly loaded with saline drainage water. As a result, rivers in these basins typically develop longitudinal salinity profiles with increasing downstream salinity levels. This phenomenon, which is generally restricted to irri- gated river basins in the arid zone, was studied in five major basins (Aral Sea, Colorado, Indus, Murray-Darling, and Nile basins). It was found that the downstream increasing salinity profiles occur in all of the studied five basins but that considerable differences exist amongst the basins. The differences can broadly be related to differences in the salt balances of the respective basins. Three generic types of profiles can be distinguished: equilibrium, salinization, and mobilization profiles. 8.1 Equilibrium Profile This type of profile is expected to form when the salt regime of the basin is in balance, i.e., when the drainage flow returns as much salt to the river as is diverted with the irrigation water. This means that the resident salt storage in the irrigated areas and the salt-load of the river both remain at the same level. The shape of the profile depends on the degree of depletion of the river flow, on the proportion of drainage return flow, and on the spatial distribution of the abstraction and drainage return points but it may generally expected to be of an exponential form. The downstream salinity may be expected to increase with an increasing degree of flow depletion and a decreasing proportion of drainage return flow. For normal depletion and drainage return percentages, the salinity at the tail end of the river may be generally expected to be about twice as high as at the head of the river. 8.2 Salinization Profile This type represents the case where the rate of salt diversion from the river exceeds the rate of salt disposal into the river. The salt balance of the land is positive as salts are accumulating in the irrigated areas. This profile typically occurs in basins and irrigated areas, which do not have adequate drainage systems. Since much of the salt will be accumulating in the upper soil layers, the land will gradually become more and more salinized. Although less than for the equilibrium case, the salt-loading of the river is generally still sufficient for the river salinity profile to show an increase in downstream salinity and also have a slight exponential shape. 42 The profile will, however, generally be flatter and have a lower salinity level all along the river than the equilibrium profile. 8.3 Mobilization Profile In this case, the salt balance of the land is negative, meaning that more salts are being disposed of into the river than are being diverted. It means that resident salts in or near the irrigated area are being mobilized and carried to the river, usually by irrigation-induced saline drainage and seepage flows. This profile typically occurs in basins and irrigated areas, which are underlain by fossil salts or otherwise have a high resident salinity and geohydrological conditions conducive to the generation of deep seepage flows. These profiles are at a higher salinity level than the equilibrium profiles and have steep gradients, especially in the mobilization reaches of the river. The study confirms that mobilization of resident salts is a widely occurring phenomenon in irrigated basins in the arid zone. By their climatic and geomorphologic setting, the irrigated areas in these basins typically abound with fossil and primary salts, which risk is being mobilized by changes in geohydrological regimes, induced by the irrigation water losses. Irrigation-induced mobilization of these resident salts was found to be the main source and cause of the increased downstream river salinity in four out of the five studied basins. A number of measures can be taken to control or even reduce the increase in downstream river salinity but none of these measures are easy to adopt in a developed basin where the water resources have been almost fully allocated and used. The principal choices are between reversing the flow depletion or otherwise increasing the fresh river flows and reducing the salt-loading of the river while both maintaining the salt balance of the land and especially preventing the salinization of the land disposal of salts by means other than via the river. Within each of these broad categories there is usually a further choice as to which specific measures are to be taken. Generally, there are no win-win solutions and the measures to be taken either place restrictions on water use or on the disposal of saline water, or add to the water use/disposal costs. Control programs have been adopted and are currently being implemented for the Colorado and the Murray basins and both programs offer useful guidance on the technical and institutional design of such programs. The control programs and measures must be generally basin-specific, taking into account the specific physical, institutional, and socioeconomic conditions of the basin. In these two basins, it was found to be of critical importance to assure that the stakeholders have a full understanding of the need for action and also that they are in full agreement with the proposed measures. It was also found that models, which can visualize and quantify the impacts of various actions (including no action), are of great help in projecting the need for action and reaching agreement on the sharing of the costs. 43 ANNEX A ARAL BASIN General The Aral Sea Basin covers major parts of five former Soviet Union countries in Central Asia (Kazakhstan, Kyrghyzstan, Tajikistan, Turkmenistan, and Uzbekistan) but extends also slightly in some other surrounding countries (see figure A1). The basin is drained by two major river systems, the Amu Darya and the Syr Darya both of which rise in mountain ranges on the Afghanistan and China border in the southeastern part of the basin. Some of the mountain peaks are in the 6,000–7,000 m range and have permanent snow and ice covers. Both rivers discharge into the Aral Sea, a huge shallow terminal lake located in the northwestern part of the basin. The Aral Sea is part of a large geomorphologic bottom area, which encompasses a number of similar undrained terminal lakes as well as numerous playas (salt-encrusted dry lake beds). Between the lower plains and the mountains, there are some intermediate plateaus and piedmont landscapes. The geology of these landscapes includes saliferous claystones and marls and highly saline aquifers. Locally, the soil material in the basin has a rather high natural salinity, made up partly by primary salinity and partly by remnant salinity from past salinization periods. The climate is of the continental type with hot summers and cold winters. There is ample precipitation in the mountain ranges (2,000 mm and more) but the rainfall reduces sharply in the middle and lower basins and amounts to only a meager 100 mm/yr. in the lower plains bordering the Aral Sea. Due to the low temperatures and severe frost during the winter, the agricultural season is essentially limited to the summer. There are some opportunities for rain- fed cropping in the upper basins but in the low-rainfall middle and lower basins little cropping can be done without irrigation. The entire Aral Sea Basin encompasses an area of 1.80 million km2 of which about 20 percent is mountainous and 80 percent is plain land. The current population of the basin is estimated to be about 40 million, up from 14 million in the 1950s. The population growth in the eighties was about 2.5 percent per year but declined to about 1.5 percent in the nineties. The municipal and industrial water use is still very limited. Water Resources Development The total precipitation on the basin is calculated to average about 500 km3/yr. out of which about 116 km3/yr. are disc