3

Human Geography
and Water Resources

Water use and scarcity are functions of water supply and demand. This chapter focuses on the demand for water, with an emphasis on the social factors affecting water use and availability: population growth, distribution, and migration; types and distribution of water use including irrigation; clean water and sanitation access; infrastructure; and institutions. Finally it includes a discussion of methods for measuring and managing water scarcity.

POPULATION DISTRIBUTION
AND MIGRATION

The study region includes some of the most densely populated areas on Earth as well as some of the least densely populated regions, with a stark increase in population density as one moves from the mountains to the ocean. As of 2010, India alone was home to about 1.2 billion people, while Bangladesh and Pakistan each contained about 149 million and 174 million people, respectively. Nepal had a smaller population of about 30 million, while Bhutan was home to about 726,000 people (United Nations, 2011a).

The water dependencies and vulnerabilities of populations living in lower-lying areas will be quite different from the dependencies and vulnerabilities of populations in higher-lying ones. However, demographic compositional data are collected through national censuses and surveys and are reported by administrative units, such as provinces and states, which typically do not conform neatly to geographic features of interest, such as river basins or zones of elevation. By using a spatial population model of gridded population counts, though, it is possible to at least estimate the current population of each country living in the watersheds of interest in this study. This can be further segmented into coarse elevation zones (low, below 1,000 m; high, above 4,000 m; and moderate, between 1,000 and 4,000 m). In this report, the Committee uses both Global Rural Urban Mapping Project (GRUMP) (CIESIN, 2004) and Landscan (2010) population data to estimate populations at risk, relying on the former when estimating urban populations1 and the latter when estimating population distribution per se.

Nepal and Bhutan are wholly contained within the Ganges and Brahmaputra river basins, and most of Pakistan is found within the Indus Basin. Bangladesh lies within the Ganges, Brahmaputra, and Meghna basins; and India has territory within all of these basins, though much of India’s land area also falls outside the main study watersheds (Figure 3.1). Note, as discussed in Chapter 1, that the population density of the Himalayan Endorheic basins is very small, and therefore the Committee focused the discussions in this chapter on the Indus and Ganges/Brahmaputra basins. Estimates of the shares of national populations living in these basins are presented in Table 3.1.

In 2010, approximately 195 million people lived in the Indus Basin, with 16 percent of them (31.9 million) living at elevations above 1,000 m (Table 3.2). In contrast, nearly all (97 percent) of the more than

_________

1 GRUMP data rely primarily on nighttime lights to delineate urban areas and therefore provide a cross-nationally consistent basis by which to measure urban areas (Balk, 2009).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 49
3 Human Geography and Water Resources W ater use and scarcity are functions of water model of gridded population counts, though, it is possible supply and demand. This chapter focuses to at least estimate the current population of each country on the demand for water, with an emphasis living in the watersheds of interest in this study. This can on the social factors affecting water use and availability: be further segmented into coarse elevation zones (low, population growth, distribution, and migration; types below 1,000 m; high, above 4,000 m; and moderate, and distribution of water use including irrigation; clean between 1,000 and 4,000 m). In this report, the Com- water and sanitation access; infrastructure; and institu- mittee uses both Global Rural Urban Mapping Project tions. Finally it includes a discussion of methods for (GRUMP) (CIESIN, 2004) and Landscan (2010) measuring and managing water scarcity. population data to estimate populations at risk, relying on the former when estimating urban populations1 and POPULATION DISTRIBUTION the latter when estimating population distribution per se. AND MIGRATION Nepal and Bhutan are wholly contained within the Ganges and Brahmaputra river basins, and most of The study region includes some of the most densely Pakistan is found within the Indus Basin. Bangladesh populated areas on Earth as well as some of the least lies within the Ganges, Brahmaputra, and Meghna densely populated regions, with a stark increase in basins; and India has territory within all of these basins, population density as one moves from the mountains to though much of India's land area also falls outside the the ocean. As of 2010, India alone was home to about main study watersheds (Figure 3.1). Note, as discussed 1.2 billion people, while Bangladesh and Pakistan each in Chapter 1, that the population density of the Hima- contained about 149 million and 174 million people, layan Endorheic basins is very small, and therefore the respectively. Nepal had a smaller population of about Committee focused the discussions in this chapter on 30 million, while Bhutan was home to about 726,000 the Indus and Ganges/Brahmaputra basins. Estimates people (United Nations, 2011a). of the shares of national populations living in these The water dependencies and vulnerabilities of popu- basins are presented in Table 3.1. lations living in lower-lying areas will be quite different In 2010, approximately 195 million people lived from the dependencies and vulnerabilities of populations in the Indus Basin, with 16 percent of them (31.9 mil- in higher-lying ones. However, demographic composi- lion) living at elevations above 1,000 m (Table 3.2). tional data are collected through national censuses and In contrast, nearly all (97 percent) of the more than surveys and are reported by administrative units, such as provinces and states, which typically do not conform 1 GRUMP data rely primarily on nighttime lights to delineate neatly to geographic features of interest, such as river urban areas and therefore provide a cross-nationally consistent basis basins or zones of elevation. By using a spatial population by which to measure urban areas (Balk, 2009). 49

OCR for page 49
50 HIMALAYAN GLACIERS: CLIMATE CHANGE, WATER RESOURCES, AND WATER SECURITY FIGURE 3.1 Population distributions in the HKH region. Population data were taken from the Global Rural/Urban Population Map- ping Project. Dashed lines indicate disputed political boundaries, following the guidance issued by the U.S. State Department, and this Committee takes no position on these boundary disputes. TABLE 3.1 Shares of National Populations Living in 600 million people in the Ganges/Brahmaputra Basin Hindu-Kush Himalayan Region Water Basins live at elevations below 1,000 m.2 Most of those living National below 1,000-m elevation are in India (79 percent) and Population Bangladesh (18 percent; Table 3.3). Even in Nepal, Population National Living in Country Basin in Basina Populationb Basins (%) where 29 million people live in this basin, almost two- Afghanistan Indus 10,636,154 31,412,000 33.9 thirds live below 1,000-m elevation. Even in mountain- ous Bhutan, more than one-quarter of its population Bangladesh Gang/Brahm 103,326,928 148,692,000 69.5 in the Ganges/Brahmaputra basin lives at an elevation Bhutan Gang/Brahm 699,847 726,000 96.4 below 1000 m. China Gang/Brahm 1,712,145 On a global scale, the countries of this region Indus 35,585 1,348,932,000 0.13 encompass some of the world's poorest and least devel- India Gang/Brahm 466,738,395 oped areas, alongside areas that are also experiencing Indus 36,074,708 1,224,614,000 38.1 rapid economic growth. The region is characterized by Nepal Gang/Brahm 28,951,851 29,959,000 96.6 relatively low shares of the population living in cities. Pakistan Indus 148,104,460 173,593,000 85.3 a Landscan (2010). 2 In Bangladesh, furthermore, nearly 60 percent of the population b United Nations (2011a). lives within 10 m of sea level and contiguous to the seacoast.

