Beijing’s hazardous blue sky

In the first analysis of US embassy “Twitter” data on pollution, Steven Q Andrews finds major failures in Chinese air quality assessments. Tighter standards are on their way, but will continue to fudge the health risks.

In the last two years, Beijing officials have announced good or even excellent air quality nearly 80% of the time, notwithstanding the persistent smog. But a monitor atop the United States embassy illuminates a different perspective: over 80% of days had unhealthy levels of pollution, and the air quality has been hazardous more often than good. Dozens of countries around the world have public air pollution reporting, but no other country describes even very low levels of pollution as excellent.

Some of the discrepancy between reported pollution levels is due to omission. The Beijing government monitors but does not report fine particulate and ozone, both of which are pollutants linked to lung disease and premature death. And some of the discrepancy is due to commission: official reports grossly understate the severity of pollution.  [See Figure 1: Reporting of Equivalent Pollution Levels in Beijing, Hong Kong, Europe, and the US: Air Pollution Index Values, Pollution Levels, and Reported Color]

Due to public pressure arising from the US embassy assessments, the government recently announced plans to include fine particulate and ozone in public reporting, but not until 2016. In the last two years, over half the days (around 55%) in Beijing exceeded the new daily fine particulate standard (75 micrograms per cubic metre) and the annual average concentration has been approximately three times higher (around 100 micrograms per cubic metre) than the proposed standard (35 micrograms per cubic metre).

But many of the misleading descriptions will remain under the new regime and, in some cases, worsen.

The main publicly reported pollutant of concern in Beijing is particulate, tiny particles of solid matter suspended in the air whose impact on human health varies according to their size. Under current guidelines, so-called excellent air quality can have coarse particulate (PM 10 – meaning particulates of 10 microns or less in diameter) levels two-and-a-half times higher than the World Health Organization (WHO) guidelines. Once China begins reporting fine particulate (PM 2.5), levels three-and-a-half times higher than WHO guidelines will still be called “excellent”. PM 2.5 concentrations considered moderate in the United States will also be “excellent”.

The US embassy monitor has a Twitter feed, @beijingair, which is widely viewed on mobile apps and reposted on Chinese microblogging sites like Sina Weibo. Many media reports have mentioned these hourly reports, which often assess the air to be hazardous, but this article provides the first discussion of daily average pollution concentrations measured by the embassy. These measurements are taken over the same time period as those recorded by the Beijing Environmental Protection Bureau (BJEPB).

Methodologies approved by the US Environmental Protection Agency for assessing PM 2.5 concentrations are based on a 24-hour average, so the embassy’s hourly tweets may actually overstate the severity of pollution in Beijing. For example, in the first 10 months of this year, 57 days had at least one tweet reporting “hazardous” pollution levels, although only 17 days had a 24-hour average PM 2.5 concentration that was hazardous. These days had air quality that was “very unhealthy” rather than hazardous.

According to Du Shaozhong, vice-president at the BJEPB, it is an “indisputable fact” that air pollution in Beijing has improved in recent years. In China, days that meet air quality standards are termed “blue sky days” and described as having “good” or “excellent” air quality. Officially, the number of blue sky days increased to 286 days (78%) in 2010, up from 100 days (27%) in 1998.

But, these so-called improvements are due to irregularities in the monitoring and reporting of air quality – and not to less polluted air. Most importantly, the government changed monitoring station locations twice. In 2006, it shut down the two most polluted stations and then, in 2008, began monitoring outside the city, beyond the sixth ring road, which is 15 to 20 kilometres from Beijing’s centre.

These changes continue to reap dividends for the authorities, not only in the increased numbers of blue sky days, but also in lower reported pollution concentrations. The reported number of blue sky days in 2010 would have been approximately 71 days less if the monitoring station locations had not changed. (Fifty-five days due to the 2006 changes, and 16 days due to the 2008 changes.)

The Chinese Academy of Sciences, using its own measurements, has found that PM 2.5 concentrations actually increased by 3% to 4% every year over the past decade. PM 2.5 is believed to pose the largest health risks, because particles of this size can be absorbed into the lungs and blood.

Two weeks ago in Beijing, while very unhealthy levels of pollution were again being officially reported as good air quality, Bie Tao, an official at the Ministry of Environmental Protection (MEP) announced that public opinion would soon be used to “assess the performance of local governments on environmental protection”. Understanding the true severity of pollution and the extent to which the public and media have been misled is now increasingly important.

Data and methods

The US embassy began publicly reporting pollution concentrations and qualitative descriptions of air quality in 2008 as an unofficial resource for its staff.  For this simple analysis, I downloaded archived Twitter data from January 2010 through to the end of October 2011 and compared it with data reported by BJEPB. Officially reported Air Pollution Index concentrations are converted to PM 10 concentrations and compared to the qualitative reporting systems used in Hong Kong and the European Union. Annual PM 2.5 concentrations and the ratio for PM 2.5 to PM 10 are also calculated based on daily US embassy data.

