Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
2.1 The arid zone in global atmospheric
circulation water resources
2.2 The Tigris and Euphrates Rivers
2.3 The Indus River
2.4 The Nile River
2.5 The Jordan River
2.6 The Colorado River
2.7 Non-renewable groundwater development in the
Middle East
2.8 Brackish-groundwater reverse-osmosis
desalination in Bahrain
2.9 Seawater desalination in the Arabian Gulf
countries
2.10 Groundwater-hydro development in Chile and
Libya
2.11 Mediterranean-Qattara solar-hydro and
pumped-storage development
2.12 Concluding remarks
Concern over global climatic changes caused by growing atmospheric concentrations of carbon dioxide and so-called greenhouse gases has been a major issue of global environment and sustainable development. Recent symptomatic global climatic changes such as El Niño and the Sahel drought and/or desertification have made us more aware of atmospheric dynamics and global climate systems. Among the most important consequences of future changes in climate will be alterations in regional hydrological cycles and subsequent effects on the quality and quantity of regional water resources, yet these consequences are poorly understood. Recent hydrological research strongly suggests that it is plausible that climatic changes caused by increases in greenhouse trace gas concentrations will (1) alter the timing and magnitude of runoff and soil moisture, (2) change lake levels and groundwater availability, and (3) affect water quality. Such a scenario raises the possibility of dramatic environmental and socio-economic dislocations and has widespread implications for future water-resources planning and management.
By looking at the variations of climate and hydrological conditions in geological time, we can simulate the analogies of future greenhouse conditions which may cause significant changes in water availability. Nicholson and Flohn (1980), using a variety of palaeoclimatic records, suggested that there have been significant climatic changes in the Sahel region that can be related to possible driving forces in global atmospheric circulation, which may explain the huge amount of fossil groundwater resources stored in the extensive Nubian sandstones of the Sahara desert during the Late Pleistocene age.
Water differs markedly from most other natural resources by its remarkable property of continuous renewal in the water cycle, the main link in which is water exchange between oceans and the land. The world ocean is a gigantic evaporator, which, in this natural cycle, is the main source of fresh water. The fresh water falls as atmospheric precipitation and is the source of all water flows and water accumulation on land. The greater part of the water on the earth, approximately 97.5%, is salt water and water that is mineralized.
The volume of fresh water amounts to 35,029,000 km³, or 2.52% of the total uncombined water on the earth. Rivers and streams account for 0.006%, fresh-water lakes for 0.26%, and water contained in the atmosphere for 0.001 % of the total quantity of fresh water. The rest of the fresh-water component occurs as soil moisture, permanent snow cover, marshes, and active groundwater (Korzun et al. 1976). Since the end of the nineteenth century the attention of many scientists has been directed to the problem of the connection between atmospheric precipitation, run-off, and evaporation in river basins. Water-resources development in the twentieth century has been directed to exploiting rivers and streams, lake water, and groundwater, which have had to be exclusively fresh. Saline waters such as seawater and brackish water, of which the storage potential is estimated to be as much as 97.5% of all the water on the earth (Korzun et al. 1976), may suggest their increasingly important role in the water-resources planning of arid to semi-arid countries in the twenty-first century.
2.1.1 Causation of the arid zone
On a world scale there is a sharp fall in rainfall near 30° latitude in both hemispheres (see the horse latitudes in fig. 2.1), and in places, notably on the west coast of Chile and Peru, rainfall is almost or entirely unknown (fig. 2.2). Since this aridity affects oceans as well as continents, it is apparent that there must be some process whereby rainfall is suppressed, for there is no shortage of water for evaporation. We shall look for this process among the dynamic factors associated with the general circulation of the atmosphere.