OCR for page 49
HUMAN GEOGRAPHY AND WATER RESOURCES 51 TABLE 3.2 Population in the Ganges/Brahmaputra and Indus Basins, by Elevation and Country Country/Basin 4,000 m Afghanistan Population in Indus Basin 1,701,185 8,901,679 33,290 Share of basin population living in elevation zone 16.0% 83.7% 0.3% Bhutan Population in Ganges/Brahmaputra Basin 193,974 496,887 8,986 Share of basin population living in elevation zone 27.7% 71.0% 1.3% Bangladesh Population in Ganges/Brahmaputra Basin 103,326,928 -- -- Share of basin population living in elevation zone 100.0% -- -- China Population in Ganges/Brahmaputra Basin 1,433 866,630 844,082 Population in Indus Basin -- 1,084 34,501 Share of basin population living in elevation zone 0.08% 49.6% 50.3% India Population in Ganges/Brahmaputra Basin 459,157,952 7,502,996 77,447 Population in Indus Basin 26,539,558 9,431,329 103,821 Share of basin population living in elevation zone 96.6% 3.4% 0.04% Nepal Population in Ganges/Brahmaputra Basin 19,239,788 9,604,421 107,634 Share of basin population living in elevation zone 66.5% 33.2% 0.4% Pakistan Population in Indus Basin 134,747,024 13,159,600 197,836 Share of basin population living in elevation zone 91.0% 8.9% 0.1% Total population in Ganges/Brahmaputra Basin 581,920,075 18,470,934 1,038,149 Total population in Indus Basin 162,987,767 31,493,692 369,448 TOTAL 744,907,842 49,964,626 1,407,597 SOURCE: Data from Landscan (2010). In 2000, Pakistan had the greatest percentage--one- rapidly, with rates of urban growth that exceed those third--of its population living in cities. Despite rela- historically found in the West. In all countries in this tively low levels of urbanization, this region is home to region, the average annual rate of change of the urban some of the world's largest cities: Dhaka (Bangladesh), population exceeds 2 percent, in Nepal and Bhutan, Delhi (India), Kolkata (India), Mumbai (India), Kara- the rates exceed 4 percent per year (United Nations, chi (Pakistan), Lahore (Pakistan).3 Cities are growing 2011b). Nevertheless, the pace of urbanization in this region is not historically unusual (NRC, 2003a). 3 One thing that makes Asia stand out, particularly the subre- Urbanization may also affect water demand. Wealthier gions dependent on the water resources of the Indus and Ganges/ Brahmaputra deltas, is the number of extremely large cities, or the and more urban populations, for example, have differ- phenomenon of urban "giganticism" (Preston, 1979). ent dietary possibilities and preferences than their rural TABLE 3.3 Urban Population Projections as Percent of Total Population, by Country 2010-2050. 2010 2015 2020 2025 2030 2035 2040 2045 2050 Bangladesh 28.07 30.80 33.89 37.35 41.045 44.84 48.69 52.57 56.415 Bhutan 34.71 38.54 42.40 46.23 49.97 53.62% 57.22 60.75 64.17 India 30.01 31.72 33.89 36.56 39.75 43.30 46.93 50.58 54.23 Nepal 18.62 21.58 24.78 28.18 31.74 35.47 39.38 43.42 47.56 Pakistan 35.90 37.67 39.88 42.53 45.62 49.07 52.54 55.98 59.37 SOURCE: Based on data from United Nations (2011b).

OCR for page 49
52 HIMALAYAN GLACIERS: CLIMATE CHANGE, WATER RESOURCES, AND WATER SECURITY counterparts, and changes in diet (e.g., toward consum- less able to move away from it" (GO-Science, 2011). ing more meat) can have implications for agricultural That said, those who migrate are by no means better off water demand. than the native population in their places of destination. Rural dwellers tend to be poorer than their urban New migrants often do not have access to land, hous- counterparts, but there is also substantial urban poverty ing, safe water and sanitation, or public services (e.g., and slum-dwelling in this region. Water supply and schooling or health care). wastewater services in cities and towns are often under- developed, and many cities in the region are located in Projected Demographic Trends areas of high flood potential, exacerbated by poor urban drainage during monsoons (Moser and Satterthwaite, Key population-related questions of interest for this 2008). Urban poverty itself often exacerbates the devel- study include whether certain watersheds and elevation opment of systems for delivery of water and sanitation, zones will experience higher rates of population growth as well as access to those resources when such systems than others, and how the demographic composition of are in place ( Johnstone, 1997). those specific areas will change. Existing demographic Nevertheless, cities serve as major magnets for methods, however, do not allow one to make such fine- migrants. Most migration takes place within country, grained projections. That being said, one thing that we between cities themselves, and between rural areas do know about the distribution of future population is and cities. International migration is a small share of that countries in this region will become increasingly total migration, although some borders are effectively urbanized (Table 3.3) and that cities (which are pre- more fluid than others (e.g., between Bangladesh and dominantly at lower altitudes) will continue to absorb India).4 migrants in search of economic and other opportuni- Although temporary migration has long been a ties. Migration has the potential to serve as an adaptive strategy used by families and households to accom- strategy to environmental change--even as the cities modate climate variability (especially flooding), there that receive the migrants cope with their own set of is little evidence that climate change or other envi- economic challenges and environmental stresses and ronmental factors (with the possible exception of vulnerabilities (Hardoy et al., 2001). However, many long-term drought) result in major migration flows complex factors beyond environmental change con- (McLeman and Smit, 2006; Tacoli, 2009). To the tribute to migration. contrary, many countries in the region are known Country-level population projections are limited for their high adaptive capacity. Yet, theoretically it by the fact that they do not have any population- is possible that long-term changes to cropland and environment interactions or feedback loops built in. losses of livelihood could make it more difficult to This will matter more as projections are made fur- earn a living from farming. Such "livelihood fragility" ther into the future. Nevertheless, it can be said with could affect migration over a much longer time span, substantial confidence that the populations of the albeit through a pathway that is less direct than a single countries in this region will grow considerably over climate-related event or change in the environment the next few decades. India alone is projected to add (Raleigh, 2011). nearly half a billion people to its population between Migrants tend to be younger and better educated 2010 and 2050 (Table 3.4). Larger populations will than those who do not move, and it is these nonmovers result in increased demand for water and may exac- who may be regarded as most vulnerable. A recent study erbate problems of resource scarcity and vulnerability. found that migration has implications not just for the Over the longer term, though, rates of population mover but for those left behind: "Poorer households are growth are projected to become smaller over time (see likely to be `trapped' in circumstances where they are Table 3.5). The slowing of population growth will be at once more vulnerable to environmental change and driven by declines in fertility rates, which are in turn caused by, among other things, rising standards of liv- 4 The bulk of international migrants from Bangladesh, Bhutan, ing, decreases in childhood mortality rates, and better India, Nepal, and Pakistan stay within the region (DRC, 2007). access to family planning.

OCR for page 49
HUMAN GEOGRAPHY AND WATER RESOURCES 53 TABLE 3.4 Population Projections by Country (thousands) 2010-2050 2010 2015 2020 2025 2030 2035 2040 2045 2050 Bangladesh 148,692 158,317 167,256 175,195 181,863 187,103 190,934 193,344 194,353 Bhutan 726 784 829 867 899 924 943 956 962 India 1,224,614 1,308,221 1,386,909 1,458,958 1,523,482 1,579,802 1,627,029 1,664,519 1,692,008 Nepal 29,959 32,581 35,164 37,653 39,943 41,977 43,749 45,257 46,495 Pakistan 173,593 189,648 205,364 220,609 234,432 246,789 257,778 267,240 274,875 SOURCE: Based on data from United Nations (2011b). PATTERNS OF WATER USE tion, it follows that a conversion from natural land cover to crops will decrease the amount of runoff that makes Types of Water Use its way into surface water or groundwater. There are many ways to measure water use, and care For water to be useful, it must be available when must be taken when comparing different metrics. The and where it is needed. Issues of water timing are often most common type of water use measurement, and the critical for both blue and green water use. For instance, subject of this report, is the quantity of water withdrawn many climates have a rainy period when available or consumed for human purposes from surface water or water exceeds human demand and a dry period when groundwater. This type is sometimes called "blue water" human demand exceeds available water. Because of the use (Falkenmark and Rockstrom, 2006; Hoekstra et al., mismatch between the natural hydrological cycle and 2011). Water withdrawal is defined as the amount of human needs for water, infrastructure is often built to water withdrawn from surface water or groundwater for help societies manage water; for example, many times some human use. Some of this water may be returned reservoirs are constructed to store water during wet to surface water or groundwater after use. Water con- periods for use during dry periods. Similarly, ground- sumption is typically defined as the amount of water water is often used in the HKH region when surface withdrawn from surface water or groundwater that is supplies are insufficient because of seasonality of water not returned to the system, usually because it is lost to flows or short-term droughts. evapotranspiration, incorporated into a final product, Water must also be of sufficient quality to be of or contaminated too badly to be reused. use, and different water uses require different quali- "Green water" use is defined as the amount of ties. Some industrial and urban water uses require rainwater used by natural vegetation, land cover, and high-quality waters; agriculture often uses lower qual- agricultural production primarily through rain-fed ity waters. When water is highly polluted because of crops (Hoekstra et al., 2011). Both green water and high salinity or concentrations of human or industrial blue water (irrigation water) are used for agricultural wastes, water quality can become the limiting factor for production in the HKH region. If agricultural devel- water availability--in some parts of the HKH region, opment replaces natural land cover with crops with a pollution may reduce overall water availability. Some- higher level of evapotranspiration than natural vegeta- times "gray water" use is also measured. In the water TABLE 3.5 Population Growth Rates, by Country, 2010-2050 2010-2015 2015-2020 2020-2025 2025-2030 2030-2035 2035-2040 2040-2045 2045-2050 Bangladesh 6.47% 5.65% 4.75% 3.81% 2.88% 2.05% 1.26% 0.52% Bhutan 7.99% 5.74% 4.58% 3.69% 2.78% 2.06% 1.38% 0.63% India 6.83% 6.01% 5.19% 4.42% 3.70% 2.99% 2.30% 1.65% Nepal 8.75% 7.93% 7.08% 6.08% 5.09% 4.22% 3.45% 2.74% Pakistan 9.25% 8.29% 7.42% 6.27% 5.27% 4.45% 3.67% 2.86% SOURCE: Based on data from United Nations (2011b).