Additionally, the US embassy data is compared to the new Chinese daily and annual fine particulate standards. While there are some gaps in the data, daily concentrations were obtained on 81% of days during this 22-month period, and analysis was only conducted on the days for which there was data both from BJEPB and the US embassy. (The US embassy refused my request for the rest of the data, responding that some gaps in the data are due to maintenance of the monitor.)

Two main concerns have been raised regarding the embassy figures: that they are not representative of the city overall – an issue mentioned on the embassy’s own website – and that the data is not accurate, a position put forward by Beijing’s environmental authorities. But while the air quality at the US embassy in Chaoyang district may not be representative of all of Beijing, air quality in the area has historically been better, not worse, than the rest of the city. It is also typical in most countries to use the worst monitoring station in the entire network to calculate attainment of standards, not a selective subset as is done in Beijing. Beijing actually reported annual average pollutant concentrations for individual monitoring stations, but only in 1998, and not since.

Moreover, the monitor model used at the embassy – the BAM 2010 PM 2.5 monitor – is the only continuous particulate monitor that the US Environmental Protection Agency allows for PM 2.5 monitoring in the United States. If the data is good enough for assessing compliance with air quality standards in America, then it should be useful for at least an initial assessment of pollution levels in Beijing.

The BJEPB has argued that air quality should not be judged from data released from foreign embassies and that the measurements are hype, but the fact that PM2.5 concentrations monitored by the BJEPB are not publicly reported makes such analysis necessary. Examining daily rather than hourly concentrations helps avoid the “hype”.


This research indicates that the misleading descriptions of the severity of particulate concentrations play an important role in understating the severity of pollution. The descriptions given for pollutant concentrations have a significant impact on how the severity of pollution is perceived by the public.

Even after the proposed revisions the severity of pollution levels will continue to be understated. According to the US embassy daily reports, there was good air quality on only 3.5% of days during the year, while under the proposed reporting system there was reported good air quality on approximately 45% of days.Figure 6. 

In the last two years, Beijing reported that nearly 80% of days had good or even excellent air quality.These same levels of PM 10 pollution would have be classified as high or worse on over 80% of the days in Hong Kong and the European Union.  These pollution levels are quite similar to the reported air quality based on independent PM 2.5 measurements from the US embassy, where over 80% of days were found to be above American standards for safe levels of air pollution. The US EPA has also recently proposed plans to significantly tighten its PM 2.5 standard.

During the past two years, PM 2.5 concentrations measured at the US Embassy were on average approximately 85% of the PM 10 concentrations reported by the BJEPB.  Annual average fine particulate concentrations monitored by the US embassy have been at approximately 100 micrograms per cubic metre for the last two years.

What next?

Most people don’t really care about governmental constructs like blue sky days and pollution indices. People worry about what really matters – the impacts of pollution on their health. As the China Daily recently wrote: “All of the residents in the city are aware of the poor air quality, so it does not make sense to conceal it for fear of criticism.” Even with the proposed revisions, the severity of air pollution in China will continue to be understated.

The average annual PM 2.5 concentration of approximately 100 micrograms per cubic metre for the last two years is nearly three times higher than the proposed annual standard of 35 micrograms per cubic metre. These levels are similar to published measurements for Beijing in 2000 and 10 times higher than the World Health Organization guidelines of 10 micrograms per cubic metre. In comparison, annual average fine particulate concentrations in America’s most polluted city, Los Angeles, were at 15 micrograms per cubic metre in 2010.  In a recent study of over 500 cities around the world, the WHO found that urban areas in Mongolia, Madagascar, Kuwait and Mexico had the highest PM 2.5 concentrations, but the pollution levels measured were only about half as severe as Beijing.

One of the most authoritative studies on the health effects of pollution, by C Arden Pope and others, published in 2009, found a decrease of 10 micrograms per cubic metre in a city’s fine-particulate concentration was associated with an estimated increase in life expectancy of approximately 0.6 years. This indicates that if Beijing’s fine particulate concentration even reached the polluted levels of Los Angeles, life expectancy may increase by over five years. Back in 1999, Chinese premier Zhu Rongji stated his own fears that air pollution in Beijing would shorten his life “at least five years” – and fine particulate concentrations have not improved since then.

There is no reason for the Beijing government to continue to wait before publicly reporting and accurately describing the hazardous air. As a first step, the government should stop describing dangerous levels of air pollution as excellent air quality. Because fine particulate and ozone levels are already measured, they should be reported to the public. With every additional polluted blue sky day the government reports, it continues its misinformation campaign that has misled the public and helped prevent real improvements in the city’s air.

Steven Q Andrews is an environmental consultant based in Beijing. He recently completed a JD at the UCLA School of Law, and previously studied geosciences at Princeton University. His research on Chinese air pollution began while he was a Princeton-in-Asia fellow in Beijing during 2006-2007.

Image and graphs by Steven Q Andrews
Homepage slider image courtesy of onekell