Fig. 2.1 Arid zone and mean annual precipitation
Fig. 2.2 Arid-region climate graphs (Arica and Antofagasta, Chile)
Fig. 2.3 Global wind convection and atmospheric pressure zones
The immediate causes of this aridity can be identified as follows: The first is the tendency towards divergent wind-flow at low levels, especially in the pole-ward parts of the trade winds (fig. 2.3). This divergence causes a general subsidence of the air column throughout the low troposphere where the water vapour is concentrated. Dynamic warming associated with this subsidence tends to lower the relative humidity and disperse clouds; it also creates a significant degree of hydrostatic stability, so that convection currents and shower-type precipitation are inhibited. At sea the trade winds have a shallow, moist layer of moving air capped by a stable or inversion layer above which the air is very dry. Over land, as in the Sahara, northern Mexico, Australia, and much of the Middle East (in summer), the dryness may extend to ground level. Hence surface divergence, general subsidence, low humidities, and an absence of deep convection form a causally connected sequence over much (but not all) of the subtropical world. The second cause of aridity, not truly independent of the first, is the existence of high-pressure zones near the latitudes of 30° north and south (figs. 2.4 and 2.5). Over the ocean, this high pressure is recorded at sea level, but inland it may be necessary to ascend to levels of 2-3 km to encounter the propagation of disturbances and the subtropical high-pressure belt, which is continuous in both hemispheres though considerably modified in form over southern Asia in summer. These subtropical high-pressure belts also separate the circumpolar westerlies from the tropical easterlies. As is well known, both westerlies and easterlies are affected by travelling wave perturbations, which account for a good part of the precipitation in both regimes. The amplitude of both sets of waves is at a minimum near the axis of the subtropical highs; hence the subtropics are least likely to be affected by rain-bearing disturbances. This constitutes the second cause of widespread aridity.
Fig. 2.4 Schematic global circulation and tropopauses
Fig. 2.5 Atmospheric circulation
In general it should be stressed that low humidities throughout a deep layer in the lower troposphere invariably lead to aridity, while very dry climates may also occur in areas of high atmospheric humidity. Thus parts of the arid south-western United States of America are dry not because of low humidities but because of the ineffectiveness of rain-making disturbances (Benson and Estoque 1954).
A further point is that remoteness from the sea is not a guarantee of drought; high moisture contents occur deep in the interior of Amazonia, more than 2,000 km from the ocean along any possible direct streamline, and rainfall is heavy. Yet extremely low rainfalls occur in many areas along oceanic coasts, as in Chile, Peru, Morocco, and southwest Africa (see figs. 2.1 and 2.2). The point is, of course, that local sources of evaporation play a very small role in precipitation, which tends in most instances to fall from moist air streams whose humidity has been derived from very remote surfaces.
The aridity of the subtropics thus emerges as an aspect of world climate dependent on deep-seated features of the earth's general atmospheric circulation. It does not arise from local or man-made circumstances but from natural causes involving exceedingly large energy transformations and momentum transports. It has not been conceived that the regime can be significantly altered by human intervention. Discussion today in the early 1990s, however, is based on the worldwide concern with global climatic changes, which may be due to the substantial increases of carbon dioxide caused by industrialization since the Second World War.
It is equally unlikely that any past climatic epoch can have experienced a complete absence of subtropical aridity. As we have seen, the maintenance of the mid-latitude westerlies absolutely requires the existence of compensating easterlies in the tropics. Similarly the transfer of momentum and heat northward in the tropics requires the existence of a Hadley cell (figs. 2.3-2.5), with subsidence (and hence low humidity and drought) at some subtropical latitude. Hence it seems likely that the arid zone can have been no more than constricted in extent and driven a few degrees towards the equator at the height of recent glacial development; it can hardly have been eliminated altogether (Hare 1964).
2.1.2 Definition and classification of the arid zone
Deserts have been mapped in regions as diverse as low and high latitude zones, inland continents, and coastal zones. Cold deserts, which are situated in high mountain ranges or high-latitude zones, are not included in this study, which deals with warm deserts situated between the tropics and the temperate zones.
One of the simplest classifications of dry climates takes 10 inches (250 mm) as the dividing line between arid and semi-arid and 20 inches (500 mm) between semi-arid and humid (refer to fig. 2.1). Although these criteria are scorned many scientific geographers and climatologists, they are actually not bad for the standard climatic classifications.
Most classifications today use combinations of temperature and precipitation, in order to make some allowance for the increasing evaporation with higher temperatures. De Martonne (1926, 1942) and Koppen (1923) both used figures of mean annual precipitation (P) and temperature (T). The basic de Martonne formula gives an index of aridity (I) which is a true sliding scale without artificial break points: I = P/(T + 10). Koppen's formula has rigid break points, but these are varied according to the season of maximum rainfall. The margin between arid and semi-arid would be R (rainfall) = T + 11. Koppen's formulas were an attempt to assign climatic values to the limits of the main vegetation types of the world.