OCR for page 49
54 HIMALAYAN GLACIERS: CLIMATE CHANGE, WATER RESOURCES, AND WATER SECURITY footprint system of water use accounting, gray water is use data are also available for China, but they are not defined as the amount of water needed to safely dilute considered here except in the context of the portions a contaminant or problematic compound to a con- of the HKH watersheds shared by China. centration at which the water will be usable for some One rough measure of water supply is the per- purpose, which is increasingly important at a river basin capita availability of water (the total average renewable scale (Hoekstra et al., 2011; note that this is a different water supply of a region divided by the population of concept from municipal gray water, or sullage). that region). Another is the ratio of water withdraw- Any alterations to the water quantity or quality also als (as an indicator of demand for water) to renewable have the potential to negatively affect freshwater biodi- water availability (as an indicator of water supply). versity or the natural ecosystem services they provide. Of the five countries, Bhutan has the largest supply Often, a certain minimum environmental flow is calcu- of water, both relative to water withdrawals and on a lated for a watershed and its watercourses, which if left per-capita basis. Total internal renewable water supply in surface water and groundwater will be sufficient to is 78 billion m3 yr-1, and because Bhutan is located maintain freshwater biodiversity and crucial ecosystem high in the Himalayas, there is no significant flow into processes (Poff et al., 2010). Necessary environmental the country from elsewhere. Total per-capita available flows vary significantly between basins depending on water remains high but has fallen slightly over time the hydrology and ecology. with population, from 174,000 m3 person-1 yr-1 in 1992 to 109,000 m3 person-1 yr-1 in 2009 as its population Key Trends in Water Use has grown. Total water withdrawals were 0.34 billion m3 yr-1 in 2009, or around 470 m3 person-1 yr-1. In general, the Committee found water use is As a downstream nation, Bangladesh has greater greater relative to natural runoff in the Indus Basin than reliance on external water supplies relative to water in the Ganges/Brahmaputra Basin, consistent with withdrawals (Figure 3.2). Total internal renewable the findings of other studies (e.g., UNEP, 2008). At a water supply is about 100 billion m3 yr-1, supplemented national level, water use relative to availability is lowest by another 1.1 trillion m3 yr-1 of water that flows in in Bhutan and Nepal, intermediate in Bangladesh and from outside the country. Total per-capita available India, and greatest in Pakistan. water has fallen from approximately 10,000 m3 person-1 yr-1 in 1992 to 7,700 m3 person-1 yr-1 in 2008. Total National-Level Water Use Statistics water withdrawals were 36 billion m3 yr-1 in 2008, or around 220 m3 person-1 yr-1. Comprehensive and consistent data on water avail- India's water withdrawals are closer to water avail- ability and water use are not available for the HKH ability than most other countries in the region (Figure region, or most other regions of the world (Gleick, 3.2). Total internal renewable water supply is 1,400 2011). One of the primary sources of data for gross billion m3 yr-1, supplemented by another 640 billion water availability and withdrawals is the Aquastat data- m3 yr-1 of water that flows in from outside the country. base maintained by the Food and Agriculture Organi- Total per-capita available water has fallen from 2,100 zation of the United Nations (FAO, 2011). Aquastat m3person-1 yr-1 in 1992 to 1,600 m3person-1 yr-1 in data come from a variety of sources, estimates, and 2008. Total water withdrawals are increasing from 500 surveys, including national-level estimates submitted billion m3 yr-1 in 1990 to 760 billion m3 yr-1 in 2010, by national governments. Because of the lack of water or from around 560 m3 person-1yr-1 in 1990 to 640 availability and water use data, the Aquastat database m3 person-1 yr-1 in 2010. India contains several large contains the most widely used estimates of these quanti- watersheds and these national average figures hide sig- ties and is used here to assess broad patterns and trends, nificant regional variations. In some basins, total water but should not be considered scientifically rigorous. As demands are approaching, or may have reached, the discussed earlier in this chapter, Bangladesh, Bhutan, limits of renewable water availability. In other basins, India, Nepal, and Pakistan have substantial portions of water use already depends on nonrenewable extraction their area contained within the watersheds that make of groundwater resources, which is unsustainable in up the study area. National-level water availability and the long run.

OCR for page 49
HUMAN GEOGRAPHY AND WATER RESOURCES 55 FIGURE 3.2 Available per-capita water and per-capita water withdrawal for countries in the study region. Countries are arranged from greater relative water stress (left) to less relative water stress (right). Note the logarithmic scale on the y-axis. Available per-capita water is shown as a stacked bar-chart for two points in time, showing the portion of available water that originates within a country (internal, black) or that originates in headwaters located in another country (external, gray). The per-capita withdrawal information is shown as a white triangle for the most current time point. PAK= Pakistan, IND= India, BAN= Bangladesh, NEP= Nepal, and BHU= Bhu- tan. For Nepal, per-capita withdrawal information was available only for the year 2002. For India, per-capita withdrawal information was linear interpolated between the years 2002 and 2010, to estimate the value in 2008. SOURCE: Based on data from FAO (2011). Nepal has a substantial supply of water in both capita basis have fallen from around 1,300 m3 person-1 absolute terms and relative to water withdrawals, on a yr-1 in 1990 to 1,000 m3 person-1 yr-1 in 2008. Pakistan per-capita basis (Figure 3.2). Total internal renewable would not be able to meet its water withdrawal demand water supply is about 200 billion m3 yr-1, supplemented without external water resources, primarily from rivers by another 12 billion m3 yr-1 of water that flows in from that flow from India into the Indus, as allocated by the outside the country. Total per-capita available water has Indus Waters Treaty. fallen from approximately 10,500 m3person-1 yr-1 in National water availability and use data for China 1992 to 7,300 m3 person-1 yr-1 in 2008 because of popu- are uninformative for this study, because they are lation growth. Total water withdrawals were 10 billion dominated by regions of the country outside the m3 yr-1 in 2000, or around 390 m3 person-1 yr-1 in 2002. HKH study area. Nevertheless, China is the upstream Pakistan's internal water supplies are very limited nation on major watersheds of the HKH region; the relative to water withdrawals, on a per-capita basis Tibetan Plateau encompasses the origins of the Ganges (Figure 3.2). Total internal renewable water supply is (whose headwaters are mostly in India), the Brahma- only 55 billion m3 yr-1, supplemented by another 180 putra (which joins the Ganges in Bangladesh), and the billion m3 yr-1 of water that flows in from outside the Indus Rivers. Thus, China can influence the long-term country. Population growth has reduced total per-capita regional management of these rivers. In addition, the available water, which has fallen from 1,900 m3 person-1 Chinese government has announced plans for "leap- yr-1 in 1992 to 1,300 m3 person-1 yr-1 in 2008. Total frog development" in the Tibetan Plateau, including water withdrawals have increased from 160 billion m3 improving the economic status of farmers and herders yr-1 in 1990 to 180 billion m3 yr-1 in 2008, but on a per- and providing services such as education, health, and

