The Effects of Population on the Depletion of Fresh Water


This article compares quantitative estimates for groundwater loss and glacier recession and considers the significance of their relative magnitudes. It concludes that the effect of food and agriculture, hence of population, may be significantly greater than that attributable to the global warming caused by industrial production and transport.

 NASA's Soil Moisture Active Passive (SMAP) mission will track Earth's water into one of its last hiding places: the soil.: Soil moisture data from SMAP will aid in predictions of agricultural productivity, weather and climate. High-resolution global maps of soil moisture produced from SMAP data will inform water resource management decisions on water availability around our planet. SMAP is scheduled to launch in November 2014 from Vandenberg Air Force Base, Calif., on a Delta II rocket. The mission is managed by NASA's Jet Propulsion Laboratory in Pasadena, California. Image courtesy of NASANASA's Soil Moisture Active Passive (SMAP) mission will track Earth's water into one of its last hiding places: the soil.: Soil moisture data from SMAP will aid in predictions of agricultural productivity, weather and climate. High-resolution global maps of soil moisture produced from SMAP data will inform water resource management decisions on water availability around our planet. SMAP is scheduled to launch in November 2014 from Vandenberg Air Force Base, Calif., on a Delta II rocket. The mission is managed by NASA's Jet Propulsion Laboratory in Pasadena, California. Image courtesy of NASA

By Robert J. Wyman

Although the international community is focusing on global warming as the paramount environmental threat, the scarcity of fresh water for agriculture may emerge as a more immediate problem. Almost one-fifth of the world’s population lives in areas where water is physically scarce (World Health Organization 2009). Rapid population growth, as well as increases in per capita food consumption and changes in the composition of the diet, has led to an enormous expansion of agriculture, which is now estimated to account for 92 percent of the world’s use of fresh water (Hoekstra and Mekonnen 2012).

Water is not used efficiently. In addition to the fundamental human need for food and fiber, water scarcity is exacerbated by a number of technological, economic, and political factors. For example, the government of India subsidizes water for rice farming in arid areas, and Kazakhstan supports cotton farming in the desert surrounding the Aral sea (Micklin 2007). The Southwest United States has similar misallocation problems. Political factors, like the use of cheap water to purchase the political loyalty of farmers, make these misallocations notoriously difficult to fix. Moreover, some water simply cannot be metered and priced owing to technical hydrological issues.

In much of the world, rainfall is insufficient and unreliable. The natural irregularity of precipitation is smoothed by the storage of water in two major compartments, glaciers and groundwater (the general term for all subsurface water) and, in much smaller amounts, in lakes or behind dams. In many regions, spring rains initiate the growing season, but much of the growth occurs during the dry season and is therefore heavily dependent on irrigation using groundwater and river-borne glacial meltwater. In South Asia, the monsoons allow a wet-season crop, but irrigation is necessary for the dry-season crop.

Buckwheat in Sangdak Village: Photograph by Sam CowanBuckwheat in Sangdak Village: Photograph by Sam Cowan

The major storage compartments for fresh water are being rapidly run down. Groundwater is depleted by irrigation. Glaciers are depleted by melting, a global warming effect, and by the deposition of black carbon (soot), discussed below. Data are now available that can separately measure the depletion of water by these two distinct routes. This article compares quantitative estimates for groundwater loss and glacier recession and considers the significance of their relative magnitudes. It concludes that the effect of food and agriculture, hence of population, may be significantly greater than that attributable to the global warming caused by industrial production and transport.

Depletion of global water stores


Expansion of irrigation using water pumped from non-renewing or incompletely renewing underground stores temporarily increases the amount of water available for agriculture, but eventually the subsurface reservoirs will be depleted, reducing the water available and increasing the amount of energy that must be expended in raising it to the surface.

The fraction of farmland that is irrigated has grown rapidly. The earth’s irrigated area has increased at 2 percent per year since 1960, almost doubling between 1961/63 and 1997/99 (IPCC et al. 2008). In recent years 40 percent of the world’s food has been supplied by irrigated land (ibid.). For countries that have rainfall-limited regions, the proportion can be much larger. For instance, 70 percent of China’s grain comes from irrigated land (Yu 2011), as does the same share of India’s agricultural output (Bhaduri, Amarasinghe, and Shah 2008). A large part of progress against hunger was the result of the green revolution, which was notably dependent on a massive increase in irrigation (Bhaduri, Amarasinghe, and Shah 2009). Much of that irrigation now depends on unsustainable groundwater use (Wada 2012). 

A successful strategy to support this increase in irrigation has been the impoundment of water in artificial reservoirs behind dams. Between 1993 and 2008, the increase in reservoir storage was roughly equal to the lower estimates of groundwater loss (Church et al. 2011, Table 1). Storage thus replaced depletion. However, dam-building reached a peak in the late 1970s with the construction of 800 major dams a year and has fallen sharply in the past 30 years down to 25 to 50 per year recently, partly because of environmental concerns (Chao et al. 2008). As the demand for food and irrigation increases, water impoundment is falling behind (Wada et al. 2012). The difference between irrigation water demand and human-engineered water storage has been filled in by a tenfold increase since 1961 in groundwater extraction (Davidson and Andrews 2013). Extraction now far exceeds recharge rates, and groundwater stores are being drawn down. 

The situation of India is illustrative. The gross irrigated area of India has quadrupled since 1950 (Ministry of Agriculture, India 2013, Table 13.2). Use of groundwater has grown even faster—by more than sixfold—until it now supplies two-thirds of irrigated land (Amarasinghe, Shah, and Malik 2009; Gandhi and Namboodiri 2009). Even though many Indians migrate from villages to the cities, India’s agricultural population continues to grow at about 1 percent per year. Net sown area, however, has remained more or less constant (Bhaduri, Amarasinghe, and Shah 2009). Over the decade 1990–2000, the net sown area per person in the agriculture-dependent population has decreased by 10 percent (from 0.29 ha per person to 0.26 ha per person). Subsistence on ever-decreasing plot sizes has been made possible by the expansion of groundwater irrigation. Bhaduri, Amarasinghe, and Shah (2009) found that a 1 percent increase in rural population density was associated with a 5 percent increase in the proportion of groundwater-irrigated area in total sown area. Their conclusion was that rural population density alone significantly accounts for the expansion of groundwater irrigation. 