Another approach analyses the temperature and precipitation ratios of individual months. The monthly indices, including mean monthly precipitation (Pm) and temperature (Tm), are then summarized into an annual figure. Thus, de Martonne's index becomes the total of twelve monthly indices, each of which is calculated as 12Pm/(Tm + 10).
Thornthwaite's basic system is more complex, but uses only mean monthly temperature and precipitation to arrive at a figure for estimated potential evapotranspiration (PEt) (Thornthwaite 1948). The arid regions are classified on the basis of an aridity index (AI) which assumes (1) AI = 0 when P = PEt, (2) AI = - 100 when P = 0, (3) AI = + 100 when P is much greater than PEt, with the following zone classifications corresponding to the AI ranges indicated:
This method has been widely used by climatologists owing to its simple methodology. For the irrigation engineer, however, the method of estimating the parameters for potential evapotranspiration was found to be not accurate enough to be used for design purposes. Thornthwaite has introduced soil factors to the point that his latest indices can no longer be considered purely climatic (Thornthwaite and Mather 1957).
An early version of the concept of evaporation balance was devised by Albrecht Penck (1910), who used the geomorphic factor of hydrographic balance for a broad climatic classification. De Martonne (1942) drew a world map upon such a basis.
Emberger (1955) uses mean annual precipitation (P), mean daily maximum temperature of the warmest month (M), and mean daily minimum temperature of the coldest month (m), all combined into a single moisture quotient (Q) using the following formula:
Unlike the formulae of Koppen, Thornthwaite, or de Martonne, Emberger's moisture quotient cannot be used by itself to make a valid climatic map. His maps appear to be based on his profound knowledge of vegetation, not on the mapping of climatic data. After mapping the vegetation zones, he determines the associated moisture quotients and other climatic values within the zones. Thus, the northern limit of his arid zone in north-west Africa varies from a moisture quotient of 16 to one of 40. In characterizing the climate of his stations, he uses the actual mean daily minimum temperature of the coldest month, as well as the moisture quotient. His maps for north-west Africa, which he calls bioclimatic zone maps, are of fundamental value for all geographers and climatologists. Because of the accuracy and detail of his mapping, the maps form a valuable test of the vegetation validity of any climatic system. A simple check shows, for example, that de Martonne's index of 10 would serve fairly well as the northern limit of Emberger's Mediterranean arid bioclimatic zone.
Meigs, a former chairman of the Arid Zone Commission of the International Geographical Union, made a global arid zone map in 1951 (fig. 2.6), whose zonal classifications have been widely used in arid-zone research since World War II. Meigs's classification is based on regional temperatures (mean monthly maxima and minima) and the duration of dry periods. His arid index map distinguishes three subregions: extremely arid, arid, and semi-arid. The extremely arid region, which receives zero annual rainfall for twelve months of continuous observation, occupies about 4% of the land area of the earth. The arid region, which has a minimum of one month of rainy season, covers about 15% of the land area of the earth. The semi-arid region has a rainy season with 100-200 mm of rainfall and occupies about 14.6% of the land area of the earth (Goudie and Wilkinson 1977).
2.1.3 Global atmospheric circulation and climatic changes in the arid zone
In the last twenty years low rainfall in the Sahel, possibly related to climatic changes, has reduced water availability in North Africa. Folland and Palmer (1986) noted that there is a strong inverse relationship between worldwide sea surface temperature (SST) and Sahelian rainfall. The SST has been rising since the late 1960s, which period corresponds with decreasing rainfall in the Sahelian region. The present increase in SST is considered likely to be linked to global climatic change, which suggests the warming of the climate by the greenhouse effect. However, the rising SST may also be linked to El Nino southern oscillations.
Fig. 2.6 World arid zones (Source: P. Meigs)
El Niño events occurred in 1972-73, 1976-77, and 1982-83, resulting in drought in the Sahel. Failure or weakening of the Guinea monsoon seriously reduced the rainfall in the Ethiopian highlands. Most of the major rivers of North Africa originate in the Central African uplands, the Atlas Ranges, and the Ethiopian highlands (see section 2.2.4). More than 80% of the Nile's water originates in the Ethiopian highlands. The reservoir level in Lake Nasser has fallen by more than 18 meters in the last seven years owing to the African drought (WPDC 1988). As a result, usable effective storages are only onefifth (20 x 109 m³) of the 1979 level.