OCR for page 49
56 HIMALAYAN GLACIERS: CLIMATE CHANGE, WATER RESOURCES, AND WATER SECURITY social security--with the express purposes of achieving area is for increased irrigation water withdrawals to put stability and national unity (Xinhua News, 2010). Min- increasing strain on hydrological supplies of available ing, which requires water (see Xiang [2010] for a study water. See Box 3.1 for a historical overview of early and of mining effects on river water quality in Tibet), and modern irrigation systems. hydroelectric dams are part of planned development in Irrigated area in the study region is shown in Figure the Tibetan region (Watts, 2010). In addition, Chinese 3.3. According to the Global Map of Irrigated Area plans to divert some regional water resources from the (Siebert et al., 2007), which is nominally for around the south to its northern regions will also play a role in year 2000, irrigation is widespread in the Indus Basin, future regional water politics (Turner, 2011). Further with a total of 15 million ha equipped for irrigation, analysis of how climate factors, including glacial retreat, the vast majority of which is in Pakistan. However, may influence Chinese needs for water and Chinese the national- and basin-level data are rough estimates. water policies is needed. According to FAO Aquastat data, Pakistan had about 16 million ha equipped for irrigation in 1990, which Water for Irrigation had risen to 20 million ha by 2008. Total Pakistan agricultural water withdrawal went from 150 billion In most countries, the largest sector of water use is m3 yr-1 in 1991 to 170 billion m3 yr-1 in 2008, or from the agricultural sector, which withdraws, and consumes 9,600 m3 ha-1 yr-1 in 1991 to 8,600 m3 ha-1 yr-1 in 2008. through evapotranspiration, large quantities of water Another useful large-scale dataset (Hoekstra and for irrigation. The general trend throughout the study Mekonnen, 2011) provides estimates of water con- BOX 3.1 Early and Modern Irrigation Systems The history of irrigated agriculture in the region goes back to development of modern submersible pumps allowed the pumping of Harappan civilization during the Bronze Age in the greater Indus Basin groundwater from great depths. Simultaneously, small inexpensive and adjacent areas. Early irrigators relied on shallow wells, flood inun- pumps were developed that allowed growers to lift water to irrigate small dation canals, and simple diversions from adjacent streams to water farm acreage (Shah, 2009). farmlands in arid and semiarid regions. With time and the advent of Again, historically water was applied through simple flood ir- canals, surface waters could be delivered to arable lands that were less rigation technologies where fields were flooded at times of high water. proximate to surface-water courses. Canal irrigation expanded most Flood irrigation was replaced in many parts of the world by furrow and dramatically in South Asia during the 19th and 20th centuries. Canals basin irrigation that allowed the grower more control over the timing also allowed irrigators some control over the timing and quantity of water and quantities of water application. Over the last 75 years of the 20th applied to cropland. Groundwater initially became a source of irrigation century, closed conduit irrigation systems were increasingly developed supply with the development of simple water lifting devices powered and employed. These systems, which include various types of sprinklers by humans and animals. The use of groundwater for irrigation allowed and, later, drip and sprinkler irrigation systems, allow water to be ap- lands to be irrigated in dry seasons during drought periods when surface plied with great precision. These systems are particularly well adapted supplies were diminished or absent altogether. to irrigate loose agricultural soils where control of infiltration rates is The Persians were early groundwater innovators who developed very important in determining irrigation efficiency, and have substantial the Qanat or Karez system of tapping and conveying groundwater. This additional potential. By contrast, surface application systems such as system required the drilling of multiple wells into alluvial water-bearing furrows and basins are best suited for farming operations on tight soils formations in foothill environments and then constructing a tunnel that where infiltration rates are low and tend to be more constant, particularly conveyed water underground to elaborate surface-water distribution where precision land leveling and watercourse improvements were systems on the plains. This clever system, which is still used in some adopted. Modern irrigation technology has resulted in more extensive locations in the Middle East and South Asia (Mustafa, 2011), permitted and more efficient irrigation, particularly when linked with institutional the extraction of groundwater by gravity and avoided the need to lift the and management reforms. Yet, in many instances, such as South Asia, water from the wells. This system required a high level of community it has increased the pressure on both surface water and groundwater cohesion to operate effectively. sources and even threatened the long-term sustainability of the basic Much later electrical and diesel pumps were developed that al- water resource (e.g., because of soil and water salinity; CAST, 1988; cf. lowed water to be lifted from wells in substantial quantities. Early pumps extensive publications on irrigation in South Asia by the International were limited in the depth from which they could lift water but the later Water Management Institute).

OCR for page 49
HUMAN GEOGRAPHY AND WATER RESOURCES 57 sumption by basin for irrigation of different crops. unreasonable, therefore, to consider further ground- Around 120 billion m3 yr-1 of irrigation water were esti- water development and use as a potential response to mated to be consumed through evapotranspiration in surface-water shortfalls that might ultimately occur the Indus. Total consumptive water use for irrigation is because of changes in the rates and magnitudes of approximately 100 billion m3 yr-1 for the Ganges, much glacial melt. greater than the 1.4 billion m3 yr-1 for the Brahmapu- Today, groundwater use is the focus of what Shah tra. More than a third of irrigation water consumed is (2009) characterizes as a "colossal anarchy." The need used for wheat, with rice taking just under a third of to feed a growing population and agricultural markets the remaining water. Compared with the other basins has resulted in intense pressure to farm arable land in the study area, a relatively large amount of irrigated as extensively and intensively as possible. In India, water consumption in the Indus Basin is for cotton pro- this has caused a transition away from the established duction, along with rice, sugarcane, and wheat. In the large-scale irrigated agriculture that relied on water Brahmaputra Basin, by comparison, around 75 percent deliveries through canals and irrigation systems that of irrigated water was used to grow rice. were managed in a more centralized bureaucratic way. Irrigation is very common in the Ganges and The combination of land scarcity, the availability of less common in the Brahmaputra Basin. The overall small, inexpensive pumps that can be used to extract Ganges/Brahmaputra Basin supported 29 million ha groundwater, and subsidized electricity has led to a of area equipped for irrigation around the year 2000, situation in which approximately 22 million farmers primarily in India but also in Bangladesh. Nepal has in India, for example, are pumping groundwater in an much more limited irrigated area because of its climate individualistically competitive fashion to intensively and topography. According to Aquastat data, India had irrigate relatively small plots of land. The resultant a total of 62 million ha equipped for irrigation in 2001, levels of groundwater overdraft are significant. Thus, which had risen to 66 million ha by 2008. Total Indian for example, Rodell et al. (2009) estimated that between agricultural water withdrawal increased from 560 bil- 2002 and 2008, groundwater extraction in three Indian lion m3 yr-1 in 2000 to 690 billion m3 yr-1 in 2008, or states (Rajasthan, Punjab, and Haryana) exceeded from 9,000 m3 ha-1 yr-1 in 2000 to 10,400 m3 ha-1 yr-1 recharge by 18 billion m3 yr-1 (Rodell et al., 2009). This in 2008. Bangladesh had 5 million ha equipped for estimate is consistent with the empirical data referred irrigation in 2008. Total Bangladeshi agricultural water to in the previous chapter that documents the magni- withdrawal was 32 billion m3 yr-1 in 2008, or around tude of groundwater overdraft. It is clear that current 6,300 m3 ha-1 yr-1. Assuming the average Indian appli- levels of groundwater overdraft cannot be sustained and cation rate of 9,000 m3 ha-1 yr-1 in 2000 is accurate for efforts to "tame the anarchy" have not been especially the irrigated area in the Ganges/Brahmaputra Basin, successful to date (Shah, 2009). total withdrawals in 2000 were 260 billion m3 yr-1. Groundwater is managed in accord with the dic- There is very little irrigation in the upstream tates of irrigated agriculture. As in other parts of the nations of the Himalayan plateau region (Figure 3.3). world, the price of water includes no scarcity value The only exception seems to be near Koko Nor (Qing- for the water itself and consists solely of the costs of hai) Lake in the Tibetan Plateau and the far northern extraction. Even extractive costs are at levels less than portions of the study area, which includes parts of their true value (or cost) because the cost of the energy northwest Gansu Province of China. Because this small has been subsidized in various ways by governments. amount of irrigated area is dwarfed by irrigated areas This results in an array of water prices that significantly elsewhere in China, the Aquastat statistics on overall understate the real value of the water, sending, in turn, irrigation in China are not informative. false signals to consumers about its relative scarcity. Groundwater and surface water are frequently That is, the prevailing prices signal irrigators that water substitutes for each other. There are many historical is more plentiful than it is in fact. The consequence examples in which emerging shortages in surface- is that cropping patterns are not always appropriate water supplies have been offset by increased reli- and the quantities of water applied are more than they ance on groundwater and vice versa. It would not be would be if the water was appropriately priced.

OCR for page 49
58 HIMALAYAN GLACIERS: CLIMATE CHANGE, WATER RESOURCES, AND WATER SECURITY FIGURE 3.3 Fraction of the land equipped for irrigation in the HKH region. Irrigation is widespread in both the Indus and Ganges/ Brahmaputra basins. A relatively large amount of irrigated water consumption in the Indus Basin is for cotton production. In the Brahmaputra Basin, by comparison, irrigation water use is dominated by rice production, while in the Ganges Basin, irrigated water is used primarily for wheat production. At least two factors militate against bringing extrac- with groundwater to poverty (Shah, 2009). Thus, India's tions into some reasonable balance with recharge. First, groundwater economy is very likely to become a source the sheer numbers of extractors mean that the transac- of additional demands for surface water as elements tion costs associated with any of the conventional forms of the groundwater resource become economically of groundwater regulation such as pump taxes (prices) exhausted. Indeed, the consequences of such economic and supervised pumping quotas would be enormous exhaustion are thought to be so severe that a contested and would outweigh any likely benefits of such regula- plan has been proposed to link the Himalayan rivers tion. Second, groundwater extractors could collectively with the peninsular rivers of India, which would allow represent a significant political force that would resist the import of some 200 km3 annually to southern and any attempts by government to require or induce western India to offset the effects of overdraft (Shah, reductions in the quantities of groundwater extracted 2009; Supreme Court of India, 2012; cf. Iyer, 2012). or energy subsidized. An additional and potentially problematic feature of this situation is that successful Water Use in Other Sectors efforts at groundwater regulation, whether by pricing, pump quotas and other means such as the control of In most countries, the sector that makes the most complementary inputs, would condemn large numbers withdrawals after agriculture is the municipal and of people who currently survive by irrigating small plots industrial sector. In some regions, withdrawals for