And India is not unique. The north-central Middle East, including portions of the fertile crescent (the Tigris and Euphrates river basins), lost 91.3 gigatons (billion metric tons or Gt) of groundwater from January 2003 to December 2009 (Voss et al. 2013).1 Focusing on a smaller region, the Central Valley of California lost, from October 2003 to March 2010, 30.9 Gt of water, essentially the total capacity of the largest reservoir in the US (Lake Mead, holding 32 Gt) (Famiglietti et al. 2011). Over large areas, water tables are falling. Declines in water tables range from centimeters to meters per year (Rodell, Velicogna, and Famiglietti 2009). Water levels in parts of the Central Valley have dropped more than 120 meters (Hiscock 2005). The breadbasket regions of the world’s most populous countries are in crisis situations (for the Indo-Gangetic plain, see below; for China see Li 2011 and Yu 2011). Because of overdrafts for agriculture, some of the world’s largest rivers (Colorado, Rio Grande, Yellow, Indus, Ganges, Amu Darya, Murray, and Nile) are now seasonally depleted so that they do not reach the sea (Postel 2010). 

Three key estimates of recent global groundwater loss are 145 ± 39 Gt per year (Konikow 2011), 204 ± 30 Gt per year (Wada et al. 2012), and 283 ± 40 Gt per year (Wada et al. 2010, 2012). All of these estimates, the average of which is 211 Gt per year, represent drastic increases over past estimated depletion. 

The icefall of Khumbu glacier, in the Nepali Himalayas…: It is …considered one of the most dangerous spots on the South Col route to Mt. Everest’s summit. Credit: NASA/GSFC/Kimberly Casey.The icefall of Khumbu glacier, in the Nepali Himalayas…: It is …considered one of the most dangerous spots on the South Col route to Mt. Everest’s summit. Credit: NASA/GSFC/Kimberly Casey.


Glaciers accrete water from precipitation and lose it through melting and calving (collapse into the ocean of the terminal edge of a glacier). Global warming (and black carbon deposition, see below) increases melting, but warmer air is also moister and thus allows more precipitation in some places. The balance for each glacier is different. However, the data are now clear that, summed over all glaciers, global warming is causing a massive retreat of the ice. While the increased melting may temporarily increase the amount of water running into rivers and thus available for irrigation, in the longer term it will reduce the amount of glacial runoff. As meltback proceeds, the decreased glacier surface area exposed to spring and summer sunlight, and the colder air temperatures at the higher altitude of the remaining glacier, will reduce the rate of melting and the amount of water available (Jansson, Hock, and Schneider 2003). The contribution of glacier melt to river flow varies in different river basins (Kaser, Grosshauser, and Marzeion 2010). In snow-dominated regions, global warming will shift the peak melting to earlier in the season and away from the time of maximum need, increasing the dependence on groundwater irrigation (Barnett, Adam, and Lettenmaier 2005).

The history of mass loss for glaciers is uncertain and appears to be variable. The longest series measuring annual mass change spans only ten years (Bolch et al. 2012). Glaciers and ice caps underwent periods of increased melting in the mid-1970s and late 1990s (Church et al. 2011), but there are recent indications that their mass loss decreased markedly beginning around 2005 (Cogley 2012; Jacob et al. 2012). Estimates of water loss for all glaciers and ice caps (not including the agriculturally irrelevant Antarctic and Greenland ice sheets2) are of approximately the same magnitude and range as those for groundwater loss cited above. Key estimates of water loss for all glaciers and ice caps, using both satellite and ground-based data, are 200 Gt per year (Cogley 2012), 354 ± 14 Gt per year (Church et al. 2011), and 148 ± 30 Gt per year (Jacob et al. 2012). The average of these estimates is 234 Gt per year. 

The range of a factor of two in these estimates has a variety of causes. Satellite data provide comprehensive measurements over broad areas, but, because of limited spatial resolution, become unreliable when the areal extent of a glaciated area is less than 5,000 km2. On the other hand, land-based measurements are sparse in time and spatial coverage. They also are biased toward smaller, land-terminating glaciers that have a more rapid melt rate (Gardner et al. 2013). Satellite mass measurements detect not only groundwater and glacier water losses, but also changes in surface and soil water, water impounded behind dams, tectonic movements, and the isostatic rebound of land freed from the weight of the ice of the last glaciation. The raw data must be corrected for these factors.

The latest global estimate of glacier losses (Gardner et al. 2013) addresses the question of why satellite measurements have shown a magnitude of glacier loss that is about half of estimates from ground-based measurements. 

The estimate reconciles land and satellite measurements (including data from the Ice, Cloud and Land Elevation Satellite). The reconciliation uses the best data from each method for each glaciated region. This consortium review, written by a number of prominent authorities and used by the IPCC (IPCC 2013a, Table 4.4), concludes that the glacier loss is 215 ± 26 Gt per year. This is essentially the same as the estimate above for global groundwater loss. Glacier melt and groundwater loss are depleting about equal amounts from the earth’s major water stores.