The increasing albedo is another climatic factor that affects water resources. Desertification has increased the albedo level in the catchment area of northern African rivers. The net effect of a high albedo level is that the sun's energy is reflected to the atmosphere to heat the air, causing increased evaporation and transpiration, resulting in a decrease in the potential water resources.
2.1.4 Palaeoclimatology and water-resources planning
About 10,000 years B.P. the global climate was moist and cooler, indicating atmospheric carbon dioxide concentrations between 260 and 290 mg/l The earth was covered with 6.2 billion hectares of forest as compared to 4.1 billion hectares today at the end of the last (fourth) glacial age (Matthews 1983).
Nicholson and Flohn (1980) explored water availability in Africa using a variety of palaeoclimatic records. They suggest that parts of the Sahel region were drier 18,000 years before the present (B.P.), became moister and cooler with frequent rains in the period 10,000-4,500 years B.P., and then became drier again up to the present. Similarly, Goodfriend et al. (1986) explored the palaeoclimatic evidence for climatic changes in the area of the Jordan River basin and the Dead Sea. They identified large fluctuations in the level of the Dead Sea, its terminal lake, in the Late Pleistocene period up to 4,300 years B.P.
Reconstructions of the record back several centuries can provide valuable information on both past climatic conditions and the vulnerability of our water resources system to future changes. In a striking example, Stockton and Jacoby (1976) used tree rings to extend the run-off record in the Colorado River basin back more than 400 years, using carbon-13 (13C) isotope in tree rings as the environment tracer (fig. 2.7). This kind of study has direct water management and policy implications. For example, the original 1922 Colorado River water allocation was based on the hydrological record available at that time: namely about 30 years from the late 1890s to the early 1920s. In 1976, when the historical record was reconstructed back to the middle 1500s, the period from 1890 to the 1920s stood out as a time of abnormally high run-off. The 400-year record now shows that more water was allocated to users than is likely to be available on a long-term average basis. If the long-term record had been available in 1922, such over allocation might not have occurred. These changes can also be related to possible driving forces in the global atmospheric circulation.
PAEAEO-CARBON DIOXIDE AND THE GREENHOUSE EFFECT. The atmospheric carbon dioxide concentrations in air particles trapped in the 2,200 metre Vostock ice cores have been analysed to examine the historical changes of the climate since 160,000 years B.P (Matthews 1983). These Antarctic ice cores also provide temperature information for the same period based on the oxygen isotope (18O) ratio. The derived temperature changes closely match changes in carbon dioxide concentrations (fig. 2.8).
Gammon et al. (1985) reviewed the history of atmospheric carbon dioxide from 100 million years ago to the present. During the Cretaceous age (100 million years B.P) the carbon dioxide level was perhaps as high as several thousand parts per million; then it dropped to 200300 mg/l during the glacialinterglacial cycles of the past few million years through 10,000 years B.P. Since the nineteenth century, the carbon dioxide level has increased to about 350 mg/l, which could be owing to the accelerated burning of fossil fuels as an energy source (see fig. 2.9). North America, western Europe, and the former USSR and Soviet-bloc countries produce 67.4% of the world's carbon dioxide emissions to sustain their industrial activities (Rotty and Masters 1985).
COHMAP AND GCMs. The scientists of the Cooperative Holocene Mapping Project (COHMAP) have assembled a global array of welldocumented palaeoclimatic data and general circulation models (GCMs). The GCM, which was developed to look at soil moisture in the mid-continental region of the United States of America, has shown that significant drying may occur if the concentration of carbon dioxide in the earth's atmosphere is doubled (Manatee and Wetherald 1986). GCM-generated hydrological data suffer from two major limitations: (1) the spatial resolution of GCMs is too coarse to provide hydrological information on a scale typically of interest to a hydrologist, and (2) hydrological parameterizations in GCMs are very simple and often do not provide the detailed information necessary for water-resources planning (WMO 1987). The hybrid COHMAP-GCM model simulates historical worldwide climatic changes in the atmosphere, geosphere, and biosphere that accompanied the transition from glacial to interglacial conditions during the past 18,000 years with geological and palaeo-ecological evidence (COHMAP members 1988), the results of which are to be used not for predicting the changes in water resources systems in future but for understanding changes in global climate on a geological time scale.
