CHINA is holding back the desert, for now. The Great Green Wall – a massive belt of trees being planted across China's arid north in what might be the largest ecological engineering project on the planet – seems to work, according to a new study.

"Vegetation has improved and dust storms have decreased significantly in the Great Green Wall region, compared with other areas," says Minghong Tan of the Institute of Geographic Sciences and Natural Resource Research in Beijing. But whether planting trees is a long-term solution remains disputed.

The Gobi and Taklamakan deserts of northern China are Asia's biggest dust bowls. Storms generated there regularly shroud Beijing in dust, which can also fall as far away as Greenland. In an effort to tame the deserts, in 1978 China began planting the wall, which is officially called the Three-North Shelterbelt Forest Programme. It is due for completion in 2050 and will eventually contain more than 100 billion trees in a 4500-kilometre belt, covering more than a tenth of the country. But opinion is divided about its success and advisability, and it has met with widespread scepticism among Western geographers.

Some credit a rise in rainfall for the decrease in dust storms across northern China over the past three decades. But Tan and co-author Xiubin Li say that their analysis of rainfall data, satellite images and an index of dust storms shows conclusively that the Green Wall is the main cause of the improvement (Land Use Policy, doi.org/xk2).

Away from the wall, vegetation cover and dust storms have risen and fallen with precipitation. But nearer to the trees, vegetation increased and dust storms diminished between 1981 and 1998, the end of the study period. Tan says the improvement has since continued. "In most places in the study area, greenness continued to increase between 2000 and 2010," he says. "In North China as a whole, we think the environment is getting well."

There are exceptions, such as Minqin in Gansu province, where the Gobi and Taklamakan come close to meeting. "But these are small regions," Tan says.

Two prominent critics of the Green Wall did not challenge the findings when spoken to by New Scientist, but warned that they do not tell the whole story. Hong Jiang of the University of Hawaii at Manoa, says China's "aggressive attitude towards nature", especially planting trees where they do not grow naturally, will not ultimately work. "Instead of controlling nature, we need to follow nature," she says. Sometimes that means "allowing sand the freedom to roll".

David Shankman of the University of Alabama in Tuscaloosa says it is not clear how permanent the Green Wall would be. "What is the mortality rate of planted trees? What happens when they die? And how do these trees affect grass and shrubs, which in general are more resistant to drought and more effective at erosion control?"

Tan agrees that the authorities should not just focus on increasing forests. "Grass may be better in most places in north China," he says.

This viewpoint is echoed elsewhere. The African Union aims to tame the Sahara with a Great Green Wall of trees, but scientists from McGill University in Montreal, Canada, reported in October that grasses and shrubs would grow faster and be more useful (Sustainability, doi.org/xk5).

More trees are still to come in China: last month the country announced that it will plant 1.3 billion trees along the Silk Road in partnership with the UN. (New Scientist)

This article appeared in print under the headline "Wall of trees keeps deserts at bay in China".