Figure 16.3 Population ratio under high water stress. Future projections of the population ratio under high water stress are depicted: (a) water scarcity index, where water withdrawal (W)/renewable freshwater (Q) exceeds 0.4; (b) water-crowding indicator, based on an assumption that less than 1000 m3 water availability (Q) per capita per year is water-scarce. The horizontal axis shows global-averaged surface air temperature increase in the future scenarios.
is diffi cult because extremes rarely occur. A recent result (Hirabayashi and Kanae 2009) shows preliminarily that by the end of twenty-first century, the annual flood-affected population (even during a year of less flooding) is likely to increase to about 300 million people. This corresponds to the number of people that are currently affected by a very devastating flood year.
One of the most apparent and important hydrologic changes attributable to global warming is the change in water supplies caused by the melting of snow cover and glaciers. Changes in snow cover and snowmelt are generally included in future projections of water availability because hydrologic and climate models often involve a snow submodel. Barnett et al. (2005) argued that more than one-sixth of the Earth's population relies on seasonal snow packs for their water supply, and that this population is thus likely to experience severe water stress in a future warming scenario. However, errors in the numerical simulation of snow accumulation and melting processes are still large among various snow submodels (e.g., N. Rutter, pers. comm.), and the amount of precipitation in cold mountain areas, where snow dominates, is poorly constrained. Thus, both the future projection and current estimation of land snow amounts are uncertain. In addition, the impact of glacier changes on water stress has not yet been assessed. When a glacier is lost, the stable water supply from it is also lost. Thus, glacier-related assessments are necessary for the sustainable assessment of water resources.
The water stress assessment shown in Figure 16.2 includes the amount of water withdrawal as well as water availability. The water withdrawal amount is much more difficult to estimate accurately than water availability. Basic data on water withdrawal estimates are available by country from, for example, the World Resources Institute or the Aquastat information system of the United Nations FAO. Agricultural, industrial, and domestic water withdrawals are separately recorded in these statistics. For spatially finer assessments, the amount of water withdrawal is distributed into grid boxes, such as the 0.5 (approximately 50 km) resolution in Figure 16.2, using the distributions of population and the irrigation area at the same spatial resolution as proxies.
There are two major sources of errors and uncertainty in the statistics and spatial distribution procedures. Water withdrawal statistics are not available for every year for every country and are particularly sparse for countries in the developing world. The date of the latest water withdrawal data varies among countries, as does the definition of water withdrawal. For example, the withdrawal amount of agricultural water in Japan is based on conventional water rights. The actual withdrawal value is unknown. Even in cases where water withdrawal is actually observed or estimated, measurement methods vary and are generally unknown. The accuracy of each measurement method is also unknown. Simple errors may be added during the information transfer from each local site to the final database. The spatial distributing procedure relies on proxies such as population and irrigation area, but validation is often insufficient, and it is ultimately nearly impossible to determine whether a proxy-based distributing method is acceptable or not.
There are several methods for projecting future water withdrawal and demand on a global scale (e.g., Vorosmarty et al. 2000; Alcamo et al. 2007; Shen et al. 2008). The projected future change (generally an increase) of water withdrawal is much larger than the change in water availability. The change in water availability is at most around 10%, but the change in total withdrawal might range from 50-150% by the middle or end of the twenty-first century in certain regions. The rate is higher in regions undergoing rapid development. Change in water withdrawal is primarily driven by population and socioeconomic factors, such as economic growth and technological evolution. SRES scenarios project more than a 20% increase in population from 2000 to the middle of the twenty-first century. In addition, the amount of water required per person will increase with development. An important driving factor is change in food consumption patterns. The amount of water needed to produce a unit of beef, for example, is approximately ten times larger than that needed to produce a unit of maize or wheat (Oki and Kanae 2004). Therefore, previous studies (e.g., Vorosmarty et al. 2000; Alcamo et al. 2003; Oki and Kanae 2006) have seen change in water withdrawal as the dominant factor controlling the future projection of high water stress populations, rather than change in water availability. Thus, the IPCC Technical Paper on Climate Change and Water (Bates et al. 2008) argued that "aside from major extreme events, climate change is seldom the main factor exerting stress on sustainability." We must keep in mind, however, that our imagination has its limits and that we must continue to monitor the relationship between water and climate change.
