Over the past decades, the rapidly expanding and increasingly urbanised Chinese capital has been hit by frequent air pollution events tied to health-threatening concentrations of fine particulate matter. The attention received as a result of the extensive international media coverage has epitomised Beijing as an emblem of air pollution in the 21st century. Beijing has become a contemporary analogy to earlier iconic imaginaries of pollution, such as the London smog of the late Victorian era or the photochemical smog of Los Angeles. Even so, regardless its critically poor air quality, Beijing currently ranks only 57th amongst the most air-polluted cities, with various urban areas in Iran, India, China, and elsewhere experiencing even more extreme levels of particulate pollution. The Chinese smog has habitually been presented as an inevitable and natural consequence of the unregulated economic-industrial expansion of the new global player. The contemporary air quality crisis in China (and in the East) has, however,  to a large extent been incited by the displacement of polluting industries from the West, and is directly linked to the relative clearing of European airspaces.

Written collaboratively by visual artist Hanna Husberg and environmental scientist Agata Marzecova, this essay builds upon accounts and imaginaries of urban air collected during a 3-month artist residency at the Institute of Provocation in Beijing, and further interviews and research done together. Using the transcripts of Beijingers who discuss their relation to air – how air is noticed, lived with, distributed and governed – it traces some of the connections between the embodied experience of living in toxic air, and the new ways in which techno-science brings air into the fold of city governance.

The air of Beijing made international headlines in 2008 as the city hosted the 29th Olympic Games. Alongside debates about human rights, and its fast urbanisation, Beijing’s airborne particulates stirred concern among a legion of scientists, athletes and bureaucrats. Most Beijingers, however, remember the weather the year of the Olympics as ‘great’. Polluting factories and plants were closed down. As much as 90 % of the traffic was removed. And the Beijing Weather Modification Office, the world’s largest weather control effort, was enlisted by the Chinese government to ensure no rain would fall on the games.

Beijing’s air was, however, never good. From the collected narratives the city’s air emerges as having  always been heterogeneous and troubled. Through the long recorded history of China, strong seasonal winds have brought severe long-distance dust storms from regions in the northwest, covering the capital and the coastal areas in a layer of yellow dust. With drought and increased desertification dust storms intensified. From being considered as yearly omen of spring, they were refigured as disaster events that needed state intervention. Dust storms became the target of an intense program of state-led ‘ecological construction’ of infrastructure, a green wall of trees, with a mission to block the movement of wind and keep the shifting sands in place.

The dust storms affected everyone, but they didn’t last for long. With the smog we learn from SQ, one of our interlocutors, that it was ‘only little by little when you realised that so many people get cancer, get a lung problem. Newspapers and media started to talk about it. Then it becomes a problem.

Sick air

JLA: When it gets smoggy I probably just get a bad mood.

XL: The inside of my nose swells a little bit, always when there is smog it swells. So my body does react to it, but my consciousness, my mind, tries to ignore it.

LY: One of my relatives is working in China Science Department of Air. I consulted him three years ago: How about the wumai? How can we protect ourselves? He said: Lydia, it’s useless, because the wumai is made up of very tiny things, only 1/20th of a human hair. You need a microscope to see it. So you cannot avoid any wumai inside your room, and also your mouth..”

VS: I don’t get sick when the air is bad, but if I can’t see blue sky I just go crazy.

VN: Today is so horribly I knew it was going to be horrible but I didn’t bring my mask.

JJ: Close your window! Don’t open your window, when you stay inside a building. And if you see it’s smoggy don’t bring your kids outside. It’s not healthy.

ZL: Recently there was this report. They did tests with white mice exposing them to smog and letting them run through tubes and do certain tasks. They were exposed during three days. The ones exposed to mai were slower, and doing much more mistakes than those who breathed normal air.

MG: So I did a quick calculation that if I stayed in Beijing for six months, I would lower my life expectancy with nine days. So I knew that my life would statistically be 9 days shorter.

CC: One of my friends told me. Don’t worry, it’s not killing us now, it will take five years to kill you so just take it easy.

A new word

Wumai was a new word,’ JJ claims. ‘It was winter something like four years ago,’ WEL confirms. ‘Yanwu’ the word which had previously been used, in particular to define the London smog of the late Victorian era, had primarily been understood as mist and fog. The emergence of a new, more loaded, term for smog suggests that although the public might have been aware of changes or deteriorations in air quality, few had realised these changes posed actual health hazards.

Both ‘wu’ and ‘mai’ are old Chinese words Wu refers to drops of water forming a cloud close to the land, and mai refers to the sky becoming dark in ancient times, sometimes for up to a month, a phenomenon similar to rain but during which clothes kept dry. ‘Mai’ is definitely not something nice. It’s something quite monstrously dangerous,’ LL explains.