 NASA | Groundwater Depletion in India Revealed by GRACE [HD]

Scientists using data from NASA's Gravity Recovery and Climate Experiment (GRACE) have found that the groundwater beneath Northern India has been receding by as much as one foot per year over the past decade. After examining many environmental and climate factors, the team of hydrologists led by Matt Rodell of NASA's Goddard Space Flight Center, Greenbelt, Md. concluded that the loss is almost entirely due to human consumption.
Groundwater comes from the natural percolation of precipitation and other surface waters down through Earth's soil and rock, accumulating in aquifers - cavities and layers of porous rock, gravel, sand, or clay. In some subterranean reservoirs, the water may be thousands to millions of years old; in others, water levels decline and rise again naturally each year. Groundwater levels do not respond to changes in weather as rapidly as lakes, streams, and rivers do. So when groundwater is pumped for irrigation or other uses, recharge to the original levels can take months or years.
More than 109 cubic km (26 cubic miles) of groundwater disappeared from the region's aquifers between 2002 and 2008 -- double the capacity of India's largest surface water reservoir, the Upper Wainganga, and triple that of Lake Mead, the largest manmade reservoir in the U.S.
The animation shown here depicts the change in groundwater levels as measured each November between 2002 to 2008. 
Completed: 02 August 2009
Animator: Trent L. Schindler (UMBC) (Lead)
Scientist: Matthew Rodell (UMBC)
Data Collected: 2002 -2008
Credit: NASA/Goddard Space Flight Center Scientific Visualization Studio

Published on February 7, 2014

Depletion of major water stores in the Indo-Gangetic plain

While global estimates, as we have seen, are characterized by uncertainty, we are on firmer ground with regional-level data estimated by satellite. The Indo-Gangetic plain is home to 800 million people, more than 10 percent of the world’s population (Bolch et al. 2012; Tiwari, Wahr, and Swenson 2009). The included countries, extending from Afghanistan to Bangladesh, add approximately 26 million people annually to the world’s population and account for almost one-third of global population growth (PRB 2011). The northwest Indian region alone encompasses 114 million people (Rodell, Velicogna, and Famiglietti 2009). The Gangetic plain of India is largely agricultural, producing grain that is an important part of the supply for the entire country (Ministry of Agriculture, India 2013, Table 4.5(b)). The Indo-Gangetic plain is one of the regions most vulnerable to water scarcity. 



In 2002, NASA and the German Aerospace Center launched twin satellites in a mission referred to as GRACE (Gravity Recovery and Climate Experiment) (GeoForschungsZentrum Postsdam 2002). These satellites follow each other around the earth 16 times a day separated by about 220 km. When the lead satellite approaches a region with extra mass (like a mountain range), it is attracted forward and accelerates, increasing the distance between it and the following satellite. After the lead satellite passes the mass, it is attracted backward; at the same time, the follower satellite has entered the extra gravitational pull and speeds up, decreasing the distance between the two satellites. The reverse process happens when the satellites fly over a region with reduced mass. Microwave ranging provides precise measurements of the distance between the satellites. Analysis of changes in the gravitational data over years allows detection of significant mass movements.

The largest broad-scale loss of mass in the world (glaciers and ice sheets are considered separately) is from a 2.7 million km2 swath extending from Afghanistan to Bangladesh (see Figure 1, region A) (Tiwari, Wahr, and Swenson 2009). For somewhat different time periods and spatial regions, the loss has been estimated at 35 Gt per year (Jacob et al. 2012) and 54 Gt per year (Tiwari, Wahr, and Swenson 2009). Within the large area of region A, the most intense mass loss occurs from a 438,000 km2 region that includes the northwest Indian states of Rajasthan, Punjab, and Haryana (including Delhi) (see Figure 2) (Rodell, Velicogna, and Famiglietti 2009). Because this region encroaches on the Thar desert, more intense irrigation is needed to sustain agriculture there. Satellite data show that the annual mass loss for this area averaged 17.7 Gt (ibid.). Per unit area, this is double or triple the rate of mass loss of the larger region A in Figure 1. 


FIGURE 1 Loss of glacier and groundwater mass in South Asia

According to GRACE satellite gravity measurements, region A, from Afghanistan to Bangladesh, is losing 54 Gt per year of groundwater. This compares with region B, encompassing the most heavily glaciated mountains of the Himalayas, which is losing only 3 Gt per year; and region C, which is gaining mass slightly. Glaciers are indicated by white stippling in regions B and C. The scale on the right is rate of water loss (or gain) in cm per year (a 1 cm depth of water has a mass of 10,000 tonnes/sq km). All of regions A and B are losing water, while all of C is gaining water (except at the extreme southeast and southwest corners).


After eliminating other possibilities, the displaced mass has been identified as water (Tiwari, Wahr, and Swenson 2009). Partitioning the lost water among the various land-storage compartments is complicated, involving land-based measurements and modeling. Rainfall in the region was slightly above normal in the measured years. Changes in surface water (lakes, reservoirs, etc.) and soil water were measured and modeled; they added mass to the region. The remaining storage compartment is groundwater. Absent mass losses in the other compartments, this is where the lost water (difference between influx and efflux) came from. The amount of water loss, as measured by the satellites, is broadly consistent with the losses due to irrigation as estimated from spatially and temporally sparse land measurements. 


FIGURE 2 Region of most intense loss of groundwater

In the northwest Indian region, GRACE satellite gravity measurements show a water loss of 17.7 Gt per year, double or triple the rate of mass loss per unit area of region A in Figure 1. The darkness of the oval regions indicates the magnitude of an averaging function used to calculate the water loss rather than the amount of loss as in Figure 1. 



The water “lost” from these aquifers does not just disappear. In India, 91 percent of the groundwater extracted is used for irrigation (Ministry of Water Resources, India 2013). After its use for irrigation, some water seeps back into the aquifers. However, the measured mass loss is a net figure, the difference between discharge and refill of the aquifers. The lost water runs off into rivers or evaporates and returns as rain; much of it ends up in the oceans, adding to sea-level rise (Church et al. 2011; Wada et al. 2012).

Underground water is interspersed with solid matter: rock, gravel, sand, and the like. To convert the mass losses to the resulting drop in the water table, one must take account of the fraction of underground space that the water occupies. Using porosities appropriate for the regions, the water table is dropping 10 cm per year, averaged across the entire Indo-Gangetic plain (Figure 1, region A) (Tiwari, Wahr, and Swenson 2009), and by 33 cm per year in the more intense northwest Indian region shown in Figure 2 (Rodell, Velicogna, and Famiglietti 2009). Local ground measurements indicate more extreme drops of 70 to 120 cm per year under widespread areas of the Indus alluvial plain in Punjab (World Bank 2010) and range up to 1,000 cm per year in some Delhi measurements (Rodell, Velicogna, and Famiglietti 2009).