OCR for page 49
HUMAN GEOGRAPHY AND WATER RESOURCES 59 electric power production and cooling are predominant, which considered only urban agglomerations of more though urban withdrawals are increasingly significant. than 50,000 people. This dataset estimates 41.5 mil- Figure 3.4 shows the location of the major cities within lion urban dwellers in Pakistan in 2000. Dividing the the study area. Most major cities in Pakistan are within Aquastat value for municipal withdrawal in 2000 of 6.4 the Indus Basin or draw their water from the Indus billion m3 yr-1 by just this urban population, that is a use (e.g., Karachi). Kabul, Afghanistan, is also in the upper rate of 154.2 m3 urbanite-1 yr-1. The GRUMP database headwaters of the basin. Aquastat data estimate 2.5 bil- shows 34.3 million urban dwellers in the Indus Basin, lion m3 yr-1 of withdrawals for municipal purposes in which with the above use rate would imply 5.3 billion Pakistan in 1991, rising to 9.7 billion m3 yr-1 in 2009, m3 yr-1 of municipal withdrawals in the Indus Basin in or from around 20 m3 person-1 yr-1 in 1992 to 54 m3 2000. This approach, however, ignores cities that are person-1 yr-1 in 2008. However, the per-capita statis- outside the Indus Basin but draw water from it through tics from Aquastat divide municipal water use by the canals (cf. McDonald et al., 2011b). total population of the country, not just that of urban The majority of the 87.9 million urban dwellers dwellers. in cities > 50,000 in the Ganges Basin in 2000 were in Another dataset, the Global Rural/Urban Map- India (69.6 million), with some in Nepal (2.6 million) ping Project (GRUMP) database (CIESIN, 2004), or Bangladesh (15.6 million). Total urban population maps cities for the region. The version of this dataset in India in 2000 was 234.9 million in cities > 50,000, developed for McDonald et al. (2011a) was used, and Aquastat lists total municipal withdrawals in 2000 FIGURE 3.4 Major urban areas in the HKH region. Urban water withdrawals are increasingly significant, and the locations of urban areas provide a rough estimate of the locations of most municipal and industrial water withdrawals.

OCR for page 49
62 HIMALAYAN GLACIERS: CLIMATE CHANGE, WATER RESOURCES, AND WATER SECURITY tive fashion; the Committee did not develop quantita- ments in improving irrigation efficiency. Scenarios tive scenarios of how hydrological flows will change in of future agricultural water use developed by Archer each of the basins in response to climate change. The et al. (2010) are similar, and the authors stressed the potential changes in water supply discussed earlier in necessity of increased water storage to allow for a this chapter could also affect water use. continued increase in agricultural withdrawals. More- over, agricultural withdrawal and consumption could Water for Irrigation be significantly reduced if Pakistan increased the productivity of agricultural water use, which would Trends in irrigation water use vary throughout the permit them to expand production without substantial region. In some countries, irrigated crop area has been increases in demand for water, through on-farm and increasing, which increases water use. At the same time, basin efficiency improvements, changes in irrigation in some countries average irrigation rates (m3 ha-1 yr-1) technology and management, changes in crop types, have been decreasing because of more efficient appli- or other practices. cations of irrigation water or changes in the kinds of The overall Ganges/Brahmaputra Basin had 29 crops produced. Future trends in irrigation water use million ha of area equipped for irrigation in 2000 are difficult to predict and depend on (among other (Siebert et al., 2007), primarily in India and to a lesser things) changes in farming practices, government poli- extent Bangladesh. Nepal and Bhutan have much more cies toward irrigation, and the magnitude and direction limited irrigated area because of their climate and of the effect of climate change on evapotranspirative topography, a trend likely to persist into the future. demand of crops. India's irrigated area increased by 0.9 percent annually In the Indus Basin, approximately 15 million ha between 2001 and 2008, with some of the increase of cropland was equipped for irrigation in 2000 (Sie- in the Ganges Basin and some of it elsewhere in the bert et al., 2007), mostly in Pakistan. Aquastat (FAO, country. Over the same time period, India's irrigation 2011) statistics show that between 1990 and 2008, area rate has increased by around 2 percent annually, from equipped for irrigation in Pakistan increased by 1.3 1995 to 2008, Bangladesh's irrigated area, which is in percent annually, up to 20 million ha, the vast majority both the Ganges and Brahmaputra basins, increased by of which is in the Indus Basin. If this trend continues, 2.3 percent annually. Time-series data for Bangladesh by 2030, Pakistan would have 25 million ha and by on irrigation rate are unavailable, although in 2008 2050 there would be 30 million ha. At the same time, it appeared to be lower than Indian rates, at around between 1990 and 2008, the irrigation rate fell by 0.5 6,300 m3 ha-1 yr-1. Other pertinent data reveal that percent annually, to 8,600 m3 ha-1 yr-1. Assuming this about 75 percent of total land cultivated in Bangladesh trend continues, by 2030, Pakistan's irrigation rate is irrigated by groundwater (Zahid and Ahmed, 2006) would fall to 7,500 m3 ha-1 yr-1 and by 2050 would and that groundwater has been decreasing (Shamsud- reach 6,500 m3 ha-1 yr-1. The net result of such a sce- duha et al., 2009). Thus, irrigation in Bangladesh has nario for Pakistan would be an increase in agricultural been increasing, with noticeable effects on groundwater withdrawals from 170 billion m3 yr-1 in 2008 to 190 depletion. billion m3 yr-1 in 2030 and 195 billion m3 yr-1 in 2050. Because the vast majority of irrigated area in the Assuming that in the future a similar fraction of water Ganges/Brahmaputra is in India, changes in irrigation withdrawals is consumed, water consumption under practices in India will drive changes in the Ganges/ such a scenario would increase from around 120 billion Brahmaputra area. If the irrigated area in the overall m3 yr-1 in 2000 to 128 billion m3 yr-1 in 2030 and 132 Ganges/Brahmaputra continues expanding at the same billion m3 yr-1 in 2050. rate as irrigated area is currently expanding in India, Whether such a scenario would be feasible depends by 2030 there would be 39 million ha of area equipped on whether there is enough water available to meet for irrigation and by 2050 some 48 million ha. It is not this increased need, whether there is adequate arable clear whether there is enough room in the crowded land to allow continued expansion of irrigated area, Ganges basin for such a continued expansion, nor is it and whether Pakistan can make the necessary invest- clear that there is sufficient water. If India's irrigation

