The Badain Jaran Desert in north China's Inner Mongolia Autonomous Region has long been known for its lakes and sand dunes.
More than 100 lakes are scattered throughout sand dunes as tall as half a kilometer, the tallest in the world.
Nobody had ever sensed a link between the two wonders of nature before Chen Jiansheng and his colleagues at the Hehai University in Nanjing, east China's Jiangsu Province, revealed their discovery underneath the sand surface.
According to Chen, there is a huge reservoir underground that provides the water for the lakes.
And the sand dunes survive in the windy desert also because water from the reservoir cements them together, he writes in an article published in the scientific journal Nature late last month.
They suspect the water in the underground reservoir is replenished by snow melting on distant mountains.
"Our findings could transform plans for the region's water resources," says Chen.
Up to 500 billion liters of water could be extracted from the desert every year, the scientists suggest.
In the last three years, Chen and his colleagues have been to Badain Jaran six times, trying to trace their origins of the water.
They began the search by analyzing the chemical elements in the lake waters, which they believe may hold clues to their origin and route.
They found calc-sinter, a kind of sediment of calcium carbonate that is prevalent in pre-historic ocean beds. This has led them to suspect the water that supplies the lake has run through rock deep underground before making its way to the ground.
They also found a fossilized plant root in a sand dune near the lake. The plant root had been there for over 4,000 years.
"It indicates that the sand dune has remained stationary there over this period of time," Chen noted in his article.
Why has it not moved in the direction the wind blows, as sand dunes in the desert normally do?
"Probably because water has cemented it," Chen suggested.
He and his colleagues dug into the dune and were surprised to discover moisture just 20 centimeters below the surface, despite the fact that they were almost 17 meters above the level of the lakes.
"The water found is clearly essential for maintaining the stationary dunes," says Ling Li, a member of the team. In other windy deserts, dunes tend to migrate the way the wind blows, he added.
The annual rainfall in the desert is less than 40 millimeters, Chen noted, while the evaporation rate is over 4,000 millimeters. Without major rivers nearby, the water must come from somewhere else, they reasoned.
The team measured the concentration of different elements in the dunes and lake-water and found the ratio of different oxygen isotopes, for example, matched the profile of snow on the Qilian Mountains, 500 kilometers to the southwest.
They also found strontium dissolved in the water, showing the melted snow seeps through deep subterranean faults to reach the desert. From looking at other elements they deduced that the water takes 20 to 30 years to make the journey from the foot of the Qilian Mountains to the lake or the sand dunes.
The annual water supply from the underground reservoir, they estimated, amounts to 500 million cubic meters.
But the water supply failed to transform the lakes there into larger ones probably due to the tremendously high evaporation rate, the scientists say.
The Chinese Government has initiated a huge project to divert water from a major river in the Qilian Mountains to support the local farmers around the Badain Jaran area.
Chen doubts the feasibility of the billion-yuan project, saying the expected great loss to evaporation over the long distance will likely make the project fall short of its goal.
Based on their discovery, Chen has suggested sending water from the Qilian Mountains to the thirsty Badain Jaran through underground tunnels and thus avoiding losses from evaporation.
If sending water through underground tunnels is not viable in economic terms, alternative plans should be drawn to make use of the water right under the desert surface to meet the local needs, he added.
Chen's point, though exciting, has faced doubts and criticism from some peer researchers.
"The idea that water stabilizes the sand dunes is interesting," says Mike Edmunds, research director of the Oxford Center for Water Research in Oxford, UK.
But he takes issue with the suggestion that the underground water could be extracted to meet the local demand.
He has collected data in the same desert and says there is no evidence that the water levels are being topped up by melting snow.
He considers Chen's theory flawed, arguing that the water in the dunes comes mainly from rainfall while the lakes are supplied by reservoirs that formed long ago, perhaps during the last glacial era.
He believes that the water is highly unlikely to have traveled from the mountains to the dunes in just a few decades.
Edmunds was also echoed by some Chinese researchers who have done research in the same field.
An unnamed Chinese researcher was quoted by the Nanfang Weekly as saying that stronger evidence is needed to hold up Chen's theory.
"Underground water normally moves several meters a year, how can the time for replenishing be only 20 to 30 years?" he was quoted as saying.
Chen and his colleagues reached their estimate of how much water can be extracted by comparing rainfall over the 23 square kilometers of desert to evaporation from the same area. The difference, they reason, is the amount of groundwater recharged.
But Edmunds counters that the difference arises only because the ancient reservoirs are drying out. If the fossil water is extracted, this will only happen faster.
He warns that locals would be relying on a water resource that would ultimately run dry and that the ecology of the area could be damaged irrevocably.
"It could be a disaster," he said.
(China Daily December 10, 2004)