The ‘wumai’ term appeared in parallel to an increased circulation of western, technoscientific notions and definitions of the air pollution problems, notably PM2.5 and the national Air Quality Index (AQI). A standardised measurement since the 1990s, but brought to a broader public attention only in recent years, PM2.5 refers to particulate matter smaller than 2.5 μm in diameter. Research has shown that particles of this size are predominantly of anthropogenic origin. Sources of PM2.5 include diesel cars and trucks, coal burning power plants, forest fires, and construction activities. While the actual aerosols that constitute smog are far more complicated in terms of size and chemical composition, high PM2.5 values have been found to correlate with a range of medical conditions and increased mortality in epidemiological studies. Because PM2.5 particles are so small, they are able to penetrate deep into our lungs or bloodstream, and have also been found in brain tissues. The AQI is calculated from  PM2.5 values, often referred to as the most health-threatening element of the index, and several other atmospheric pollutants.

In China, PM2.5 concentrations in the air were first monitored at the US Embassy in Beijing. In 2008, the monitoring device began automatically sharing tweets with air quality data every hour. In 2009, Twitter was blocked in China and only those with a VPN service were able to access it. In 2010, the US embassy tweeted ‘Crazy bad’. The message was quickly corrected to ‘Beyond Index’. The PM2.5 concentration crossed the 500 upper limit, 20 times exceeding the level considered safe by the World Health Organization. In 2013, Chinese PM2.5 data opened to the public. A number of air quality index applications appeared, and a four-color alert system based on the air quality index was launched. In 2016, this alert system was revised. It was decided to double the levels that trigger orange and red alerts.

Home becomes like an airlock

‘My oldest air purifier is 10 years old. At that time almost none of my acquaintances understood what I wanted it for. The thing was expensive too, 1000 USD! And I’d have it at home. No one understood what it was good for.

Sometimes the PM2.5 index stops at 500 and doesn’t nudge. But the levels might still be doubled. 500 is considered so bad, that you wouldn’t be allowed to keep chickens in it in Europe. Absolutely not, it would be animal cruelty. But my children are expected to go to school even if the health index would be at 1000 if they continued to count. As it was on Christmas day last year.

It easily becomes a bit technical, but it’s quite interesting. It’s like watching birds or anything else, you learn from where you live. I keep track of these particles that I cannot see through the measuring device.

Home becomes like an airlock. We have hooks at the entrance for the kids air masks. When we go out even our two-year-old knows that he should put on the mask before going out. When we come in he’ll say: ‘Shut the door quickly so we don’t let out the fresh air!’ It’s a bit like living on a space station, like in the science fiction movies you could see in the 1980s when I was a child. And the kids are part of it.

It’s politically sensitive. The leadership wants to sell Beijing as a political showcase for the policies they want to pursue. That’s why they avoid setting criteria for indoor air quality. If they did so, they would need to say, that by definition, the air that we force our children to sit in is harmful – acutely dangerous – much of the year. Naturally they avoid it. Because of this, no regulations. My kids get to sit with masks in the classroom. They may take them of when eating.’

HS, journalist and mother of four

The rise of techno(eco)logical sensing

Owing to recent developments in sensing, computation and networking technologies, PM2.5, like all other parameters that constitute the AQI index, can nowadays be detected using automated environmental sensors that produce continuous, near real-time data transmissions. The new automation and algorithmic techniques assist in collecting information from a multitude of sensors positioned in the city  in order to automatically process, integrate and analyse the data into detailed spatio-temporal models and predictions of changes in air pollution. These are then shared digitally through ubiquitous data streams, providing citizens with constantly updated estimates of the particulate content in the air.

This development arrived in parallel to the emergence of the ‘wumai’ concept. Arguably, the automatisation of digital data processing and dissemination of air quality data in Beijing played a role in achieving a broader public awareness about the air pollution and its toxicity. Moreover, many Beijingers emphasized, that the attunement to the invisible terrain of particles, enabled by these recently emerged real-time pollution measurements, influenced the way they went about their daily routine, and even how they perceived outdoor and indoor spaces.

Unlike smoke and dust, which are perceptible to our senses, and can be observed directly, PM2.5 particulate matter is, due to its microscopic size, imperceptible to humans. It is detectable only with the help of scientific instruments and technology. Encountering the AQI and PM2.5 values in the news, through smartphone apps, in weather predictions, or even transmitted through smart sensors installed in homes, the data and its technologies become a ‘naturalised’ extension of the human senses. Rather than representing the ‘urban ecosystem’ or a certain ‘environmental condition’, they become part of the relationships that make up the urban fabric, producing technoecologies, where sensors, data, particles, and people, are interconnected, one influencing the other.

These new technoecologies produce an augmented air, adding new information to our experience, or adding an altogether different experience. Dynamic data is delivered to a mobile user. Augmenting the user also comes to mean augmenting the whole space in which s/he lives, or through which s/he passes. It proposes a new level of storytelling.