OCR for page 49
HUMAN GEOGRAPHY AND WATER RESOURCES 63 rate remains at 2008 levels of around 10,400 m3 ha-1 people) by 2050. Assuming the urban use rate of water yr-1, such a scenario would increase agricultural water stays the same, expected population growth implies an withdrawals in the Ganges/Brahmaputra from 260 bil- increase in municipal water use to 13 billion m3 yr-1 of lion m3 yr-1 in 2000 to 400 billion m3 yr-1 in 2030 and municipal withdrawals in the Indus Basin in 2030 and nearly 500 billion m3 yr-1 in 2050. It seems unlikely 22 billion m3 yr-1 of municipal withdrawals in the Indus that there is enough water in the Ganges to support Basin in 2050. this level of withdrawal, particularly in the dry months In 2000, there were 88 million urban dwellers in (see discussion below). cities > 50,000 in the Ganges/Brahmaputra Basin, This scenario for the Ganges/Brahmaputra is primarily in India (70 million) and to a lesser extent similar to one developed by the 2030 Water Resources Bangladesh (16 million) (CIESIN, 2004). India had Group (2009), which predicted that, if current trends an urban use rate of 179 m3 urbanite-1 yr-1, and Ban- continue, by 2030 Indian agricultural withdrawals gladesh had an urban use rate of 155 m3 urbanite-1 yr-1. would have nearly doubled from current levels. This For India, urban population is projected to increase by report also notes that changes in the agricultural sector's 2.4 percent per year between 2000 and 2030, or by 1.7 water use also represents the most cost-effective way percent per year between 2000 and 2050. Bangladesh to reduce India's water withdrawals, especially imple- is forecast to have similarly rapid urban population mentation of measures that increase yields from India's growth, increasing by 3.1 percent per year between cropland without increasing water applied as irrigation. 2000 and 2030, or by 2.1 percent per year between 2000 Similarly, Cai et al. (2010) found that if less productive and 2050. Assuming these urban population growth areas in India were to achieve the same water productiv- rates, and that urban use rate stays the same, India's ity as more productive areas, agricultural withdrawals municipal withdrawals from the Ganges/Brahmaputra could be reduced by 31 percent without reducing agri- Basin would grow from 12 billion m3 yr-1 in 2000 to cultural production. 26 billion m3 yr-1 in 2030 and 29 billion m3 yr-1 in 2050. Similarly, Bangladesh's municipal withdrawals Municipal Water Use from this basin would grow from just over 2 billion m3 yr-1 in 2000 to 6 billion m3 yr-1 in 2030 and 7 billion Trends in the municipal sector are consistent across m3 yr-1 in 2050. Municipal withdrawals for Nepal and countries, with an increasing urban population result- Bhutan will likely increase as well, driven by increases ing in an increase in municipal withdrawals. Trends in in urban population in these countries, although their the use rate (m3 urbanite-1 yr-1) are generally not avail- municipal withdrawals will continue to be a very able as time series, although economic development is small proportion of the municipal withdrawals in the likely to increase this use rate over time. Ganges/Brahmaputra Basin. In 2000, the GRUMP database (CIESIN, 2004) This scenario of growth in municipal water demand shows 34 million urban dwellers in the Indus Basin in the Ganges/Brahmaputra Basin appears consistent (cities > 50,000 people). The vast majority of these with the scenarios developed for India by the 2030 urbanites are in Pakistan, where the average urban Water Resources Group (2009). That report predicted use rate was around 150 m3 urbanite-1 yr-1, a relatively municipal demand would double from 2000 to 2030, high rate, which with the above use rate would imply a slightly slower rate of growth than in the Commit- over 5 billion m3 yr-1 of municipal withdrawals in the tee's scenario of growth developed specifically for the Indus Basin in 2000. For Pakistan as a whole, urban Ganges. population is projected to increase by 3.1 percent per year between 2000 and 2030, or by 2.8 percent per year Industrial Water Use between 2000 and 2050 (United Nations, 2011b). If this rate of urban population growth holds for the Indus Industrial water use appears likely to increase over Basin as a whole, population might grow to nearly time, driven by increases in population and economic 85 million urban dwellers (cities > 50,000 people) by development. In the Aquastat database, there is little 2030 and 140 million urban dwellers (cities > 50,000 consistent trend in industrial water use per capita, with

OCR for page 49
64 HIMALAYAN GLACIERS: CLIMATE CHANGE, WATER RESOURCES, AND WATER SECURITY some countries increasing and some countries decreas- water scarcity will change with increases in population ing. However, even if water use per capita stays constant and water use as well as with climate change. Because at 2008 levels, it appears likely industrial water use quantitative scenarios of how the hydrological cycle will will continue to rise because of fast rate of population be affected by climate change were beyond the scope of growth in many of these countries. this report, most of the discussion of climate change is The majority of Pakistan's 1.4 billion m3 yr-1 of narrative, describing the likely direction of change. industrial withdrawals in 2008 was likely from the As noted earlier, the simplest way to define physi- Indus Basin, and if industrial water use per capita stays cal water scarcity is to take the amount of water in a the same might rise to 2.3 billion m3 yr-1 in 2030 and region and divide by the population. One common set 3 billion m3 yr-1 in 2050. India's industrial withdrawals of thresholds defines regions with more than 1,700 in the Indus are unknown, but are likely smaller than m 3 person -1 yr -1 as "water sufficient," while those Pakistan's, because few major Indian cities are located below this threshold have some degree of water stress in the Indus Basin. (<1,700 m3 person-1 yr-1), chronic scarcity (<1,000 m3 Trends in India's industrial withdrawals from the person-1 yr-1), or absolute scarcity (<500 m3 person-1 Ganges/Brahmaputra are unknown, but are likely yr-1) (Falkenmark, 1989; Falkenmark and Lindh, 1974; increasing over time. One study by the 2030 Water Falkenmark and Widstrand, 1992; Falkenmark et al., Resources Group (2009) suggests that industrial with- 1989). Using this metric, with population growth drawals for all of India will quadruple, reaching 196 (ignoring potential changes in water availability) billion m3 yr-1 in 2030. A large portion of Bangladesh's Pakistan will move from water stress in 2000 (1,400 0.8 billion m3 yr-1 of industrial withdrawal in 2008 m3 person-1 yr-1) to chronic scarcity in 2030 (900 m3 was taken from the Ganges/Brahmaputra Basin and, person-1 yr-1) and 2050 (700 m3 person-1 yr-1), even if industrial water use per capita stays the same, might without factoring in climatic changes to regional rise to 1.1 billion m3 yr-1 in 2030 and 1.2 billion m3 yr-1 hydrology. Any reductions in flow in the Indus caused in 2050. Similarly, Nepal's industrial withdrawals of 0.4 by climate change would further intensify this scarcity. billion m3 yr-1 in 2000 are all taken from the Ganges The next driest country by this metric is India, which Basin, and if industrial water use per capita stays the stays classified as water stressed but moves from 1,600 same might rise to 0.6 billion m3 yr-1 in 2030 and 0.7 m3 person-1 yr-1 in 2009 to 1,300 m3 person-1 yr-1 in billion m3 yr-1 in 2050. Bhutan uses a relatively trivial 2030 and 1,200 m3 person-1 yr-1 in 2050. By this simple amount of water for industrial purposes, a trend that is measure, and ignoring potential changes in water avail- likely to continue. ability, Bangladesh, Bhutan, and Nepal remain classed as water sufficient through 2050. Trends in Water Scarcity Another way to define physical scarcity is the ratio of water use to water availability. By this metric, and As discussed in Chapter 3, we primarily pre sent ignoring potential changes to water availability, the simple metrics of physical water scarcity. This approach Indus Basin seems the most likely of any of the study is driven by the limited data available for more sophisti- area basins to face problems of water scarcity (Figure cated measures that take into account, for example, eco- 3.6). Significant increases in irrigation water use in this nomic water scarcity or water quality issues. However, basin, particularly during the dry months of November we stress that these other issues may be important and to March, may result in essentially all flow in these deserve future study. Levels of water stress in the future months being used for irrigation. Increased irrigation will also be affected by adjustments to human behavior water use in the Ganges, particularly in the dry period or technological interventions to make existing water of November to May, may likewise result in essentially use more efficient, both of which are difficult to predict. all flow in these months being used for irrigation Even without climate change affecting water avail- (Figure 3.7). One potential response by policy mak- ability in the study area, many countries would have a ers would be to attempt to increase storage during the significant challenge providing enough water to meet monsoon season so that water would be available for their needs under traditional projections. In this sec- irrigation during the dry season. Even with increases in tion, the Committee explores how some metrics of irrigation water use by Bangladesh, the Brahmaputra

OCR for page 49
HUMAN GEOGRAPHY AND WATER RESOURCES 65 FIGURE 3.6 Natural runoff (solid line) and blue water consumed (dashed line) per month for the Brahmaputra. Note that the y-axis is a log scale. The Brahmaputra has very little water scarcity, except in February and March, although water consumption still does not exceed natural runoff during those months. The large spike in natural runoff in the period June to September corresponds to the monsoon period. SOURCE: Based on data from Hoekstra and Mekonnen (2011). Basin seems less likely to be water stressed according The picture that emerges is nevertheless one in to this metric, except for a brief dry period in February which water scarcity, be it generalized or seasonal, will and March (Figure 3.6). intensify in the coming decades. Climate change will also influence the extent and severity of intensifying FIGURE 3.7 Natural run-off (solid line) and blue water (dashed line) consumed per month for the Ganges. The Ganges has very little water scarcity, except in February and March, when water consumption exceeds natural runoff. The large spike in natural runoff in August corresponds to the monsoon period. SOURCE: Based on data from Hoekstra and Mekonnen (2011).

