Crisis on tap? Pollution and burgeoning populations stress earth's water resources.

Earth gets one of its nicknames, the Blue Planet, from the way it looks from space. About 70 percent of the planet's surface is covered with water, a substance that known types of life can't do without. All told, the oceans, land, and atmosphere hold the equivalent of almost 1.4 billion cubic kilometers of liquid water. About 96.5 percent of that total is salty ocean, a little more than 2 percent of the total is locked up in ice, and a smidgen wafts as vapor in the atmosphere. That leaves just over 1 percent as water that's readily available for human use.

And that small fraction is getting smaller day-by-day. In many areas, the amount of fresh water that falls as rain or snow and eventually reaches lakes and rivers isn't sufficient to meet the current or projected demand for drinking water, irrigation, and industrial activity.

To supplement this precipitation, which often doesn't occur when or where it's needed, people are pumping vast amounts of water from aquifers, layers of sediment or soil that hold moisture in the spaces between their particles. In fact, people have pumped enough water from aquifers during the past century to measurably raise global sea levels. The removal of water from some types of sediments can cause them to compact, forever destroying some of their capacity to hold future rainfall.

Because aquifers aren't being recharged nearly as quickly as they're being depleted--and because people are becoming ever more dependent on aquifers to fulfill their various thirsts--scientists are striving to better understand how groundwater systems interact with the water that flows across Earth's surface. Indeed, the pressures of pollution and a growing world population threaten to transform the lonely lamentation of Samuel Coleridge's ancient mariner--"Water, water, everywhere, nor any drop to drink"--into a cry legitimately uttered by millions of future landlubbers.

ALL PUMPED UP Many cities and towns originally cropped up near rivers that could provide fresh water. Sooner or later, however, a large number of those grew into metropolitan areas where the demand for water began to outstrip the supply.

A region is considered to have a relative scarcity of water if more than 20 percent of the local river's flow is diverted for household use, agricultural irrigation, and industrial purposes. In 1995, more than one-third of the world's population of 5.7 billion lived in such areas, says Richard B. Lammers of the University of New Hampshire in Durham. Of those people, about 450 million lived in areas of severe water stress, where more than 40 percent of a river's flow was diverted for human use.

That growing overuse of surface water has been making aquifers a critical source of fresh water. Scientists estimate that these underground reservoirs, which can reside at depths just below the surface to more than 1 kilometer down, hold in excess of 1,000 times the amount of water that falls on land as precipitation each year, says William M. Alley of the U.S. Geological Survey (USGS) in Reston, Va. More than half of the U.S. population and more than one-quarter of people worldwide depend on groundwater as their primary source of drinking water, he notes.

Much of the water in these reservoirs has come from rain and snow. Shallow aquifers may contain water that fell locally in the previous few days, but moisture held in deep layers of sediment may have originally rained down in far-off regions hundreds of thousands of years ago. For example, isotopic analyses of dissolved elements in water deep underground in parts of the western United States suggest that it fell as rain about 15,000 years ago. That was at the height of the last ice age, when rainfall in the region was vastly more plentiful and aquifers absorbed water at about 20 times the rate they do today, says Alley. in the June 14 Science, he and several colleagues describe recent research recognizing and quantifying the interactions between water that flows on Earth's surface and that stored in aquifers.

Because deep aquifers are slow to recharge, the reservoirs are essentially a nonrenewable resource that's being mined. In the past 50 years, says Alley, groundwater depletion has become a problem in many areas of the world. In substantial portions of the High Plains aquifer, which underlies a 450,000-square-kilometer area that stretches from South Dakota to the Texas Panhandle, more than half of the subterranean moisture has been pumped out. Water levels in that aquifer have dropped more than 45 meters in places. On the arid plains of northern China, the depletion of shallow reservoirs has forced people to sink wells into aquifers more than 1 km below the surface.

Problems can arise even if the rate of groundwater withdrawals doesn't exceed the rate at which precipitation recharges the aquifer, says Alley. Aquifers, especially shallow ones, aren't completely isolated from surface water such as lakes, streams, and rivers. In some regions, groundwater provides a major source of water for rivers as it flows from the ground into the depths of the streambed. A 30-year study by the USGS of 54 streams across the continental United States showed that, on average, more than half of the stream's annual flow came from groundwater. If significant amounts of shallow groundwater are diverted for agricultural or other uses, the flow patterns and the ecology of the rivers can suffer.

In arid and semiarid regions, the pattern typically is reversed. Bodies of water, many of them short-lived, contribute water to the subsurface. In some parts of Niger, a dry landlocked nation in sub-Saharan Africa, virtually all the water that ends up in the region's patchy aquifers seeps down from ponds that form after occasional rains.

In the U.S. Southwest, the rate of natural recharge through ephemeral streams is one of the least understood factors in the groundwater equation, says John P. Hoffmann, a hydrologist at the U.S. Geological Survey in Tucson. To chip away at this ignorance, he and his colleagues are using untended temperature sensors buried in streambeds to estimate the rates at which water percolates down into the subsurface.