If water availability changes due to global warming have less of an impact on water stress than changes in water withdrawal, it is natural that more attention should be paid to the future projection of water withdrawal. Previous studies (e.g., Vorosmarty et al. 2000; Alcamo et al. 2003; Shen et al. 2008) have applied different sets of equations for the projection of water withdrawal, using various proxies and parameters, such as GDP and electricity consumption. A review of those different methods (N. Hanasaki, pers. comm.) revealed that the future projections vary considerably. Validations of these methods (and of the results thereby calculated) are far from adequate. However, uncertainty in future withdrawal projections does not greatly affect the main qualitative message to the general public in terms of future water stress: the number of people impacted by high water stress is likely to increase until the middle of this century. The scenario chosen is a major factor differentiating future high stress populations (Oki and Kanae 2006). Put simply, our future behavior is the key to our future conditions. However, the uncertainty of future withdrawal projections will have great importance if we evaluate sustainability rather than water stress, and if we focus more on quantitative regional details.
Measuring the Sustainability of Water Resources An Integrated Water Resources Model
As described above, the causes of high water stress can be classified into three types: (a) water is scarce, and people in the region want more, (b) the ecology is signifi cantly damaged, and (c) groundwater is in danger. To measure the sustainability of water resources, therefore, we need to measure (a) the output of human activities that require water resources, (b) the stability of aquatic ecology, and (c) the decline of the groundwater table. It is virtually impossible, however, to measure the first item solely from the viewpoint of water, because any output of human activity is a product of myriad complicated social (and natural) conditions. A possible proxy would be the yield of agricultural crops, which are highly dependent on water availability. The second item, aquatic ecology, can be measured for a specific region. However, it may be difficult to measure it continuously for a long period and on a global scale. The third item, which is literally the sustainability of the stock of water, is able to be measured on regional and global scales, but as yet only limited regional information is available. This makes it extremely difficult to measure the present and past status of sustainability from a global perspective.
Our goal, however, is to measure the future sustainability of water resources, as the future is the nature of sustainability. Interdependence, trade-offs, and interactions between resources such as water and land must be taken into account when doing so. In addition, we must overcome the apparent flaws in the assessment presented in Figure 16.2; namely, no consideration was given to green water for nonirrigated crops or to the temporal variations of water availability and demand.
A candidate tool for measuring the sustainability of water resources from past to future is an integrated numerical model of water resources, which incorporates or represents water availability, water withdrawal, agriculture production and yield, the impact on ecology, and the impact on groundwater. The applicability of future scenarios for population change, socioeconomic change, and climate change is a necessary function of such an integrated model.
Hanasaki et al. (2008a, b) have recently developed the first stage of such an integrated water resources model, although it needs further development. Specifically, the model consists of six modules for land surface hydrology, river routing, crop growth (both irrigated and nonirrigated), reservoir operation, environmental fl ow requirement estimation mainly for aquatic ecology, and anthropogenic water withdrawal. The model can run with daily time steps, at a spatial resolution of 1° x 1° (longitude and latitude). Finer spatial resolution is possible, but the output is meaningful only on a weekly or monthly, not daily, basis. The environmental fl ow requirement module was incorporated based on case studies from around the world. It provides a possible framework for assessing aquatic ecosystem sustainability (i.e., the second item noted above).
The environmental flow requirement could be calculated using the method of Smakhtin et al. (2004) instead of that employed by Hanasaki et al. (2008a, b). The spatially distributed irrigation water requirements for agricultural crops yields are computed. Domestic and industrial water requirements are computed in advance by the method of Shen et al. (2008), which could be replaced by another method. The computation of these three (agricultural, industrial, and domestic) water requirements provides a possible framework for the assessment of the first item: the demands of human activities. Given that domestic and industrial water requirements tend to have a higher priority than irrigation, the degree to which the irrigation water requirement is fulfilled (thereby affecting crop yields) would become the measure of the first item. The model uses the water supply from major artificial reservoirs and dams, as well as that from unsustainable groundwater, and can provide a possible framework for the assessment of unsustainable groundwater decline (i.e., the third item). Another similar model (Alcamo et al. 2003) exists, and comparable functions can be expected if there is further development of that model as well. In addition, an analogous model (Rost et al. 2008) is used to compute green water as well as the implicit impact of unsustainable ground water.