OCR for page 49
66 HIMALAYAN GLACIERS: CLIMATE CHANGE, WATER RESOURCES, AND WATER SECURITY scarcity. It follows, then, that one means of addressing TABLE 3.6 Share of Population with Access to Improved potential changes in hydrological circumstances that Drinking Water and Sanitation, by Country are uncertain is to ensure that current water manage- Improved Drinking Water Improved Sanitation ment practices are as effective and equitable as possible. Urban Rural Total Urban Rural Total This means making all efforts to increase and maintain Afghanistan 78 39 48 60 30 37 the productivity of water that is currently available. Bangladesh 85 78 80 56 52 53 Such productivity increases need to be sought for both Bhutan 99 88 92 87 54 65 consumptive and in-stream uses. Thus, for example, China 98 82 89 58 52 55 employment of more rational water pricing practices India 96 84 88 54 21 31 and the development of flexible schemes of water Nepal 93 87 88 51 27 31 allocation that will allow droughts and unanticipated Pakistan 95 87 90 72 29 45 shortfalls to be managed in both a timely and effec- tive way could strengthen the capacity of the region to SOURCE: Data from WHO/UNICEF (2010). manage its water resources more effectively in the face of climate change and other hydrological uncertainties. and sanitation, even when stratified by urban and rural residence, tend to mask considerable spatial variation in CLEAN WATER AND SANITATION ACCESS access. Data from the nationally representative Demo- graphic and Health Surveys (DHS),5 for example, A basic measure of water development and condi- suggest that there is greater variation in access to clean tions is access to improved drinking water and sanita- water and sanitation in rural districts than urban dis- tion systems, as measured by the United Nations on tricts. It is possible, in principle, to use these data to a regular basis. This measure forms the standard by create district-level mappings of access to clean water which the water-related Millennium Development and sanitation. This could provide finer-grained insight Goals were set. Political and social stability is affected into the geographic distribution of water-related need by the societal capacity to cope with and adapt to large- and vulnerability, including by river basin and proxim- and small-scale changes in water availability that may ity to the Himalayas. result from glacial retreat and other impacts of climate change on the hydrological system. This societal capac- MEASURING WATER SCARCITY ity, often called vulnerability or resilience, depends on conditions that promote human well-being generally. With the preceding analysis in mind, it is not sur- The water-related aspects commonly measured include prising that there are many different ways to measure water availability or scarcity (discussed in the next sec- water scarcity. Water scarcity can occur when problems tion) and clean water and sanitation, which are impor- of water quantity, water quality, or timing mean there tant for human health and general well-being. is not enough water to meet people's wants. Scarcity in Countries in Africa have the most serious problems the physical sense is often defined in terms of arbitrary in terms of the fraction of population without access but useful criteria such as those discussed in the next to these basic water services, but Asia has the largest paragraph. Economic scarcity is customarily defined absolute number of people without access (hundreds in terms of the cost of making water available and of millions), with a wide divergence among countries a willingness to pay. One significant manifestation and within countries. Table 3.6 shows the fraction of of economic scarcity is the economic exhaustion of the populations in countries of the region with access groundwater, which contrasts with physical exhaustion, to "improved drinking" water systems and "improved as explained in Box 3.2. In this section, we focus on sanitation" systems, as of 2008 (WHO/UNICEF, simple measures of physical water scarcity, primarily 2010). As these data show, urban dwellers are typically because they are easy to estimate from available data. more reliably served than rural dwellers, and access to water is typically higher than access to sanitation. 5 Data available at http://www.measuredhs.com/data/available- However, national-level estimates of improved water datasets.cfm.

OCR for page 49
HUMAN GEOGRAPHY AND WATER RESOURCES 67 yr-1), chronic scarci ty (<1,000 m3 person-1 yr-1), or BOX 3.2 Economic absolute scarcity (<500 m3 person-1 yr-1). By this simple Exhaustion 0f Groundwater measure, Bangladesh, Bhutan, and Nepal are classed as Groundwater is an important element of hydrological sys- water sufficient, and Pakistan and India are classed as tems throughout the world. It is known to be a significant, if not water stressed. As noted above, these national averages completely understood, part of the hydrology of the Himalayas, hide important regional differences. Additionally, the the Ganges Plain to the south, and the southern Peninsula of India thresholds are somewhat arbitrary, and this measure (Bookhagen, 2012; Shah, 2009). Over time, extractions of water from an aquifer must be roughly equal to recharge if the aquifer is of water scarcity does not account for water that flows to remain economically viable. Overdraft is the situation in which across a border and it ignores different uses of water in extractions exceed recharge. Overdraft itself is not problematic different climates. as long as periods of intermittent overdraft are punctuated with Another common, but more descriptive, way to periods in which recharge exceeds extractions and the aquifer consider physical water scarcity is to look at the ratio recovers. Over such periods of time the aquifer is in equilibrium. of water withdrawals or consumption to total water Circumstances characterized by persistent overdraft are not sus- tainable in the sense that extraction cannot be economically viable available, which is defined as natural streamflow by indefinitely. Where extractions are consistently greater by volume Hoekstra and Mekonnen (2011). Figure 3.6 shows than recharge, the water table falls and the elevation from which estimates of natural runoff and water consumption for water must be extracted grows. the Brahmaputra basin, on a monthly time step. This The costs of groundwater extraction include the costs of river has very little water scarcity, except in February needed energy and those costs are highly sensitive to the depths and March, and even then, water consumption does not from which the groundwater must be pumped. As the water table is drawn down through overdraft the costs of extraction rise. Ulti- exceed natural runoff. The large spike in water available mately, these costs rise to the point where it is no longer economi- in the period June to September corresponds to the cally feasible to continue pumping and the extractor or extractors monsoon period of the year. in question cease to pump. In some instances this will allow the The Ganges basin (Figure 3.7) has a similar annual aquifer to recover as recharge then exceeds extraction. There are pattern, but water consumption exceeds natural runoff other instances where the storage capacity of the aquifer is altered for February and March. There is also a sharper peak in (consolidated) or rates of recharge are very small or nonexistent, that it is no longer economical for any extractor to withdraw water. natural runoff in the single month of August. At this point the aquifer is said to be economically exhausted. Eco- Finally, the Indus Basin (Figure 3.8) has a much nomic exhaustion is not necessarily identical to physical exhaus- higher percentage of the natural runoff consumed over tion, however, because pumping depths from which extractions the whole year than the Ganges or the Brahmaputra are no longer economically feasible can be significantly smaller basin, with the percentage highest in the period Octo- than the pumping depths that would prevail if all of the water in ber to March. However, water consumption does not the aquifer were extracted. These latter circumstances constitute physical exhaustion (Glennon, 2002; NRC, 1997). exceed natural runoff in any month for the Indus, in contrast with the Ganges. More sophisticated measures of water scarcity are available in the literature. Sometimes measures of Such an approach, however, should not be taken to scarcity explicitly set aside a portion of available flow as imply that other facets of water scarcity are not also an environmental flow (EF). Sometimes issues of water extremely important to human livelihoods. quality are also considered, which can be a limiting Perhaps the most common metric of physical factor for many applications (McDonald et al., 2011a). water scarcity, and certainly the simplest to calculate, Similarly, measures of water scarcity sometimes incor- is to take the average amount of water in a region and porate information on how well water is actually deliv- divide by the population (Falkenmark, 1989; Falken- ered to people. In many cities, for instance, a substan- mark and Lindh, 1974; Falkenmark and Widstrand, tial fraction of people live in neighborhoods without 1992; Falkenmark et al., 1989). These indexes usually consistent access to safe drinking water, simply because define regions with more than 1,700 m3 person-1 yr-1 infrastructure is absent (UN-HABITAT, 2006). as water sufficient, while those below this threshold Because data are sparse, it is not clear how increased had some degree of water stress (<1,700 m3 person-1 glacial melt will affect total runoff. However, it could

OCR for page 49
68 HIMALAYAN GLACIERS: CLIMATE CHANGE, WATER RESOURCES, AND WATER SECURITY FIGURE 3.8 Natural runoff (solid line) and blue water consumed (dashed line) per month for the Indus. A higher percentage of the natural runoff is consumed than for the Ganges or the Brahmaputra basin, with the percentage highest in the period October to March. Water consumption does not exceed natural runoff in any month for the Indus. SOURCE: Based on data from Hoekstra and Mekonnen (2011). affect the timing of runoff, which would in turn affect of this report, this section highlights patterns in the the ratio of water consumption to natural runoff. For relationships between climate change and water man- example, in the Brahmaputra, if the change in runoff agement in South Asia. Institutions that focus more due to glacial melt occurs in February and March, on disasters (e.g., devastating floods, droughts, and there would be a larger change in the ratio. If the run- GLOFs), vis--vis water management more broadly, off were to change during the monsoon peak between are addressed under the rubric of environmental secu- June and September, there could also be issues with rity in Chapter 4. flooding. Increasing International Assessment WATER MANAGEMENT, INSTITUTIONS, AND HYDROCLIMATE CHANGE Considerable progress has been made in assess- ing water resources implications of climate change The water-use patterns described above are medi- in South Asia at international and national levels of ated by institutions at multiple levels in the region -- analysis over the past decade. Some advances have come from international to national, state, and local water about through regional intergovernmental organiza- management, and from mountain headwaters to plains tions (IGOs) such as the South Asian Association for and coastal environments. In addition to multiple lev- Regional Cooperation (SAARC6), whereas others stem els of management, water institutions span a range of from multilateral donor-sponsored initiatives. Regional subsectors from irrigation to domestic, industrial, and IGOs such as SAARC are devoting increasing atten- environmental uses that often have separate agencies tion to climate change, extreme events, and disaster risk housed under different ministries, and whose perspec- reduction (e.g., SAARC; cf. also Regional Integrated tives vary on hydroclimate change. Although a full Multi-Hazard Early Warning System for Africa and description, let alone discussion, of these complex water management systems lies beyond the scope 6 http://saarc-sdmc.nic.in/.