To assess the implications of the greenhouse effect for water resources in the arid to semi-arid regions, regional-scale details of future changes are needed for temperature, precipitation, evaporation, soil moisture, and other hydro-climatological variables. It is simply assumed that increasing concentrations of greenhouse gases will have an adverse effect on water resources in tropical and temperate arid zones.
2.1.5 Nature of the hydrological cycle and global water balance
Hydrology is concerned with the occurrence and movement of water on the earth. Water is one of the commonest substances in nature. It occurs in chemically combined forms, in free states, and in the biosphere. The free states include groundwater and soil moisture in the upper layer of the lithosphere and in the soil cover, and water in oceans, seas, lakes, rivers, glaciers, and permanent snow cover. A small amount of water occurs in the atmosphere as water vapour, water drops, and ice crystals. Unlike other natural resources, water is continually moving and changing from one form to another. The movement is a characteristic of all forms of water. The movements of ocean currents, rivers, groundwater runoff, humid air over oceans and continents, and transpiration are links in the interconnected water cycle of nature.
HYDROLOGICAL CYCLE. The general concepts of the global hydrological cycle are illustrated in fig. 2.10. The rectangles of the figure denote various forms of water storage: in the atmosphere, on the surface of the ground, in the unsaturated soil moisture zone, in the groundwater reservoir below the water table, in the channel drainage network, and in the oceans. The arrows in the diagram denote the various hydrological processes responsible for the transfer of water from one form of storage to another.
Thus the precipitable water (W) in the atmosphere may be transformed by precipitation (P) to water stored on the surface of the ground. In the reverse direction, water may be transferred from the surface of the ground by evaporation (E) or from the unsaturated soil by transpiration through vegetation and subsequent evaporation from the leaf surface (Et).
Some of the water on the surface of the ground will infiltrate through the surface into the unsaturated soil (F), but some of it may find its way as overland flow (Qo) into the channel network. During precipitation, if the field moisture deficit of the soil which has arisen since the previous precipitation is substantially satisfied, then there will be either recharge (R) to the groundwater or else lateral interflow (Qi) through the saturated soil into the channel network.
Groundwater storage is depleted by groundwater outflow (Qg), which enters the channel network and supplies streamflow during dry periods. During prolonged droughts, soil moisture may be replenished by capillary rise (C) from groundwater to the unsaturated zone and subsequent loss to the atmosphere by evapotranspiration.
Overland flow (Qo), lateral interflow (Qi), and groundwater outflow (Qg) are all combined and modified in the channel network to form the run-off (Ro) from the area for which the balance is being calculated. These various hydrological processes form the subject matter of physical hydrology.
GLOBAL WATER BALANCE. The quantities of water in the ocean, atmosphere, ice masses, lakes, and rivers may be evaluated without too much difficulty. It is more difficult to determine the amounts contained in living organisms and in the lithosphere. The figure of greatest uncertainty is that for inactive groundwater. A general idea of the world water reserves is shown in table 2.1. The total reserves are preliminarily estimated to be about 1,386 million km³. The breakdown of the volume of each form of water is shown below (Korzun et al. 1978):
>> Ocean and seas. The volume of water in the ocean is estimated to be 1,338 million km³ by assuming that the ocean covers an area of 361.3 million km² with an average depth of about 3.7 km. This represents about 96.5% of the total water reserves of the earth.
Table 2.1 World water reserves
| Volume (km³) | Share of world reserves | ||
| Share of total water(%) | Share of fresh water(%) | ||
| Atmospheric water | 12,900 | 0.001 | 0.01 |
| Glaciers end permanent snow cover | 24,064,000 | 1.74 | 68.7 |
| Ground ice in zones of permafrost strata | 300,000 | 0.022 | 0.86 |
| Water in rivers | 2,120 | 0.0002 | 0.006 |
| Water in lakes (fresh, 91,000 km³) | 176,400 | 0.013 | 0.26 |
| Water in marshes | 11,470 | 0.0008 | 0.03 |
| Soil moisture | 16,500 | 0.001 | 0.05 |
| Active groundwater (in aquifers), including brackish and fossil | 10,530,000 | 0.76 | 30.1 |
| Inactive groundwater (in lithosphere) | 23,400,000 | 1.7 | |
| World ocean | 1,338,000,000 | 96.5 | |
Source: Korzun et al. 1976.