It should be stressed that all kinds of observed and statistical data (e.g., the amount of water withdrawal, crop yield, irrigated area, river discharge, and groundwater depth) are necessary for calibrating, validating, and assimilating an integrated water resources model. The development of a successful model requires input from a limited number of top researchers; however, the acquisition, management, and distribution of the necessary data necessitates a global collaborative effort. The value of data should not be underestimated.
Potential Outcomes of Integrated Water Resources Model to Quantitative Evaluation
In this section, I present an example application and potential outcomes of such an integrated water resources model. The model used for this purpose is not the current version of Hanasaki et al. (2008a; 2008b) nor does it resemble the current version of other similar models. It represents a possible extended version of the former, which requires further modification and development. To carry out future projections, the model should be paired with a data set representing climate, population, land use, and possible socioeconomic conditions of the future. In addition, it must be capable of characterizing the adaptation of human beings to the changing environment of the future.
The integrated water resources model will enable us to measure sustain-ability and to evaluate possible solutions quantitatively. For example, when fulfi lling both environmental requirements and water demands for domestic, industrial and agriculture use, it is possible to calculate how much water withdrawal from unsustainable groundwater is necessary. If the volume of the unsustainable groundwater reservoir is already known, then it becomes possible to estimate the time of groundwater extinction. Another example is a computation that forbids the withdrawal of unsustainable groundwater as long as society's demands for water remain the same. Under this condition, the degree to which the environmental flow requirement is fulfilled, a proxy of aquatic ecological sustainability, can be evaluated. Consider a further example of a computation of the decrease in crop yield when there is insufficient irrigation water, under the condition that the environmental flow requirement is fulfilled and withdrawal of unsustainable groundwater is forbidden. In the real world, there is a mixture of insufficient crop yield, insufficient environmental fl ow, and the use of unsustainable groundwater. This mixture makes it difficult sometimes to evaluate sustainability completely, although the trade-off between the items must be represented in the model. These numerical experiments would give us an opportunity to evaluate the degree of sustainability in one dimension. Of course, holistic evaluation of the mixtures and trade-offs is necessary. Possible solutions, such as land use and green water management through better agricultural practices and control of demand, must be applicable in the experiments to enable an evaluation of these for the future.
The model must be able to evaluate the impact of infrastructure development. Currently, Hanasaki et al.'s (2008a, b) version incorporates only the major dams of the world. Thus, prominent aqueducts and water canals in the western U.S. and China, as well as minor dams, need to be incorporated into the model. The acquisition and management of GIS data on the global infrastructure are also still unresolved issues. Infrastructure developments may provide a solution to water stress; however, they might also lead to the loss of another kind of sustainability. On a global scale, major dams have already had significant impacts on large river systems (Nilsson et al. 2005; Hanasaki et al. 2006).
Finally, if the various numerical experiments are to be eventually successful, it is necessary to establish criteria to evaluate the degree of sustainability. An integrated model allows us to evaluate the services and impacts that water resources provide, rather than the simple ratio of water availability and withdrawal amount. Nevertheless, evaluating sustainability from the viewpoint of services and impacts is difficult, partly because the evaluation is subjective.
Conclusions and Outlook Critical Issues for Assessing the Sustainability of Global Water Resources
There is considerable uncertainty in the measurement of present water availability and withdrawal on a global scale. Future projections of water availability and withdrawal are even more uncertain. Currently, projections of future water availability rely on numerical climate model simulations, the errors and uncertainty of which are still under debate. Uncertainty is generally larger for water withdrawal than for availability, and data on water withdrawal for model calibration and validation are too sparse. Future projections of water withdrawal rely on the projections of future population, lifestyle, industrialization, technology, and land use for the regions in question. Because irrigation water is the major component of withdrawal, data on land use changes, especially those involving irrigation, are vital. The water stress assessment depicted in Figure 16.2 was made possible by the global distributed irrigation area data developed by Siebert et al. (2005). Future irrigation area distribution might also be derived from these data. To project future irrigation areas, several kinds of socioeconomic projection data are necessary. However, future irrigation information is itself necessary for the projection of future socioeconomic conditions. Thus, an interactive projection system involving both socioeconomics and irrigation is necessary. Such a system has not yet been developed. For this, we need to combine a socioeconomic model and an integrated water resources model. The model discussed by Rosegrant et al. (2002) and its extension are candidates that incorporate many necessary items; however, it also appears to lack several components of distributed natural and anthropogenic water cycles, represented in the models developed by Hanasaki et al. (2008a, b), Rost et al. (2008), and Alcamo et al. (2003). Further development is therefore expected. In addition, similar to global climate models for climate change issues, several different models should be developed to obtain a better range of information and uncertainty.
