During the Beijing Olympic Games in August, fears of choking pollution were quickly forgotten thanks to a bevy of anti-smog measures and good (partially man-made) weather. According to the government, all but one of the days in August were "Blue Sky" days—Beijing's term for a day with only "moderate pollution," when the Air Pollution Index hits 100 or below. (The API is based on an average of sulfur dioxide, nitrogen dioxide and fine dust readings at 27 monitoring stations—ozone data is not released—and has nothing to do with the actual color of the sky that day.)

As a result, attention turned toward more pressing controversies, like lip-synching and underage gymnasts. But now, a paper published in the latest issue of Environmental Research Letters (pdf) by researcher Steve Andrews concludes that Beijing's claims to air quality improvement over the past decade may be more a matter of cooking the books than cleaning the skies.

Even for those of us with an unlikely and perhaps unhealthy optimism about China's environmental situation, it wasn't hard to be suspicious of Beijing's pollution progress in recent years. In 2007, the Beijing government managed to meet its target of 246 "Blue Sky" days largely thanks to a disproportionate number of days when the API reading was exactly 100. More than a few people wondered if pollution readings weren't being shaved to meet the official goal. One of those people was Andrews, a young researcher working for a green NGO in China. After digging into official data available on public web sites (like this one and this one), Andrews found that Beijing had been moving its air sampling stations to areas with less traffic and industry to create the appearance of less pollution. His research, first published last year in the Wall Street Journal, showed that if the same monitoring station locations used in Beijing from 1998 to 2005 continued to be used in 2006, 38 of that year's 'Blue Sky' days would have exceeded the "Blue Sky" standard. It's as if New York City were measuring its pollution by putting monitoring stations in Stony Brook.

When I asked Du Shaozhang, a senior Beijing environmental official, why the city had moved its monitoring stations, he would switch between condemning Andrews's findings (saying, "This phenomenon does not exist, because this is a misunderstanding") and pledging "improvements" to the monitoring network, conceding that Beijing needed "to enhance observation and enforcement and supervision" of the air-quality data.

Andrews's study, though focused on Beijing, raises questions about the monitoring of pollution in other large Chinese cities. (In China's 14 largest cities alone, air pollution is responsible for the deaths of 50,000 newborns each year, according to the Shanghai Star newspaper.) It also raises questions about the air during the Olympics. During the first few smoggy days of the Games, government-issued numbers on pollution would not only hover suspiciously below the "blue sky" cutoff, but differ significantly from independent surveys made at populated areas. On August 8, the day of the opening ceremonies, for instance, the official API was a "Blue Sky" 94, which translates to a PM10 (fine dust) concentration of 138 micrograms/m3 (incidentally, that is almost three times the World Health Organization's recommended exposure). But, that same day, the BBC measured a much higher concentration of 156 mg/m3, and the Associated Press measured a PM10 concentration of 345 mg/m3 at mid-afternoon on the Olympic green.

Even in the months after the Games, the Beijing Environmental Protection Bureau has declined to release details about exactly what was in the air in August, hampering efforts by researchers to verify the number of "Blue Sky" days. The government did take serious steps to combat pollution before the Games, including closing factories, phasing out coal heating, and upgrading national emissions standards (cars are said to be the biggest smog contributor in Beijing). And the city continues to attempt improvements of its air quality, with a new trial project that takes one-fifth of the city's car population off the roads each day. But without dependable information on what's in the air, it will be hard to tell if anti-smog initiatives are really working—and harder still to breathe.

--Alex Pasternack