The total volume of water in the aquifers is unknown, so it is not possible to estimate the number of years that the aquifer water will last. Nevertheless, we know the loss is large compared to other storage compartments. For comparison with the aforementioned 35–54 Gt annual groundwater loss, the total storage capacity of the 203 main reservoirs and the largest lake in the Indian grain belt is 39.5 Gt (Rodell, Velicogna, and Famiglietti 2009).


Data from the GRACE satellites are also used to measure the melting of glaciers. Essentially all mass loss in the Himalayas is from a region bordering India and Nepal for 1,500 km and encompassing the most heavily glaciated mountains (Figure 1, region B) (Tiwari, Wahr, and Swenson 2009); this region had a mass loss of only 3 Gt per year. The remainder of the glaciated region (region C) had mass gains. The glaciated gain and loss regions together (regions B and C), facing the full front of the Indo-Gangetic plain (region A) and encompassing about an equal area, showed negligible mass change. The 35–54 Gt annual loss due to irrigation in the agricultural region south of the Himalayas (region A) is roughly 10–20 times the 3 Gt per year loss in the adjacent glaciated region (regions B and C).

The effects of food and agriculture

One might roughly attribute groundwater loss to food and agriculture via irrigation, and glacier melt to global warming via greenhouse gas emissions. The situation, however, is actually more complicated. 

Food and agriculture have been estimated to contribute between 19 and 29 percent of global anthropogenic greenhouse gas emissions (Vermeulen, Campbell, and Ingram 2012). Of this, 80–86 percent originates on farms and in livestock husbandry—specifically, nitrous oxide (N2O) emissions from soils (38 percent of the farm and livestock total), methane from enteric fermentation (32 percent), biomass burning (12 percent), rice production (11 percent), and manure management (7 percent). The contributions of the various gases are made commensurable by converting the effect of each into carbon dioxide (CO2) equivalents.

Black carbon is the soot produced when burning fossil fuels or biomass. Black carbon is the dominant absorber of visible solar radiation in the atmosphere, but it is a particulate solid and thus is not considered a greenhouse gas. A large fraction of the earth’s population produces black carbon as it burns wood, grass, and dung for energy (Kerr 2013). Globally, 60 percent of black carbon emitted is from biofuel burning in developing countries and the widespread burning of crop residues (Weaver 2011). This open burning is incomplete and produces aerosols containing other organic carbon species as well as black carbon. The organic aerosols emitted in open burning generally act as cooling agents, while the black carbon is warming.

It is uncertain whether the sum effect of open biomass burning on global warming is positive or negative (Bond et al. 2013), but at least in South Asia, black carbon appears to be a specific cause of Himalayan glacier melt (Menon et al. 2010). The Indo-Gangetic plain is the source of more than half of the 8 million tons per year of black carbon released globally (Qiu 2013; Ramanathan et al. 2007). The air heated by black carbon flows over the Himalayas and causes an estimated 60 percent of the warming which occurs in that region (Meehl, Arblaster, and Collins 2008; Ramanathan and Carmichael 2008; Ramanathan et al. 2007). In addition, deposition of black carbon on Himalayan slopes increases the melting rate of a typical glacier by up to one-third (Qiu 2013). It has been estimated that the warming effect of black carbon is as important as greenhouse gases in the observed retreat of over two-thirds of the Himalayan glaciers (Ramanathan and Carmichael 2008).

The Appendix provides more detail on the contributions of the major agriculturally produced greenhouse gases and black carbon.

The global estimates of water released by glacier melting and from groundwater irrigation are roughly similar. Groundwater is overwhelmingly used for agriculture. Glacier melting, however, is only partly driven by greenhouse gas emissions from industrial production and transport. There is a large contribution from food chain greenhouse gasses and from food chain black carbon. In an accounting that distinguishes water loss due to the need of vast populations for food from that due to fossil fuel burning for industrial production and transport, the food-related fraction of lost glacial water must be subtracted from industrial production losses and added to food-related losses. The accounting process is summarized in Figure 3. After taking account of that correction, water loss resulting from people’s need to eat may dominate the loss resulting from industrial production and transport.3


FIGURE 3 Accounting for fresh water loss 



Food, population, and affluence

Grains provide more than half the calories in the world’s diet. Globally, income elasticity for grains is small across all income levels (Amarsinghe and Singh 2009; Baffes et al. 2009; World Food Program 2008): richer people do not consume much more grain than poorer people. The total consumption is fairly closely proportional to population and not very sensitive to affluence.4

Meat is more sensitive to affluence; it has a higher elasticity that depends on income level (Baffes et. al. 2009). Meat is currently about 6.5 percent of the global diet (by weight) and is expected to rise to 7 percent by 2030.5 The FAO (2012) warns that “there is less of a meat revolution than commonly asserted, mainly because of lack of development and income growth in many countries [and] cultural and religious factors. … The perception of revolutionary change in the meat sector reflects the extraordinary performance of world production and consumption of poultry meat.” Poultry is also the most efficient converter of grain to meat.6 For chicken, when optimally produced, the feeding efficiency, kg grain fed/kg edible meat, is as low as 1.5 to 2.0 (Smil 2002).

Future prospects

A recent US Department of Agriculture report (Fuglie, Wang, and Ball 2012) argues that “improving agricultural productivity has been the world’s primary safeguard against a recurring Malthusian crisis.” In the last 50 years, agricultural output has about tripled while population and groundwater depletion have each increased by approximately a factor of 213. Early in this last half-century, agricultural growth came predominantly from additional resource inputs, including the large increase in irrigation described above. More recently, growth has been the result of technology used to increase the effectiveness of these inputs. FAO (2013) data show that the global land area used for agriculture has remained stable since 1994. Ausubel, Wernick, and Waggoner (2012) argue that, largely as a result of improvement in yields, an expansion of the area of agricultural land has so far not been needed to accommodate increased population and affluence. However, Rudel et al. (2009) conclude that agricultural intensification has not been generally accompanied by decline or stasis in cropland area. Similarly, Fuglie, Wang, and Ball (2012) assert that the rate of expansion of natural resources (land and water) has slowed only slightly.