>> Glaciers and permanent snow cover. The amounts of water in the ice of the polar regions and in the glaciers of mountainous regions are estimated to be about 24 million km³, which accounts for 68.7% of the earth's total resources of fresh water.
>> Inactive groundwater. Substantial amounts of water are stored in the lithosphere. The inactive groundwater in the earth's crust-that is, gravity water contained in the pores and cracks of saturated strata-is estimated to be 23.4 million km³ by assuming an effective porosity of 5%-15% and a maximum depth of 2,000 m.
>> Active groundwater. The depths of fresh-water accumulations vary, depending on local geological characteristics. By assuming an effective depth of aquifers between 200 and 600 m, the volume of active groundwater is preliminarily estimated to be 10.53 million km³, which is 30% of the total volume of fresh water.
>> Soil moisture. Soil moisture is more closely related to weather conditions than is groundwater. During the wet seasons, moisture is stored in the soil, while it is removed by evaporation and tran spiration in the dry seasons. The storage of moisture in the soil is estimated to be 16,000 km³, by assuming a soil layer of 2 m thick with 10% of moisture on average.
>> Lakes and reservoirs. There are numerous lakes in the world. Large lakes with an area of more than 100 km² may store 95% of the total reserves in all lakes in the world. The total water volume in the world's 145 large lakes amounts to 168,000 km³, which is 95% of the world's total of 176,000 km³. Of this, 91,000 km³ is in freshwater lakes, while approximately half the water (85,000 km³) is salty. Most of the salty lake water is concentrated in large lakes without outlet, such as the Dead Sea, the Caspian Sea, and the Aral Sea.
Intensive construction of large dams, especially since the Second World War, has created large reservoirs on major rivers. The total capacity of the 10,000 reservoirs of the world amounts to about 5,000 km³, with a net capacity of about 2,000 km³, which now controls approximately 14% of the total annual river run-off of 445,000 km³.
>> Marshes. Marshes occur in many areas of the earth. These are mostly peat marshes in countries with temperate climates and their equivalents in tropical and equatorial areas. The total amount of marsh water in the world is preliminarily estimated to be 11,000 km³.
>> River channels. The total water storage in river channels of the world at any given moment is estimated at 2,000 km³, which accounts for only 0.006% of all fresh water. However, it is of great importance to development as a renewable source for water supply.
>> Atmosphere. Water is contained in the atmosphere in the form of water vapour, water drops, and ice crystals. The total moisture in the atmosphere amounts to about 12,000 km³, which is equivalent to 25 mm of water if spread over the whole surface of the globe.
2.1.6 Remarks on global water reserves and water resources
The total amount of water in the hydrological cycle is constant and can neither be increased nor diminished. From the global-scale water budget study outlined above, it looks as though there is more than enough fresh water to meet the demands of human survival, both now and in the foreseeable future. However, water is often available in the wrong place, at the wrong time, or in the wrong quality. This uneven distribution is highlighted in the arid regions. In many of the countries in the Middle East, the hydrological cycle is being disturbed by overexploitation, depletion, or deterioration of the fresh waters in rivers and groundwater aquifers.
Salt waters, including seawater and brackish waters in rivers and aquifers, have been conceived as either useless or harmful and as being outside the scope of water-resources planning except where neither conventional river water nor groundwater of good quality exists. The volume of conventional fresh-water reserves on the earth, however, is minimal compared with other forms of water such as seawater, which accounts for 96.5% of the total water on the earth. Potential groundwater reserves are estimated to be as high as 30% of the total freshwater reserves. However, brackish water is predominant in the major aquifers of the arid region owing to the minimal rainfall and hence minimal groundwater recharge.
Most of the known conventional water resources such as river water and groundwater of good quality or low salinity have already been developed or will soon be fully exploited in most countries of the Middle East. Non-conventional waters such as seawater and brackish waters, including both surface and groundwater, therefore, seem likely to play an increasingly important role in water-resources planning of the arid region for the twenty-first century, when the advances and innovations of desalting technologies, including reverse osmosis applications, are taken into account to save energy and cost.
