About 12 kilometres north of the Huitengliang wind farm at Xilinhot, Inner Mongolia, herdsman Siqinbateer has observed a weird phenomenon in his pasture that baffles even meteorologists. 'The ground heats up quickly, like a pan on a stove, the wind blows like a headless fly and not a single drop of rain falls,' he said in August, during the rainy season. He pointed at the spinning blades of the wind turbines over the horizon. 'This started happening after they came.' It is not just a herdsman's superstition or his distaste for modern technology. Siqinbateer's claim is backed up by government statistics. Li Qinghai, an engineer with the Water Statistics Bureau in Xilingol League, said the precipitation data collected by the bureau showed that adjacent to big wind farms there was an obvious decline in annual rainfall since 2005 - in some areas by as much as 80 per cent. 'The issue is often overlooked as much of Inner Mongolia is suffering an unprecedented drought,' he said. 'But after spending more than two decades studying the rise and fall of water levels in the region, I have a strong feeling that the wind turbines are playing a disruptive, if not destructive, role in this, because the droughts in these areas developed much faster than in the turbine-free regions.' Li said he wanted to study the issue more deeply, but nobody would fund the research. Given the nationwide hype of wind-power development, the topic is considered politically incorrect. Scientifically, warming the ground should enhance rain formation rather than suppress it. One possible explanation is virga, which are streaks of rain that are so thin that they evaporate before reaching the ground due to the soaring heat, but many scientists reject that possibility. The confusion exemplifies how little we know about the long-term environmental impact of the wind turbines. Large wind farms have been growing exponentially on the mainland in recent years, and Greenpeace estimates that by 2020 they would seize from the atmosphere as much energy as 13 Three Gorges Dams could generate. While the dams have been criticised for draining life-sustaining energy from natural rivers, there is little understanding, let alone concern, about 'damming' the wind. Scientists, who have conducted some research on the subject, came to one conclusion: the large-scale use of wind power can alter local and global climate by extracting kinetic energy and change the wind's swirling forces. In other words, wind power is not completely green. As China will probably be the world's biggest producer of wind energy by the end of this year, researchers in China and worldwide have called for the wind industry sector and the government to look at the issue seriously. They caution that until wind farms' effect on regional and global climate systems is better understood, building even more could lead to unexpected disasters. Dr Gao Hu, deputy director of the National Development and Reform Commission's Centre for Renewable Energy Development, said the government had never heard of the issue and would not fund research. 'Everybody wants to see the rapid development of wind energy,' he said. 'We don't want anything to get in its way. We have never heard of such complaints, and we are not worried about it. Study is unnecessary because it tries to look at something that can't possibly happen.' Theoretically, though, the possibility is there. In 1961, American mathematician and meteorologist Edward Lorenz discovered by computer modelling that in a chaotic and sophisticated system such as atmosphere, a disturbance as small as the flap of a butterfly's wings could set off a tornado thousands of kilometres away. It has come to be known as the Butterfly Effect. A typical wind turbine on the mainland is more than 100 metres tall with three enormous blades, each requiring a heavy-duty truck to transport and the most muscular crane to assemble. A large wind farm has hundreds of them. For decades, wind industry designers have been mining the catalogue of airfoil profiles developed by the US National Advisory Committee for Aeronautics, the predecessor of Nasa, to determine the shape of the most efficient blades. The chosen blades were tested in a wind tunnel to ensure the maximum lift and minimum drag in a given environment, such as Inner Mongolia. The result is a highly efficient energy sucker. To spin a triple-blade turbine weighing as much as 40 tonnes, all it takes is a light breeze moving as low as 3 metres per second, which according to the Beaufort wind force scale, causes leaves to rustle and can be felt on exposed skin. So, it was no surprise that in a 2004 study on the influence of large-scale wind power on global climate, a team led by Professor David Keith, of the University of Calgary in Canada, discovered that a 'very large amount of wind power can produce a non-negligible climate change at continental scales'. It cited the wintertime cooling over most of Europe happening at the same time as warming over the central United States as one example. The disturbance could widen the gap between temperature extremes by as much as 4 degrees Celsius, enough to raise havoc in a region, the study found. The results were obtained by a computer simulation of two different general circulation models to achieve a more reliable result. But neither model included the significant atmospheric effect of turbine-generated turbulence, which, the authors admitted, may underestimate the impact on climate. Due to the scientific complexity of the issue and the novelty of large wind farms, only a few prominent meteorologists in Canada, the United States and Germany have begun to study the problem. These studies are not detailed enough for quantitative evaluation of wind turbines' climate impact, globally or regionally, but there is no doubt in the meteorological community that it should be done. In an e-mail interview, Keith said that the Inner Mongolian locals' observations were interesting and worthy of state-funded research. 'Good mesoscale modelling, good data on climate and measurements of turbulence' would generate valuable scientific findings that could, at least in Canada, help support local residents in getting financial compensation from wind power companies or the government, he said. With so many different variables factored into the make-up of climate, he said, it would be hard to pinpoint any one cause of a given effect. Hu Yongyun, professor of the School of Physics' Department of Atmospheric and Oceanic Sciences at Peking University, said wind turbines were unlikely to dramatically alter regional climate, such as precipitation, as long as they stayed small. 'In my opinion, wind farms of less than a dozen square kilometres should have little impact on climate, because a thunderstorm, for instance, requires a much larger area to form,' Hu said. But large-scale wind farms were another matter, he said. China is building seven such major bases. One in Jiuquan, Gansu, covers nearly 200,000 square kilometres. 'Like hydroelectricity, wind power originates from the sun, and the speed of its renewal is slow in certain regions,' Hu said. 'Wind plays an important role in transporting heat and moisture. I think the wind energy companies and the government had a responsibility to give people an answer about the wind project's possible impact on climate before putting up their turbines.' Professor Wang Hongqing, a computer modelling expert in the same department as Hu, said the Butterfly Effect had been accepted by most scientists but it remained little more than a theory. 'The Butterfly Effect certainly exists, but rarely do scientists consider it in practice. Even something as large as a wind turbine is often neglected because we consider only the air movement high up in the atmosphere, more than a dozen kilometres above sea level,' he said. 'But large-scale wind farms will almost definitely have an impact on regional climate. It could be something bad, as the locals in Inner Mongolia experienced, or something positive. The problem is that we don't know, and I think nobody will know unless somebody does some research about it. Considering the fast development of wind power in China, we have to do some research about it.' But Professor Gerd Tetzlaff, director of the Institute for Meteorology at the University of Leipzig in Germany, cautioned that such research would be difficult, if not impossible, at the moment. The herdsmen's observation that the ground was getting warmer made Tetzlaff question how long the period of observation was. 'To prove such a statement would require something like 20 or 50 years of observations,' he said. 'I assume that large-scale wind farms are a development of the past five years at most, with a strong upward trend ruling out most statistical techniques to be applied. 'In principle, scientific results can always help to prove a certain position or claim. Whether or not this leads to success depends on many influences, the lack of practice-relevance and of quality of the scientific findings often being the first obstacle.'