As emphasized, observed and statistical data (particularly on water withdrawal and related topics) are extremely important to model development. Currently, the level of human-oriented and natural hydrologic cycle (e.g., river discharge, precipitation, snow, ice, and frozen ground) data is insufficient. Continuous monitoring is indispensable, yet observation networks have declined over recent years. A globally coordinated effort is thus necessary to make such data sets available to researchers and decision makers worldwide.
Although quantitative uncertainty has been emphasized, there has been a considerable improvement in the data, knowledge and modeling methodologies of global water cycles over recent decades. Thus, the global view and conclusions shown in or derived from analyses, such as Figure 16.2, are sufficient to examine our possible future and to consider necessary actions on a global scale (Oki and Kanae 2006). Similar to the current status of climate change projections led by IPCC, reducing uncertainty will be of great benefit for better regional assessments and for considering and taking necessary actions on a regional or finer scale.
While geographically distributed information is necessary, sustainability should not only be evaluated by region. A global perspective is often very important because virtual water trade is possible, and indeed common, in the modern world. Nevertheless, environmental flow and other water requirements should be fulfilled within each region to strengthen the regional economy, social welfare, and ensure ecological sustainability. It should not be forgotten that virtual water trade is not versatile; poor countries cannot import what they need. The incorporation of virtual water trade into a water resources assessment is highly challenging.
Gaining a perspective over a longer time frame is very important. For example, it is probably acceptable to exploit fossil water in the era of maximum global population projected for the first half of the twenty-first century. However, the minimum condition for accepting such exploitation is that society will move toward sustainable water usage and finally cease withdrawing unsustainable groundwater before the sources are exhausted.
The use of water does not necessarily imply the consumption of bulk water as a substance. Gravity and water temperature are major resources for industrial purposes. Water quality is the resource in the case of domestic water. Then, treated water is usually released again into the terrestrial hydrologic cycle. Therefore, the quality, temperature, and gravity of water should be incorporated into integrated water resources models and utilized to assess water stress and sustainability. This has not yet been attempted. Among water quality issues, groundwater contamination threatens to become a serious problem on a global scale. Thus, modeling and projection studies need to be conducted.
The projections made so far lack the aspect of "ethics," which should not be neglected. Future projections of water withdrawal are made as extensions of our current status, similar to the so-called business-as-usual scenario. Usually, developed countries do not have the option of drastically decreasing water withdrawal. Developing countries are usually not projected to become high consumption countries. Although such projections are performed for practical purposes, we need to consider to what extent luxury lifestyles are acceptable and how the world should be from the viewpoint of globally equal social welfare and quality of life. Since a major part of water resources are used in food production, just how much water is necessary for the entire world depends on food consumption—specifically, the logistical distribution system for food, the rate of waste food, and diet (meat- vs. vegetable-based). Changes in behavior can alter the necessary amount of water far more rapidly than can changes in natural water availability. An example of the demand was projected for the case of China (Liu and Savenije 2008), from the viewpoint of water footprint. However, nobody knows whether the projected future will accommodate sustainability. Definitions of sufficient environmental flow require an ethical viewpoint because it is possible for a small portion of anthropogenic activities to interrupt natural environmental conditions.
"Ethical" considerations include questions such as: Which kind of solutions is acceptable, and to which degree? For example, the degree to which modification, usually accompanying construction of structures, of natural land and hydrologic systems is acceptable is an ethical consideration. The construction of structures having a local impact involves a trade-off relationship with virtual water imports, which necessarily has impacts in remote areas. This trade-off is also a target of ethical considerations. Trade-offs impact the distribution of water between upstream and downstream areas. When limited water availability occurs in a river basin, the balance between up- and downstream is an ethical consideration; in the real world, this situation usually becomes a political issue.