OCR for page 49
HUMAN GEOGRAPHY AND WATER RESOURCES 69 Asia [RIMES],7 which is an association of country of Water Resources (2009-2011) earlier published a hydrometeorological agency directors in the greater two-volume work that compiled Comprehensive Mission Indian Ocean region). Water management remains Documents for a National Water Mission Under National a sensitive topic that has historically been addressed Action Plan on Climate Change (Government of India, through bilateral relations, although a 2012 SAARC- 2009). Following a summary of recommendations, it Chamber of Commerce and Industry conference presented supporting documents on the current policy dealt with climate, energy, and water, and SAARC is context of water management, surface water, ground- increasingly active in addressing hydroclimate disaster water, and basin planning for climate change. preparedness. Bangladesh has joint concerns about the coastal Comparative international water resources analysis hazards of climate change, inflows from major interna- is also advancing through the work of the South Asia tional rivers and tributaries, and domestic hydroclimate Water Initiative which convenes countries in the region risks. To address these issues Bangladesh created an to analyze water management issues related to climate interagency Institute of Water Modeling chaired by change (SAWI, 2011). The World Bank has conducted the Ministry of Water Resources as a center of climate substantial analytical studies of water management and change modeling, and a Climate Change Cell11 for climate risk in several countries and basins in the region adaptation and mitigation programs including irriga- (Pahuja et al., 2010; Sadoff and Rao, 2011; Yu, 2010). tion agriculture under the Ministry of Environment. These studies are helping increase awareness, analysis, The Government of Pakistan established a Global and cooperation. Change Impact Study Centre (GCISC) that prepared Among international nongovernmental organiza- 16 research reports, including one on Climate Change tions contributing to these advances, the Asia-Pacific Implications and Adaptations of Water Resources in Paki- Network for Climate Change organized an early study stan (Ali et al., 2009), which included a section on policy of water resources impacts of climate change in South needs. The GCISC studies contributed to a Planning Asia (APN, 2004). More recently, ICIMOD has sup- Commission Task Force Report on Climate Change. The ported cooperative international programs on snow and federal cabinet adopted a climate change policy that ice hydrology in the Himalayan region (e.g., ICIMOD's has substantial water resources provisions and a new INDUS8 and HKH-HYCOS9 programs). Other inter- Ministry on Climate Change in 2012. These national national nongovernmental organizations (NGOs) have actions occur at a time of constitutional devolution of undertaken comparative case study analyses of water authority for environment, agriculture, and other sec- management in the context of climate change across tors to the provinces and political uncertainty. the region (e.g., ISET, 2008; Moench and Dixit, 2004; Nepal has been an early leader in scientific assess- see also the International Water Management Institute ments of climate change on water resources man- [IWMI] website10). agement (Gyawali, 2011). Nepal created a National Adaptation Programme of Action followed by a Nepal Formulation of National Water and Climate Change and Development Portal,12 which Climate Policies has a national branch of the Climate Action Network, youth alliances, and NGO associations. These efforts Several countries have articulated national water occurred within what has become a larger context of policies related to climate change. In 2012, India issued constitutional and governance uncertainties. a new Draft Water Policy that includes numerous refer- Bhutan has placed growing emphasis on climate ences to and a major section on adaptation to climate change and hosted the SAARC meeting on cli- change (Government of India, 2012). The Ministry mate change at Thimpu in 2010. Its water policy is situated under its Ministry of Agriculture, though it 7 http://www.rimes.int/about_overview.php. 8 http://www.icimod.org/?q=265. 9 See http://www.icimod.org/?q=264. 10 See http://www.iwmi.cgiar.org/Topics/Climate_Change/ 11 See http://www.climatechangecell.org.bd/. default.aspx. 12 See http://www.climatenepal.org.np/main/.

OCR for page 49
70 HIMALAYAN GLACIERS: CLIMATE CHANGE, WATER RESOURCES, AND WATER SECURITY also has a substantial hydropower program under its states and districts is associated with increasing refer- Department of Energy. ences to climate change in water sector planning docu- Afghanistan is concentrating on reestablishing ments at those levels, the actual implications for bud- hydrological monitoring and water resources plan- geting, management, and governance are less evident. ning after a prolonged interruption. Water policy is The largest cities in South Asia are developing climate addressed in the Ministry of Agriculture, Irrigation, adaptation plans that include water systems, but the and Land, which raises an interesting point of compari- rapidly growing proportion of secondary and tertiary son across countries where water and climate policies cities show limited evidence, capacity, or higher-level are variously situated in ministries of agriculture, water support for addressing hydroclimate effects on water resources, and environment. supply, drainage, or wastewater management. At local scales of irrigation management, however, there is Uneven Subnational, State, and Local Water substantial practical experience with hydroclimate Management Capacity adjustment. The robustness of these local forms of adaptation in relation to longer-term and larger-scale Recent decades have witnessed trends toward trends in climate variability associated with snow and devolution of water governance from national or con- ice hydrology is less well known. current jurisdiction to state and local levels (i.e., the "subsidiarity principle" in the Dublin Statement on CONCLUSIONS Water and Sustainable Development, 199213). Irriga- tion departments in India and Pakistan are situated at Key features of the human geography of the HKH the state and provincial level of government. In the case region were identified at the workshop by the break- of Pakistan, although some recent annual plans for the out groups on Hydrology; Water Supply, Use, and provinces refer to climate change, provincial budgets Management; and Demography and Security. Starting and department plans for 2012 do not yet include from those concepts, the Committee used its expert major climate change adaptation analyses, programs, judgment, reviews of the literature, and deliberation to or policies analogous to those being developed at the develop the following conclusions: national level. In comparison with government, however, the past Rural and urban poor may be most at risk, in generation of social science research on local irrigation part because the poor are least likely to be able to ret- management by farmers in South Asia underscores rofit, move, or rebuild as needed when faced with risks. their keen perception of and adjustments to hydro- However, the environment is not the only driving factor climate variability (see, e.g., numerous IWMI field for migration and other adjustments. research studies of irrigation in South Asia;14 IUCN, Social changes in the region have at least as 2008; Kreutzmann, 2000). much of an effect on water demand as environmental It seems reasonable to summarize that at this time factors do on water supply, leading to stress. Existing international and national water organizations are stress could be exacerbated by climate change in the making significant advances toward water resources future. assessments of climate change that include snow and Future population growth in specific watersheds ice hydrology. National water agencies are devoting or elevation zones is uncertain, but there is certainty increasing emphasis to climate impact and adaptation that the region will become increasingly urbanized. assessment. It is less clear how these assessments affect Water resources management and provision of management, policy, and interagency water coordina- clean water and sanitation is already a challenge in the tion. Similarly, while devolution of water governance to countries that share the Himalayan watersheds. The adequacy and effectiveness of existing water manage- 13 http://www.wmo.int/pages/prog/hwrp/documents/english/ ment institutions is a reasonable, if coarse, indicator of icwedece.html. how the region is likely to cope with changes in water 14 http://www.iwmi.cgiar.org/Publications/index.aspx. supply.

OCR for page 49
HUMAN GEOGRAPHY AND WATER RESOURCES 71 Changes in seasonal streamflow could have Water management assessments have advanced significant impacts on the local populations by altering over the past 5 to 10 years, although their implemen- water availability patterns and affecting water manage- tation in water policies and programs is less clear; to ment decisions and policies for irrigation, municipal, date, there is limited penetration to lower levels of industrial, and environmental use. governance or support for local water managers who Average aggregate water-use data provide face the greatest risk. crude estimates of water supply, demand, and scarcity. Water use has been increasing over time in both Increasing the detail, consistency, and accessibility of the Indus and Ganges/Brahmaputra basins, and this water-use data (along with streamflow and aquifer data trend will continue for the next several decades, with as discussed in Chapter 2) are key priorities for sound irrigation by far the largest use. regional water assessment.

OCR for page 49