Reducing the materiality and multiple relationalities of air into the numerical measurements of the AQI index contributes to a new “perceptual regime” ( a diagnosis proposed elsewhere by Antoinette Rouvroy), in which the focus shifts from ‘incalculable’ social, economical and environmental contexts and perceptual experiences to the ‘invisible’ but ‘calculable’ particles and the calculation of statistical risks.

The everyday economics of numbers

Dr. WT: Sometimes visually when you wake up it’s superbad, but by noon when the sunshine comes it gets better. Usually the numbers are getting worse because the cars come out. That’s why in the morning sometimes it looks bad, but it probably has more element of fog than smog. So if you really want to know if it’s fog or smog you have to check the numbers.

AR: I would say it became visible with technology, with the apps. All of a sudden you’re being introduced to this mechanism. The apps are measuring this and telling you it’s already been three years that you are exposed to very high standards. You called them white skies, but actually it’s pollution.

Dr. WT: As a doctor I’m trying to guide the people. First, if the air is not good, you’d better not go out so often and you’d better put on your mask. Second, if you really don’t feel good, and it’s influencing your daily life, we can sometimes give medicine to protect the breathing system. This medicine will make you feel better.

AR: I studied pharmacy in university, and I’ve done my best to find credible informations to understand what kind of danger we are exposed to. I do that every day. When the big smog started and peaked in 2013-14, all of a sudden the air became such a primary thing for everyone here, especially the foreigners. Then you started having apps popping up everywhere that measure air. People were measuring air all the time. We’ve been doing that too, not panicking, because we have this scientific background.

PZ: People more and more pay attention to their personal health, but not in a very correct way, for example a lot of students wear masks that do no good to their health, and many parents prefer to drive their child to school rather than let them walk, because they believe that the air quality is better inside the car than outside. It is absolutely wrong.

Dr. WT: Air should be controlled as water was. The government is doing that now.

PZ: Even if we cannot dramatically change the urban form we can do some adjustments. The air pollution is much higher for cars traveling at low speed than at high speed so if we can speed up the traffic we can somehow solve the problem. It’s not necessary to put down existing buildings or do some very big change.

Dr. WT,  respiratory physician; AR, artist and researcher; PZ , urban planner


The Beijingers we met described air as having always been heterogeneous, multiple, and at times paradoxical. In recent years Beijing’s air has, however, become increasingly numerical.

Big data and the Internet of Things ‘promise’ cities the potential to gain multitude of insights from a large amount of data collected through various sources, and allow the automatic integration of environments, data and sensor technologies using highly networked services. As AR puts it: ‘One of the strategies the government uses for dealing with a billion and six hundred million people, is that whatever you throw at them you have big data.’ In this context numerical air has rich applications. It produces endless data flows for machine-based forecasting and algorithmic calculation of risk, and provides the public opportunities to sense and respond to fluctuations of particles in the air.

This shift towards technologically mediated citizen sensing is not unique to Beijing. Rather Beijing provides a fitting example of an increasingly technologically mediated and datafied urban ecology, that characterises the broader shift to the regime of environmentality, i.e. a governing of and through the environment. Speculating about the future in an increasingly urbanised world, it has become commonplace to imagine the future city as the smart city, in which the automated administration and governance of digital data flows are mostly presented as means for increasing the efficiency and the well-being of city dwellers. Managed through an extensive and constant flow of citizen, environmental and other data, that feed into holistic algorithms, these smart cities are imagined to enable optimal performance, if not harmony.

Air is particularly fitting for the smart city paradigm. Unlike many other urban phenomena, the air  quality indexes (AQI, PM2.5) can be fully automated and measured in real time. As they provide immediate information on changes in air quality they perfectly comply with the ‘smart ecosystem’. Indeed, the ubiquitous technological sensing of air has allowed for new levels of environmental awareness and public scrutiny in Beijing. It has enabled (middle-class) citizens to minimise health risks, organise into communities and engage with the topic, admittedly predominantly for strategies of adaptation.

Even so, computing the healthy city based on ‘toxic’ data feels ill at ease with smart city imaginaries that speak to matters of comfort and ease. In the case of urban air, the produced data derives from suspended particulate pollution. It measures a ‘lack of quality’. As an instance of slow violence polluted air is an environmental condition that cannot be opted out: breathing is necessary whether the air is toxic or not.

Data-driven environmental sensing technologies enable a certain reduction of health risks and some strategies of adaptation. But, there is a less clear road to other forms of public engagement. If collective engagement with atmosphere is configured as the means for self-governance of urban inhabitants we need to pose questions about the possibility, if not urgency, of other-than-preemptive algorithmic technologies of air. Further, we need to ask, what are the side effects of computationally augmenting the relationship between citizens and their surrounding in a system where air is treated as a number, and the algorithms of health metabolism are constructed from pollution data?


Stay tuned with us – Subscribe to our newsletter and follow us on social media: