World water use and water availability: trends, scenarios, consequences

358 Water Resources Systems Hvclrologica! Risk, Management and Development (Proceedings of symposium HS02b held during IUGG2003 at Sapporo, July 2003). IAHS Publ. no. 281, 2003. World water use and water availability: trends, scenarios, consequences IGOR A. SHIKLOMANOV & JEANNA A. BALONISHNIKOVA Slate Hydrological Institute (SHI), Second Line 23, St Petersburg 199053, Russia ishiklom(gjzb3627.spb.edu Abstract New results of an assessment of future world water use obtained at SHI in 2001 are presented. Using data from Shiklomanov (2003) as well as new data on water use all over the world, a new scenario has been developed for sustainable global water use before 2025, i.e. a "Sustainable Development" scenario. This scenario is an alternative to that used in Shiklomanov (2003) which was developed in 1997 and is known as the "Conventional Scenario". According to the Sustainable Development scenario, world water use until 2025 will be approximately stable. It will increase by about 6% before 2010, and then gradually decrease. Despite the fact that stabilization (and even a decrease of water use in the future) can significantly decrease the pressure on water resources, this will not greatly affect the specific water availability in most regions in the world due to population growth. Key words global water resources; water availability; water use INTRODUCTION Water availability is one of the most important indices of sustainable development in modern society. When characterizing water availability in any region (country) of the world it is necessary to have data on both water availability in the study area and on the dynamics of water resources use. Water resources comprise a stable index during a stationary climate, while water use varies in time and depends on physiographic and socio-economic factors. Moreover, major water use indices are the volumes of water used in different sectors of the economy (agriculture, industry and municipal water use) as well as the volume of water lost to evaporation from reservoirs. Two aspects of the problem are usually discussed when the dynamics of global water use are studied. First, trends of water use in certain regions (countries) for the previous decades are analysed; second, a forecast (scenario) for the future is made taking into account population number, level of the socio-economic development, available water and land resources, and opportunities for introducing novel water-saving technologies. These two issues have been addressed at the State Hydrological Institute (SHI) and the research results are presented in Shiklomanov (2003) which contains detailed data on the dynamics of water use in the 20th century for each continent, in natural-economic regions and in selected countries. The monograph also contains the results of forecasts by the so-called "Conventional Scenario" (CS). This scenario suggests a projected development of water use from the model of previous decades. On the basis of the available detailed information on countries and regions, a new scenario has recently been developed at the SHI for projected water-use development in the world (Sustainable Development scenario: SDS). This scenario takes into account the progressive tendencies in the dynamics of water use. It is oriented towards

World water use and water availability: trends, scenarios, consequences 359 a wide application of different projects for the modification of technologies of freshwater use, mainly in irrigation and industry. This paper briefly describes the new SDS. Water availability and the load on global water resources until 2025, based on the simulation results of this scenario are presented. Finally, the present situation and the predicted development scenarios are compared. BACKGROUND AND PRESENT SITUATION To forecast the dynamics of water use and water availability in any large region it is necessary to generalize and analyse different physiographic (including climatic) and socio-economic factors. Physiographic factors include renewable water resources and their space-time distribution, land suitability for agriculture and irrigation, as well as climatic characteristics. Socio-economic factors include population number (urban and rural population), area of irrigated land, volumes and water surface of reservoirs, and level of socio-economic development (as GNP in US$ per capita), which explain the volume and structure of water use in each economic sector. All these factors together with the actual data on water use in many countries have been obtained and generalized at the SHI which made it possible not only to analyse the dynamics of water use and water availability during the 20th century (before 1995) for each continent and natural economic region of the world, but also to make a forecast up to 2025 (Shiklomanov, 1997, 2000, 2003). This forecast has been based on the CS of a projected development in water use from the model of previous decades. This scenario is realistic, because it is based on the analysis of tendencies of water use in different countries and regions during 1950 1995 by all main water users, i.e. population, industry and agriculture, taking into account the dynamics of the additional water loss to evaporation from man-made lakes. Thus, according to the CS, it is assumed that no fundamental changes would occur in the attitude of human society to freshwater use. As the results of this scenario show, it is possible to expect a future critical situation in water supply on the global scale. According to computations of low-water periods, about 40% of the global population would live under the conditions of a catastrophically high pressure on water resources before 2025, whereas 58% of the population (about 4.5 billion) would have a very low or catastrophically low water availability. Taking into account the results of this assessment of future world water resources by CS, the International Scenario Development Panel (SDP) organized by the World Water Council developed three new scenarios in 1999-2000 on global water use. These scenarios are based on an assumed fundamental increase of the efficiency of freshwater use during the next decades, both in developed and developing countries. These scenarios are entitled business-as-usual (BAU), technology, economics and private sector (TEC), and values and lifestyle (VAL). According to these scenarios, forecasts based on SHI data before 1995 provide quite optimistic results. For example, according to the BAU scenario, total water use if compared with the present one (1995) would only be higher by 13%; according to the TEC and VAL scenarios, it would be lower by 4% and 26%>, respectively. The results of the forecasts according to the above three scenarios, as well as according to the CS of the SHI, were submitted at the Second World Water Forum (Rijsberman, 2000).

360 Igor A. Shiklomanov & JeannaA. Balonishnikova Meanwhile, a critical analysis of the three scenarios developed by the SDP has been made for different countries and regions of the world. The analysis shows that all these scenarios provide a too optimistic picture of water use, and that some of the assumptions on which these scenarios are based are not realistic because they do not take into account the opportunities and interests of many countries. For example, according to the BAU scenario, it is assumed that there is no increase of irrigated land area in the world after 1995-2000, which is absolutely unrealistic. It is hardly realistic or substantiated that domestic water use in highly developed countries decreases to one half or one fifth, which is assumed in the TEC and VAL scenarios. It is doubtful that the projected increase of freshwater use efficiency is possible in irrigation and industries in all countries of the world (Rijsberman, 2000). SUSTAINABLE DEVELOPMENT SCENARIO (SDS) The SHI Sustainable Development Scenario is based on the following, quite realistic, assumptions for the period of 1995-2025: Population number would attain 7.9 billion persons; this corresponds to the mean value of all the assessments available. Irrigated land area would increase by 20%, with variations in individual regions from 10% (highly developed countries) to 30-40%) (countries with transient economies and some developing countries in Africa and South America). The efficiency of water use in irrigation would attain 15% on average, with variations in individual regions from 20% (highly developed countries) to 10%. Domestic water use (1 day" 1 per capita) would be lower by 10% to 20% in the regions of Europe, by 40% in North America. In Africa water use would be higher by 10-20%) (northern and southern Africa) to 100-120%o (other regions). In the southern region of Asia an increase of 50% is expected. In the other regions there would be no changes, or a decrease by 10-20%> may be expected. In the poorest countries of Africa and Asia, at the end of the period, domestic water use would not be less than 50 1 day" 1 per capita. In the richest countries of Asia, water use would not exceed 300 1 day" 1 per capita. In South America and Australia water use would be either lower by 10-25% or would not change. Industrial water use in the regions with highly developed countries would amount to 0.5-0.6 of the water use observed in 1995. In the regions of the former Soviet Union this value would be 0.8. In the regions of Asia, South and Central America, in north and southern Africa and in Oceania an increase in water use by a factor of 1.2-2.0 is expected, depending on the region. In the other regions of Africa, the industrial water use would be 6-7 times higher. The decreased water withdrawal would be accompanied by an increased water consumption (as % of the total value). Water loss by evaporation from reservoir surfaces would be higher by 5-10% in North America, Europe, the foimer Soviet Union and Australia. In the other regions these losses would be higher by 20-30%). The assessments according to SDS and CS were made individually for selected countries and for each region. The results so obtained were then aggregated for continents and for the entire globe.

World water use and water availability: trends, scenarios, consequences 361 RESULTS According to the SDS analyses, global water use would be approximately stable until 2025. It would increase by about 6% until 2010, then it would decrease slightly and by 2025 it would attain 3890 km 3 year" 1 which would be greater by only 2.7% of that in 1995. Industrial water use would be lower by 10%. Agricultural water use would be practically the same, and domestic water use would be higher by about 25%. The dynamics of water consumption are expected to fluctuate in a similar range (Shiklomanov & Balonishnikova, 2002). More significant changes in the dynamics of water use may be expected on individual continents (Table 1). Compared with 1995, predicted water use in Europe, North America and Australia will be lower by 22%, 24% and 5%, respectively. On the other continents water use would be higher: by 3%o in Africa, by 21% in South America and by 12% in Asia. Even more differentiated dynamics in water use are expected in the natural-economic regions of the world which mainly depend on physiographic features and on socio-economic factors of the countries included in the region. Analyses of the forecast data by water use for each continent and natural-economic region yield the following pattern. In Europe and North America a decrease in agricultural water use of 10% is expected, whereas a slight increase (1-4%) of this water use is expected on the other continents. In the regions the dynamics vary from a decrease by 10-12% to an increase by 15-20%. Industrial water use in most regions of Europe, North America and Australia would probably decrease by 33-43%, in some regions even less. However, in most regions of Asia and South America one would expect a higher industrial water use by a factor of 1.5-2.5, in some regions of Africa by a factor of 7. Domestic water use as a whole in Europe and North America would be slightly lower (by 12-20%»). On the other continents it is expected to be higher. For example, in Africa it would be more than three times higher. It is assumed that specific Table 1 Dynamics of water use in the world by continents. First line = water withdrawal, second line = water consumption (km 3 year"'). Assessment Continent 1900 1940 1950 1960 1970 1980 1990 1995 Forecast CS 2000 2010 2025 SDS 2000 2010 2025 Europe 37.5 96.1 136 226 325 449 482 455 463 535 559 444 416 353 13.8 38.1 50.5 88.9 122 177 198 189 197 234 256 193 198 201 North 69.6 221 287 410 555 676 653 686 705 744 786 669 634 527 America 29.2 83.8 104 138 181 221 221 237 243 255 269 237 234 223 Africa 40.7 49.2 55.8 89.2 123 166 203 219 235 275 337 232 259 292 27.5 32.9 37.8 61.3 87.0 124 150 160 170 191 220 165 176 182 Asia 414 682 843 1163 1417 1742 2114 2231 2357 2628 3254 2310 2476 2487 249 437 540 751 890 1084 1315 1381 1458 1593 1876 1428 1509 1470 South 15.1 32.6 49.3 65.6 87.0 117.0 152.0 167.0 182.0 213 260 175 190 201 America 10.8 22.3 31.7 39.6 51.1 66.7 81.9 89.4 96.0 106 120 92 97 99 Australia 1.60 6.83 10.4 14.5 19.9 23.5 28.5 30.4 32.5 35.7 39.5 30.5 30.8 29.1 & Oceania 0.58 3.30 5.04 7.16 10.3 12.7 16.4 17.5 18.7 20.4 22.3 17.9 18.9 18.8 Totals (rounded) 579 1088 331 617 1382 768 1968 1086 2526 1341 3175 1686 3633 1982 3788 2074 3973 2182 4431 2399 5235 2764 3860 2133 4006 2233 3889 2194

362 Igor A. Shiklomanov & Jeanna A. Balonishnikova Central region of North America 1940 1955 1970 1985 2000 2015 2030 Southern region of Asia 1940 1955 1970 1985 2000 2015 2030 Agricultural Use A Industrial Use Municipal Use Reservoirs CS SDS Fig. 1 Dynamics of water use in the Central region of North America and the Southern region of Asia by economic. Dashed lines show SDS, continuous lines show CS. water use (in 1 day" 1 per capita) in North America (USA and Canada) would be reduced to 400 1 day" 1 (instead of 700 1 day -1 in 1995). In Europe this reduction would be only 10-20 %. In some regions of Africa and Asia, where specific water use was only 30-50 1 day" 1, one would expect an increase to more than twice the current value. As an example Fig. 1 shows the dynamics of projected water use according to two scenarios for two regions of developed and developing countries. When comparing the predicted water use volumes and the amount of renewable water resources in a region, it is possible to assess the load on the water resources by the coefficient of water resources use (X,,,) which is usually determined as the ratio (in %) of water use to water availability. According to the SDS, the load on water resources in the world would be lower. If the water use situation follows the CS, most regions in the world would face a catastrophically high load on water resources. On the basis of the K w computation, Table 2 shows the population distribution over the world

World water use and water availability: trends, scenarios, consequences 363 Table 2 Number of population on the Earth inhabiting regions of different rates of water use. Population number K m % 1950 1995 2025 CS 2025 SDS Number: Number: Number: Number: 10 6 % 10 6 % 10 6 % 10 6 % < 10 1197 46.4 1032 18.1 1236 15.7 1412 17.9 11-20 752 29.1 1582 27.7 2587 32.8 2411 30.6 21-40 580 22.5 722 12.7 426 5.4 936 11.9 41-60 51.4 2.0 1914 33.6 510 6.5 2363 30 >60 0.0 0.0 451 7.9 3118 39.6 755 9.6 K, K, < 10, low load on water resources = 11-20, average load = 21-40, high load = 41-60, very high load > 60, catastrophically high load in regions with different stress on the water resources in 1950, 1995 and 2025 according to the CS and SDS. In 1950 only 22% of the global population lived with a high load on water resources and there was no region in the world with a very high load. In 1995 more than 40% of the global population lived at very high and catastrophically high loads. SDS is hence a much more desirable scenario. The stress on water resources cannot fully characterize the state of water resources in a region because it does not take into account the population number and the amount 3 3 1 of water consumption. In the opinion of the authors, water availability (10 m year" per capita), obtained by dividing the renewable water resources during low-water periods minus water consumption, by the population number, is a more complete factor. Water availability values have been derived for each natural-economic region, each continent and for the whole world for 1950-1995, and for the future until 2025, according to the CS and SDS. It should be noted that water availability is subject to great space-time variability over the regions. For example, in 1995 the highest values of water availability (140-160 x 10 J m 3 year" 1 ) were observed in Canada, Alaska and Oceania. In contrast, the present water availability in densely populated regions is around 1-4 x 10 3 m 3 year" 1 per capita. In North Africa and on the Arabian Peninsula it does not exceed 0.1-0.3 x 10 3 m 3 year" 1. It should be noted that water availability of less than 2.0 x 10 3 m 3 year" 1 per capita is usually considered as very low. If it is lower than 1.0 x 10 3 m 3 year" 1 per capita this water availability is catastrophically low (Shiklomanov, 1997). As the analyses show, the water situation in 1950 was quite favourable: catastrophically low water availability was not observed in any region. By 1995 the situation had changed dramatically. Of the global population, 38% lived under conditions of catastrophically low water availability. The situation will be even worse in 2025. It is expected that the majority of the population (90%>) would live under conditions of low, very low and catastrophically low water availabilities. Moreover, it should be noted that this would occur according to both scenarios (CS and SDS) at similar population growth. This means that any future stabilization or even a decrease of water use would be favourable for the stress on water resources but it would not greatly affect water availability, which mainly depends on the population number.

364 Igor A. Shiklomanov & Jeanna A. Balonishnikova CONCLUSIONS From the research results, there are two projections for water use development in the world. If water management is developed according to the CS, the global water use up to 2025 would exceed 5000 km J year" 1, i.e. it would be about 38% higher than that in 1995. The stress on water resources would increase dramatically and water availability would decrease dramatically (58% of the global population would have very low and catastrophically low water availabilities). If water management is developed according to the SDS, global water use by 2025 would be practically stable and the stress on water resources would be much lower in many regions. Even in this case, the specific water availability in many regions, however, would tend to be lower due to population growth. Thus, the following two measures should be taken for a multi-purpose solution of water problems in regions with restricted water resources. First, new technologies for freshwater saving should be used by all water users. Secondly, the availability of water resources should be increased by runoff reservoir control, by using long-term freshwater storage, salt water desalination, water transfers, and other measures. Such an approach would not only reduce the stress on water resources but also counteract the decrease in water availability in many regions of the world. REFERENCES Rijsberman, F. R. (ed.) (2000) World Water Scenarios: Analyses. Earthscan, London, UK. Shiklomanov, I. A. (ed.) (1997) Assessment of Water Resources and Water Availability in the World. WMO/Swedish Environmental Institute. Shiklomanov, I. A. (2000) Appraisal and assessment of world water resources. Water International 25(1 ), 11-31. Shiklomanov, I. A. (ed.) (2003) World Water Resources at the Beginning of the 21st Century. Cambridge University Press, UK (in press). Shiklomanov, I. A. & Balonishnikova,.1. A. (2002) Water use in the world: tendencies, scenarios, consequences (in Russian). Proceedings of the Rosgidromet Scientific Conf. vol. 2 (St Petersburg, April 2002), 10-13. WMO, SEI (1997) Comprehensive Assessment of the Freshwater Resources of the World, p. 34. WMO/Swedish Environmental Institute.