It should be noted that an exclusive viewpoint of "less water consumption is better" or "doubling water productivity" is misleading. Rice paddies, for example, are a major land use type in Southeast and East Asia, apparently consuming large amounts of water. However, they are naturally induced and thus inevitable. The wide, flat plains of Southeast and East Asia are alluvial floodplains, which were naturally very wet, like marshland. Thus, rice paddy fi elds with surface ponding water were the only possible dominant land use until modern drainage technology emerged. Similar examples can be found all around the world.
Land planning and management are inevitably closely related to the sustainability of water resources. Major reasons for this are that the price of water should be low and vast amounts should be available. In addition, green water management has become an important issue. One analysis (Rockstrom et al. 2007) argues that a continuous expansion of nonirrigated agricultural area is necessary over the near future if we are to reach the Millennium Development goals of the UN for hunger. As discussed in this chapter, the aquatic ecosystem is damaged by water withdrawal. The green water concept reminds us that terrestrial ecosystems, in addition to aquatic ecosystem, could also be damaged by our water use, either through the degradation of water quality or through degradation/change of water quantity. The treatment and reuse of water may seem preferable; however, we should not forget the energy that would be needed. Thus, water is primarily a site-dependent resource—it is attached to the land. The integrated model system that I have described must be a primary tool for the simultaneous planning of water and land use. In addition to land use change determined by human activities, future projections need to consider that the anticipated global warming could alter the natural landscape of the world and the distribution of appropriate places for each crop. Sometimes, the virtual water concept has a close relation to land resources. Japan, for example, is one of the heaviest virtual water-importing countries. This is not due to a lack of water, but rather because there is not enough fl at cropland in Japan; thus, the amount of virtual water imported is very large.
Energy is becoming more closely related to water resources. It was conventionally believed that only hydropower had a close relationship to water resources. Currently, hydropower occupies a small portion of global total electricity generation, and it is not expected to contribute a major portion of the total in the future. However, two additional items have recently appeared that demand consideration: biofuel and the cooling of power plants by water. Biofuel is currently a controversial topic, and new opinions are released almost daily. Hence, I will avoid a detailed discussion here. Although the first generation of biofuels has not been positively received, due to its total carbon budget and close relation to food security, the possibility remains that another, improved type of biofuel will become more widely accepted and prevail. In such a case, additional water supplies, as well land, will most likely be needed. The amount of land or water will, of course, depend on the type of biofuel. The integrated water resources model has the capacity to assess the necessary water for biofuels and the sustainability of water.
Water for cooling, and of course for hydropower, may become a serious concern, even in developed countries, when climate change occurs. Hightower and Pierce (2008) reported that a severe drought in France in 2003 caused the loss of up to 15% of the nuclear power generation capacity for five weeks as well as a loss of 20% of the hydropower capacity. The 2007 drought in eastern Australia raised similar concerns. In both developing and developed countries, energy requirements may increase in the future, and more water for cooling may become necessary. Even without additional energy requirements, a rise in droughts caused by global warming may decrease the power generation capacity. These issues have not yet been taken into account in global water resources assessments and projections.
Development of an integrated water resources model is necessary because conventional water stress assessment is not sufficient for measuring the sustainability of global water resources. Efforts to acquire and manage data need to be enhanced and continued. Moreover, an understanding of how to apply the model and evaluate the results is crucial for measuring the sustainability. Although continuous efforts for model improvement and data acquisition are desirable, we cannot expect a perfect model or set of data. Ultimately, we need to interpret subjectively the results and take actions; however, interpretation is not solely a matter for science.
Water interacts strongly with other resources, because water is widely and inexpensively available all over the world. Therefore, considerably more effort is needed to measure the sustainability of water as a resource that intimately interacts with other resources.
Overleaf (left to right, top to bottom): Fabian Dayrit, Marco Schmidt, Mohamed Tawfic Ahmed Klaus Lindner, Heleen De Wever, Johannes Barth Thomas Ternes, Thomas Knepper, Paul Crutzen Motomu Ibaraki, Group discussion, Shinjiro Kanae
Was this article helpful?
Your Alternative Fuel Solution for Saving Money, Reducing Oil Dependency, and Helping the Planet. Ethanol is an alternative to gasoline. The use of ethanol has been demonstrated to reduce greenhouse emissions slightly as compared to gasoline. Through this ebook, you are going to learn what you will need to know why choosing an alternative fuel may benefit you and your future.