The researchers studied a stream near Tucson that flowed only 27 days in the year of the study. Because water carries heat from the streambed surface, the scientists could usually detect the onset of water flow within 20 minutes by looking at temperature data collected from sensors buried 1 m deep. They could also detect within 2 to 3 hours when the stream went dry.

Using measurements from sensors even deeper in the streambed, the scientists could also estimate the amount of water that leaves the surface to recharge shallow aquifers beneath the transient waterway. Such a network of buried electronic sensors could be less expensive and more effective for estimating the rate of aquifer recharge than are surface flow meters, says Hoffmann. His research team reported its findings at a meeting of the American Geophysical Union in San Francisco last December.

THAT SINKING FEELING Aquifers can receive more than just the water that trickles down from the surface. In many places, particularly those with distinct wet and dry seasons, people drive massive volumes of excess water into underground reservoirs when it's available--and before it flows out to sea--and then pump it out later when demand is high (SN: 1/29/00, p. 73).

The region that includes Los Angeles is a good example. Water districts in this 14-million-person metropolitan area continuously stock their aquifers with water from local precipitation and water diverted from the Colorado River and sources in northern California. When water demand is low, the ground swells like a wet sponge. But when summertime rolls around, the net withdrawal of groundwater causes the ground to compact and subside. In some spots in the Santa Ana basin, southeast of Los Angeles, the ground rises and falls as much as 11 centimeters over the course of a year (SN: 8/25/01, p. 119).

Some of the subsidence is permanent, says Gerald W. Bawden, a geophysicist at the USGS in Menlo Park, Calif. Layers of fine-grained sediment underlying the region can be as much as 70 percent water by volume, he notes. When the water is drawn out, the grains of the rock can shift and settle closer together, which leaves less space for water to occupy when it's time to store the bounty of the next rainy season. When Bawden and his colleagues first analyzed satellite measurements of the Los Angeles area taken between 1992 and 1998, the data suggested there had been around 6 centimeters of unrecoverable compaction of sediments in the region, or about 12 millimeters each year on average.

A recent reanalysis of the data shows the compaction actually was minimal before 1995, which boosts the post-1995 net loss of thickness to about 20 mm per year. And the rate of subsidence may be accelerating, Bawden notes. Satellite observations made between 1998 and 2000 suggest that during that period, the basin's net compaction was a little more than 22 mm per year. He has requested measurements made since 2000 to see whether the trend continues and the compaction is accelerating.

In areas where water isn't pumped back into aquifers, subsidence can be even more substantial, says Alley. In the desert corridor between Phoenix and Tucson, withdrawals of groundwater for agriculture have caused water levels in aquifers in some small areas to drop more than 90 m and the subsurface sediments to compact more than 3 m. In California's San Joaquin valley, an area of more than 13,000 square km has subsided at least 30 cm due to groundwater depletion, and in the worst spot, the surface of the ground has dropped more than 9 m.

Agriculture isn't the only culprit: Consider Venice, Italy. Groundwater extractions there for industrial uses from the 1930s to the 1970s significantly contributed to subsidence of sediments, complicating the city's woes of rising sea levels and sinking buildings (SN: 7/24/99, p. 63).

GROWING DEMAND According to the World Meteorological Organization, world water demand in 1995 was six times that of 1900, even though the global population only tripled during that same period. Much of the surge in demand stemmed from increased industrial activity and irrigation-intensive agriculture, which now supplies about 40 percent of the world's food crops, says Lammers. If current patterns of water consumption and population growth continue, in the year 2025 at least 3.5 billion people will live in river basins where water scarcity significantly affects household and economic activity. That's about half the projected world population then, says Nels Johnson, formerly at the World Resources Institute (WRI) in Washington, D.C. Demographers also predict that much of the forthcoming world-population growth will occur in urban areas, which are poised to hold 5 billion people a quarter century from now.

The growing population is already making surface water scarcer. On top of that, add the stress from increased industrial and agricultural activities that inject more contaminants into that surface water, says Johnson, who's now with the Nature Conservancy in Harrisburg, Pa.

Contamination affects the water supplies of as many as 3.3 billion people today. In the developing world, up to 90 percent of wastewater, including sewage, is discharged directly into rivers and streams without treatment. In developed nations, industrial pollutants pose a potent threat. Also, pesticides and fertilizers taint surface waters in agricultural areas worldwide, Johnson notes.

Eventually, that pollution will push increasing numbers of people to seek groundwater to slake their personal, industrial, and agricultural thirsts. Therefore, says Alley, it's vital that scientists understand the interactions between the water that streams across Earth's surface and that seeps through the sediments below. In time, he notes, much of the world will yet again be following a Southern California trend: They'll be getting water from aquifers that they've artificially recharged from above.

That assumes, of course, that the water pumped into aquifer storage is itself unpolluted.