Future agricultural productivity gains are certainly possible, but will be harder to achieve, since potential yield gain is increasingly related to greater use efficiency of solar radiation (FAO 2009b). Global wheat and rice yield increases are slowing and are now just below 1 percent per annum, while yield increase for maize is 1.6 percent. Diminishing returns may be setting in. Production efficiency gains in land use for grain production in China leveled off 30 years ago, but inputs continue to increase, including a tenfold rise in groundwater extraction to support irrigation since 1961 and a nearly 17-fold increase in fertilizer use between 1961 and 2009 (Davidson and Andrews 2013).

The FAO Expert Meeting on How to Feed the World in 2050 (FAO 2009a) concluded that agricultural production will have to increase by 70 percent (nearly 100 percent in developing countries) by 2050 to cope with a 40 percent increase in world population and to raise average food consumption to 3,130 calories per person per day. This translates into an extra billion tons of cereals and 200 million tons of meat annually (see also Smil 2002). Under this scenario, harvested irrigated land would expand by 17 percent, all in developing countries. Mainly because of slowly improving water-use efficiency, water withdrawals for irrigation would grow at a slower pace but still increase by almost 11 percent (286 Gt). 

A variety of factors may improve or worsen the situation in the future. A few obvious examples related to groundwater use include the adoption of more efficient technology currently available or development and dissemination of new technologies. Market forces may raise prices (at the cost of rising food insecurity for the poor) to limit worsening water scarcity. Public policy may promote more efficient allocation systems. Governments may adopt policies encouraging conservation of groundwater resources. A similar range of factors could affect greenhouse gas emissions and hence glacier melt. Other eventualities may add further stress. Land may be degraded by overuse and climate change. Rising incomes may facilitate increased incorporation of meat into the diet. Large-scale economic or political dislocation may increase population growth or decrease the rate of technology adoption. Political shifts may block government actions to reduce greenhouse gas emissions.

This analysis has suggested that the current world population, through its demand for food and fiber, is probably a larger contributor to the drawdown of freshwater stocks than the greenhouse gas emissions and global warming attributable to industrial production and transport. Absent major positive changes in the contexts discussed above, as population increases steadily (see the linear rise in global population from 1950 to the present in United Nations 2013, Figure 1) and as per capita consumption of both food and manufactured goods increases, we can expect further acceleration of both glacier melting and aquifer drawdown.


Appendix: Food and agriculture contributions to glacier melting

Carbon dioxide

CO2 is not a major product of the growing process. It is absorbed during plant growth, but is released during decomposition or respiration. The burning of biomass waste from agriculture and cooking over open fires are estimated to produce as much as 40 percent of global CO2 (Levine et al. 1995), but much of the CO2 is reabsorbed as the biomass re-grows. Agricultural soils and plants act as both sinks and sources of CO2. The net flux of CO2 added to the atmosphere from these processes is estimated to be small (Vermeulen, Campbell, and Ingram 2012). However, agriculture and food processing call for many inputs that require combustion of fossil fuel. Among these are fertilizer production, energy for on-farm machines, irrigation pumps, transport, processing, and refrigeration. The most recent summary is in Vermeulen, Campbell, and Ingram (2012).

Land-use conversion

There are other, off-the-farm contributions to total emissions. By 2005, agriculture had expanded to cover 37 percent of the earth’s terrestrial surface. About 80 percent of new land for crops and pastures comes from replacing forests, particularly in the tropics. Land-cover change is a major source of CO2 emissions. Considering all factors, land-cover change accounts for between 12 percent (van der Werf et al. 2009) and 17 percent (Barker et al. 2007) of total anthropogenic greenhouse gas emissions. Other important contributions come from fertilizer production and fossil fuels used in mechanized agricultural processes.


Anthropogenic methane retains about 30 percent as much heat in the atmosphere as does CO2 (IPCC 2007). Agriculture (with contributions from ruminants, rice culture, biomass burning, and organic material in waste) produces between one-half (Vermeulen, Campbell, and Ingram 2012) and two-thirds (average of the six studies cited in IPCC 2007 that include estimates of biomass burning) of total anthropogenic methane. Hence “food methane” retains about 15–20 percent as much heat in the atmosphere as does CO2.

Nitrous oxide

Direct emissions from agricultural production account for about 60 percent of global anthropogenic N2O emissions. The primary source of net emissions is the breakdown of synthetic fertilizers. Other emissions are part of the nitrogen cycle, including organic nitrogen from animal excreta, crop residue, and biological nitrogen fixation (Mosier et al. 1998).

Black carbon (soot)

Global warming is the difference between the warming caused by incoming, mostly visible, light from the sun and the outgoing, invisible, infrared radiation. Greenhouse gases absorb outgoing infrared radiation, thus preventing the release of heat from the earth. Black carbon works via the incoming sunlight. A pollution-free atmosphere is mostly transparent to visible light, leaving black carbon as the dominant absorber of visible solar radiation in the atmosphere.

Annual emissions of black carbon have been estimated to be 40 percent from fossil fuel combustion (diesel and coal), 40 percent from open burning of biomass in the location where it is grown (associated with deforestation and crop residue burning) (Bond et al. 2013), and 20 percent from cooking with biofuels (wood, dung, and crop residues) (Ramanathan and Carmichael 2008). Hence the majority (60 percent) is associated with the production and consumption of food while the minority (40 percent) is due to industry and transport. 

After CO2, black carbon is the largest contributor (about two-thirds that of CO2) to global warming (Bond et al. 2013; Ramanathan and Carmichael 2008). The 60 percent of black carbon that is food associated then has a magnitude equal to about 40 percent of the global warming effect of CO2.

A new comprehensive survey of black carbon’s effect on global warming (Bond et al. 2013) finds that soot is warming the climate about twice as fast as estimated by the IPCC (2007). It concludes that global atmospheric absorption attributable to black carbon should be increased by a factor of almost 3 in most models (Bond et al. 2013). It also argues that black carbon may be playing a significant role in the retreat of mountain glaciers, particularly in the Himalayas, and may also have contributed to the marked decline in Eurasian springtime snow cover since 1979, a finding that has not been reproduced by models that only account for warming attributable to greenhouse gases.

Aside from its effect via warmed air carried over the Himalayas, discussed above, even very small quantities of black carbon deposited over snow and ice reduce albedo (reflectance) and enhance solar absorption (Flanner, Zender, and Rasch 2007). Much of the albedo-reducing effect occurs through local melting and refreezing of the snow into larger grains (Doherty et al. 2010). The magnitude of black carbon’s contribution to global average surface warming is estimated to be three times greater than that of CO2 (Flanner, Zender, and Rasch 2007; Ramanathan and Carmichael 2008). Other organic (brown) carbon adds to the effect (Doherty et al. 2010).


I thank Themba Flowers for preparing the maps and flow chart.

1 1 Gt is the mass of 1 cubic km of water.

2 The loss from Antarctica and Greenland is comparable in magnitude to the glacier loss discussed in this article (Church et al. 2011; Gardner et al. 2013; Shepherd et al. 2012).

3 The two routes of loss have very different properties in the time domain. The need for food each year is responsible for all of the annual water loss via irrigation. If irrigation were halted for a year, there would be no drawdown of aquifers in that year. On the other hand, the current level of global warming is the result not just of this year’s emissions, but of the long-term atmospheric accumulation of greenhouse gases. If the world stopped emitting greenhouse gases for a year, glacier melting resulting from global warming would continue unabated (IPCC 2013b). In short, while the food needs for each year are responsible for the full annual amount of irrigation, it has taken since the beginning of the industrial revolution to build up enough atmospheric greenhouse gases to cause the current annual amount of glacial melt. 

4 This article has not reviewed the enormous Impact = Population . Affluence . Technology (I=PAT) literature, but a quick survey of major recent studies confirms that changes in population size and affluence are the principal drivers of change in anthropogenic environmental stressors (Dietz, Rosa, and York 2007; Rosa and Dietz 2012). Some other widely postulated drivers (e.g., urbanization, economic structure, age distribution, increased education and life expectancy) appear to have little effect. Other variables (population aging and urbanization) have less than proportional effects (O’Neill et al. 2012). Evidence for the role of yet other factors (trade, culture, and institutions) is ambiguous (Rosa and Dietz 2012). Analyses of historical trends tend to show that CO2 emissions from energy use respond almost proportionately to changes in population sizes (O’Neill et al. 2012). Another summary finds that greenhouse gas emissions rise at least proportionately with population and possibly as rapidly as the square of population (Rosa and Dietz 2012). A comprehensive study including data from 86 countries over 45 years (Jorgenson and Clark 2010) found that regression coefficients relating CO2 production to population were large and more than twice as high as those relating CO2 emission to gross domestic product per capita. This relationship held for both less and more developed countries and was stable over time. A study of OECD countries confirms that, for richer countries, population has about double the effect of income per capita (Menz and Welsch 2012).

The I=PAT literature estimates how much of a change in environmental impact is caused by a percent change in one of the causative variables (Dietz et al. 2007). The present article undertakes a quite different task. It does not focus on changes in population or affluence. Instead, it compares current water losses from the total needs of the current population for food with the water losses due to the current rate of global warming stemming from the use of fossil fuel for industrialized production and transport. It is differentiated from previous work by attempting to account for total current losses rather than incremental losses due to historical change or projected future change.

5 Percents are calculated from table 2.5 in Alexandratos and Bruinsma (2012), with carcass weights converted to consumed weight by a reduction of one-third, roughly in line with the range of estimates found in Smil (2002).

6 Many food animals graze on open, often unirrigated land or are fed from refuse; these do not add to water needs. Other animals are grain fed for at least part of their life, and that grain is converted into a lesser weight of meat (Smil 2002).


Alexandratos, N., and J. Bruinsma. 2012. World Agriculture Towards 2030/50: The 2012 Revision. Food and Agriculture Organization.

Amarasinghe, Upali A., Tushaar Shah, and R.P.S. Malik. 2009. “India’s water futures: Drivers of change, scenarios and issues,” in Upali A. Amarasinghe, Tushaar Shah, and R.P.S. Malik (eds.), Strategic Analyses of the National River Linking Project (NRLP) of India. Colombo, Sri Lanka: International Water Management Institute, pp. 3–24.

Amarsinghe, Upali A. and O.P. Singh. 2009. “Changing consumption patterns of India: Implications on future food demand,” in Upali A. Amarasinghe, Tushaar Shah, and R.P.S. Malik (eds.), Strategic Analyses of the National River Linking Project (NRLP) of India. Colombo, Sri Lanka: International Water Management Institute, pp. 131–146.

Ausubel, Jesse H., Iddo K. Wernick, and Paul E. Waggoner. 2012. “Peak farmland and the prospect for land sparing,” Population and Development Review 38(Supp.): 221–242.

Baffes, John, Dominique van der Mensbrugghe, Andrew Burns, and Israel Osorio. 2009. “The macroeconomic environment for agriculture and rural areas to develop over the next 50 years, Table: Income elasticity for grains is small even at low income levels,” Expert Meeting on How to Feed the World in 2050, FAO, Rome.

Barker, Terry et al. 2007. “Technical Summary. Climate Change 2007: Mitigation of Climate Change,” Contribution of Working Group III to the Fourth Assessment Report of the IPCC.

Barnett, T.P., J.C. Adam, and D.P. Lettenmaier. 2005. “Potential impacts of a warming climate on water availability in snow-dominated regions,” Nature 438: 303–309.

Bhaduri, Anik, Upali Amarasinghe, and Tushaar Shah. 2008.” Future of foodgrain production in India,” International Water Management Institute,

———. 2009. “Groundwater expansion in Indian agriculture: Past trends and future opportunities,” in Upali A. Amarasinghe, Tushaar Shah, and R.P.S. Malik (eds.), Strategic Analyses of the National River Linking Project (NRLP) of India. Colombo, Sri Lanka: International Water Management Institute, pp. 181–196.

Bolch, T. et al. 2012. “The state and fate of Himalayan glaciers,” Science 336: 310–314.

Bond, Tami C. et al. 2013. “Bounding the role of black carbon in the climate system: A scientific assessment,” Journal of Geophysical Research: Atmospheres 118: 5380–5552. 

Chao, B.F., Y.H. Wu, and Y.S. Li. 2008.” Impact of artificial reservoir water impoundment on global sea level,” Science 320: 212–214.

Church, John A. et al. 2011. “Revisiting the Earth’s sea-level and energy budgets from 1961 to 2008,” Geophysical Research Letters 38: 1–8.

Cogley, Graham. 2012. “The future of the world’s glaciers,” in Ann Henderson-Sellers and Kendall McGuffie (eds.), The Future of the World’s Climate. Elsevier, pp. 197–222.

Davidson, Debra J. and Jeffrey Andrews. 2013. “Not all about consumption,” Science 339: 1286–1287.702 Populat ion and the De pl e t ion of Fre sh Wat er 

Dietz, Thomas, Eugene A. Rosa, and Richard York. 2007. “Driving the human ecological footprint,” Frontiers in Ecology and the Environment 5: 13–18.

Doherty, S.J., S.G. Warren, T.C. Grenfell, A.D. Clarke, and R.E. Brandt. 2010. “Light-absorbing impurities in Arctic snow,” Atmospheric Chemistry and Physics 10: 11647–11680.

Famiglietti, J.S. et al. 2011. “Satellites measure recent rates of groundwater depletion in California’s Central Valley,” Geophysical Research Letters 38: 1–4.

FAO. 2009a. “The resource outlook to 2050,” in Expert Meeting on How to Feed the World in 2050. Food and Agriculture Organization, pp. 1–33.

———. 2009b. “Can technology deliver on the yield challenge to 2050?” in Expert Meeting on How to Feed the World in 2050. Food and Agriculture Organization, pp. 1–46.

———. 2012. “World agriculture towards 2030/2050: The 2012 revision,“ Food and Agriculture Organization, pp. 1–147.

———. 2013. “World + (total) agricultural area (all years) in FAOSTAT database,“ Accessed 6 November 2013.

Flanner, Mark G., Charles S. Zender, and Philip J. Rasch. 2007. “Present-day climate forcing and response from black carbon in snow.” Journal of Geophysical Research 112: 1–17.

Fuglie, Keith, Sun Ling Wang, and V. Eldon Ball (eds.). 2012. Productivity Growth in Agriculture: An International Perspective. Wallingford, UK: CAB International.

Gandhi, V.P. and N.V. Namboodiri. 2009. “Groundwater irrigation in India: Gains, costs and risks,” Working Paper #2009-03-08, Indian Institute of Management, Ahmedabad.

Gardner, Alex S. et al. 2013. “A reconciled estimate of glacier contributions to sea level rise: 2003–2009,” Science 340: 852–857.

GeoForschungsZentrum Postsdam. 2002. “GRACE newsletter,“

Hiscock, Kevin. 2005. Hydrogeology Principles and Practice. Blackwell Publishing.

Hoekstra, Arjen Y. and Mesfin M. Mekonnen. 2012. “The water footprint of humanity,” Proceedings of the National Academy of Sciences, USA 109: 3232–3237.

IPCC. 2007. “Intergovernmental Panel on Climate Change 4th Assessment Report.” Working Group I: The Physical Science Basis.

———. 2013a. “Intergovernmental Panel on Climate Change 5th Assessment Report.” Working Group I: The Physical Science Basis. 

———. 2013b. “Approved summary for policymakers,” Twelfth session of Working Group 1. Intergovernmental Panel on Climate Change. SPM–1 to SPM–36.

IPCC, Bryson Bates, Zbigniew Kundzewicz, Shaohong Wu, and Jean Palutikof. 2008. “Climate change and water,” Technical Paper of the Intergovernmental Panel on Climate Change.

Jacob, Thomas, John Wahr, W. Tad Pfeffer, and Sean Swenson. 2012. “Recent contributions of glaciers and ice caps to sea level rise,” Nature 482: 514–518.

Jansson, Peter, Regine Hock, and Thomas Schneider. 2003. “The concept of glacier storage: A review,” Journal of Hydrology 282: 116–129.

Jorgenson, Andrew K. and Brett Clark. 2010. “Assessing the temporal stability of the population/environment relationship in comparative perspective: A cross national panel study of carbon dioxide emissions, 1960–2005,” Population and Environment 32: 27–41.

Kaser, Georg, Martin Grosshauser, and Ben Marzeion. 2010. “Contribution potential of glaciers to water availability in different climate regimes.” PNAS 107: 20223–20227.

Kerr, Richard A. 2013. “Soot is warming the world even more than thought,” Science 339: 382.

Konikow, Leonard F. 2011. “Contribution of global groundwater depletion since 1900 to sea-level rise,” Geophysical Research Letters 38: 1–5.

Levine, Joel S., Wesley R. Cofer III, Donald R. Cahoon, Jr., and Edward L. Winstead. 1995. “Biomass burning: A driver for global change,” Environmental Science and Technology 29: 120A–125A.

Li, Jiao 2011. “Water shortages loom as northern China’s aquifers are sucked dry,” Science 328: 1462–1463.Rob e rt J . Wyman 703 

Meehl, Gerald A., Julie M. Arblaster, and William D. Collins. 2008. “Effects of black carbon aerosols on the Indian monsoon,” Journal of Climate 21: 2869–2882.

Menon, S. et al. 2010. “Black carbon aerosols and the third polar ice cap,” Atmospheric Chemistry and Physics 10: 4559–4571.

Menz, Tobias, and Heinz Welsch. 2012. “Population aging and carbon emissions in OECD countries: Accounting for life-cycle and cohort effects,” Energy Economics 34: 842–849.

Micklin, Philip 2007. “The Aral Sea disaster,” Annual Review of Earth and Planetary Sciences 35: 47–72.

Ministry of Agriculture, India. 2013. Directorate of Economics and Statistics, “Agricultural statistics at a glance 2012,” Accessed 3 November 2013.

Ministry of Water Resources, India. 2011. “Ground water resources of the country,” Accessed 3 November 2013.

Mosier, Arvin et al. 1998. “Closing the global N2O budget: Nitrous oxide emissions through the agricultural nitrogen cycle,” Nutrient Cycling in Agroecosystems 52: 225–248.

O’Neill, Brian C. et al. 2012. “Demographic change and carbon dioxide emissions,” The Lancet 380: 157–164.

Postel, Sandra. 2010. “Water adapting to a new normal” in Richard Heinberg and Daniel Lerch (eds.), The Post Carbon Reader: Managing the 21st Century’s Sustainability Crises. Healdsburg, CA: Watershed Media.

PRB. 2011. “World population data sheet,” Population Reference Bureau, Washington, DC.

Qiu, Jane 2013. “Pollutants capture the high ground in the Himalayas,” Science 339: 1030–1031.

Ramanathan, V. and G.H. Carmichael. 2008. “Global and regional climate changes due to black carbon,” Nature Geoscience 1: 221–227.

Ramanathan, V. et al. 2007. “Atmospheric brown clouds: Hemispherical and regional variations in long-range transport, absorption, and radiative forcing,” Journal of Geophysical Research 112: 1–26.

Rodell, Matthew, Isabella, Velicogna, and J.S. Famiglietti. 2009. “Satellite-based estimates of groundwater depletion in India,” Nature 460: 999–1002.

Rosa, Eugene A., and Thomas Dietz. 2012. “Human drivers of national greenhouse-gas emissions,” Nature Climate Change 2: 581–586.

Rudel, Thomas K. et al. 2009. “Agricultural intensification and changes in cultivated areas, 1970–2005,” Proceedings of the National Academy of Sciences, USA 106: 20675–20680.

Shepherd, Andrew et al. 2012. “A reconciled estimate of ice-sheet mass balance,” Science 338: 1183–1189.

Smil, Vaclav 2002. “Eating meat: Evolution, patterns, and consequences,” Population and Development Review 28: 599–639.

Tiwari, V.M., J. Wahr, and S. Swenson, S. 2009. “Dwindling groundwater resources in northern India, from satellite gravity observations,” Geophysical Research Letters 36: 1–5.

United Nations. 2013. World Population Prospects: The 2012 Revision. New York.

van der Werf, G. et al. 2009. “CO2 emissions from forest loss,” Nature Geoscience 2: 737–738.

Vermeulen, Sonja J., Bruce M. Campbell, and John S.I. Ingram. 2012. “Climate change and food systems,” Annual Review of Environment and Resources 37: 195–222.

Voss, Katalyn A. et al. 2013. “Groundwater depletion in the Middle East from GRACE with implications for transboundary water management in the Tigris-Euphrates-Western Iran region,” Water Resources Research 49: 1–11.

Wada, Yoshihida 2012. “Nonsustainable groundwater sustaining irrigation: a global assessment,” Water Resources Research 48: 1–18.

Wada, Yoshihida et al. 2010. “Global depletion of groundwater resources,” Geophysical Research Letters 37: 1–5.

———. 2012. “Past and future contribution of global groundwater depletion to sea-level rise,” Geophysical Research Letters 39: 1–6.704 Populat ion and the De pl e t ion of Fre sh Wat er 

Weaver, Andrew J. 2011. “Climate change: Toward the second commitment period of the Kyoto Protocol,” Science 332: 795–796.

World Bank. 2010. “Deep wells and prudence: Towards pragmatic action for addressing groundwater overexploitation in India.” Washington, DC.

World Food Program, United Nations. 2008. “PDPE market analysis tool: Price and income elasticities,”

World Health Organization. 2009. “10 facts about water scarcity,” Accessed 5 July 2012.

Yu, Chaoqing 2011.” China’s water crisis needs more than words,” Nature 470: 307.


This article has been reprinted with permission from Wiley Online Library.

Citation:  Wyman, R. J. (2013), The Effects of Population on the Depletion of Fresh Water. Population and Development Review, 39: 687–704. doi: 10.1111/j.1728-4457.2013.00634.x

The article is available at:

The videos and photographs were added by Horizon International. This article was published on the Horizon International Solutions Site on 4 April 2014.  

Book CoverBook CoverThis article is presented as part of the Supplementary Material that accompanies the book 
Water and Sanitation Related Diseases and the Environment: Challenges, Interventions, and Preventive Measures, a Wiley-Blackwell publication in collaboration with Horizon International, written by 59 experts. Janine M. H. Selendy, Horizon International Founder, Chairman, President and Publisher, is Editor.  Professor Robert Wyman, Yale University,  is author of the chapter, "Water and Health: The demographic Background."

There are now more than 48 additional articles as part of the Supplementary Material published on the Horizon International Solutions Site.

The book’s 4 hours of multimedia DVDs are included with an abundance of multidisciplinary resources, covering diverse topics from anthropology to economics to global health are being distributed free of charge by the Global Development And Environment Institute (GDAE) at Tufts University. 

These are being given to thousands of libraries, organizations, and institutions in 138 less-wealthy countries. They are valuable additions to resources for use in classrooms and communities, by researchers and government decision-makers. 

Map of CountriesMap of Countries

Map of countries

As of 28 March 2014, these resources have been made available in over 1,500 entities in 60 countries.

Read more: PDF Version is available at





Latest articles


Air Pollution



Endangered Species




Global Climate Change

Global Health


Natural Disaster Relief

News and Special Reports

Oceans, Coral Reefs



Public Health



Toxic Chemicals


Waste Management


Water and Sanitation

Yale Himalaya Initiative