- Openness I - Open Data
- Eilís O'Neill
- Jerry Randall
In 2011, Christa Hasenkopf, an atmospheric scientist, and her husband Joe Flasher, a software developer, stepped off a plane in Ulaanbaatar, Mongolia. The first thing they noticed was the air.
“You can smell it in the airplane before you land. You can smell it on people’s clothes before you even board the airplane. You can see it hanging over the city as you fly in,” Flasher remembers.
They weren’t going to escape this thick haze anytime soon. Hasenkopf would spend the next two years doing post-doctoral research at the National University of Mongolia while Flasher wrote software and became involved in the local tech startup scene. During that time, Ulaanbaatar’s air permeated every aspect of their lives.
“When you’re in it, everything just seems a bit hazy,” Flasher says. “Your eyes sting and your throat is scratchy and has an odd taste to it. If you leave a cup of water out overnight, it tastes like charcoal in the morning. Forget running outside—I’d try to walk very slowly wherever I was going so I could breathe as shallowly as possible. We’d pay someone to come in and seal up the windows in the apartment for the winter so no air could get in. You can tell when you go to a place that hasn’t sealed the windows because it smells just like the outside.”
Ulaanbaatar’s air quality is so low in the winter because more than half the city’s residents live in the ger district, which has no running water, no electricity. Instead, residents burn raw coal to cook and to heat their homes during the city’s icy winters. The smoke billows out and settles in a dense haze over the city. Ulaanbaatar is in a valley, locked between mountains. The air has nowhere to go.
As bad as its air quality is, Ulaanbaatar isn’t the worst. Delhi holds the title for world’s most polluted city, followed by other cities in India and Pakistan. Even in small doses, air pollution is a major threat to human health. According to the World Health Organization, it’s linked to 7 million premature deaths every year. In particular, carbon monoxide, sulfur dioxide, nitrogen dioxide, and particulate matter are linked to heart disease, and air pollution could also be a contributing cause of strokes, respiratory problems, and increased lung cancer.
Malavika Vyawahare, who grew up in Delhi, talks about air pollution in her city and her own respiratory problems.
Air quality is a sticky environmental justice issue that divides not only the Global North from the Global South but also the wealthy from the poor within individual cities and countries. Ulaanbaatar’s poorest residents live in the ger district. Wealthier residents of Ulaanbaatar and of polluted cities all over the world can afford to live near parks, in sealed and air-conditioned apartments, in the cleanest air. Now, open data activists like Hasenkopf and Flasher are aiming to tackle the entire global problem through open data.
Proponents of open data—data gathered by the government, researchers, or NGOs and then released in an open format so anyone can use it—have postulated that this information could help track progress toward—and even to achieve—Sustainable Development Goals as well as address persistent development issues like corruption and poverty. Skeptics counter that open data isn’t always as open as it appears and may just support a predetermined narrative. Furthermore, they argue, data can only go so far to spur change; public awareness, strong institutions, and political commitment are more important.
Hasenkopf believes open data about air quality really can lead to concrete change—and she points to the US Environmental Protection Agency (EPA) as an example. The Clean Air Act of 1963 stipulated that any locale that might be out of compliance with the EPA’s air quality standards must have an air quality monitor constantly registering levels of ozone, particulate matter, carbon monoxide, sulfur dioxide, nitrogen dioxide, and lead. That data then gets uploaded to and stored on the EPA’s website, where anyone can download the historical data for any of the EPA’s monitors. Researchers use that data to model air quality across the entire country and thus to understand correlations between air quality and health.
Pete Spragins talks about air pollution in Los Angeles, CA.
Thanks to the Clean Air Act, the EPA, and interested academics, we know an enormous amount about the health effects of air pollution at relatively low levels—the levels present in the US and Europe. We know comparatively little about the health effects of the kind of pollution present in cities like Ulaanbaatar and Delhi.
“It seems that, man, if you want to have that understanding anywhere, you’d want to have it in the most polluted places,” Hasenkopf says.
So she and Flasher decided to take a stab at the problem.
“We were inspired by the air pollution monitor on the roof of the US embassy in Beijing that tweeted out and shared air quality data in real time. We thought it’d be a good idea to do something similar in Ulaanbaatar and that, together—me with my software skills and Christa with her scientific skills—we could do it,” Flasher explains. So they started monitoring particulate matter at the National University of Mongolia. “We tweeted and Facebook-posted real-time air quality data. We were surprised by the influence of doing something so simple [the data even got mentioned in Mongolia’s presidential debates], and that’s when we realized the power of open, real-time air quality data.”
So they decided to grow their project beyond Ulaanbaatar: to the whole world.
Hasenkopf estimates that, every day, there are about one million air quality data points “that are shared somehow in the internet ether and then are lost for the record or for easy access immediately after the fact.” She and Flasher set out to “store that stuff and make it openly available and then attribute it to the air quality agency that did the work of capturing it in the first place.”
Since August, Hasenkopf and Flasher’s website, OpenAQ, has been gathering data from government air pollution monitors all over the world, putting the data into a standard format, storing it, and making it openly available and freely accessible. Unlike previous efforts, such as AQICN, OpenAQ gives raw numbers instead of just a scale indicating how polluted a given location is.
Hasenkopf and Flasher have grants from Thriving Earth Exchange (a collaboration between the American Geophysical Union and Amazon Web Services) and the Earth Journalism Network and in-kind support from Development Seed to run the website, but they haven’t paid themselves for their time. Nonetheless, Flasher says, “I don’t see it so much a volunteer effort as an investment into a long-term project.”
Every month, their website gets about 100,000 hits from about 90 different countries.
Hasenkopf hopes academics will use the information published by OpenAQ to do large-scale epidemiological studies that could contribute to “the world’s understanding of how air quality affects health…at really polluted levels, levels you would find in China or India or a good chunk of the world.”
Thus far, academics have used the data for visualizations. One group of researchers is trying to use it to help calibrate satellites’ measurements of on-the-ground air quality. Others are looking to review the success of policies in New Delhi that only allow even license plates to drive on some days and odd license plates to drive on other days. Hasenkopf and Flasher believe that the website will become even more useful from a policy perspective as the data set gets older and people can calculate monthly or even annual averages.
The act of using open data for research actually exposes the holes in the data that governments must be most eager to hide. OpenAQ shows how few air quality monitors there are in Africa: the continent draws a near blank on OpenAQ’s maps. The one air quality monitor in the highly polluted capital of Chile is located in the heart of one of the city’s largest parks and thus is not representative of the city as a whole when researchers and activists make international comparisons. This gerrymandering of data is part of what OpenAQ can expose, Hasenkopf says.
“As long as we have the data and its geographic coordinates, people can make their own judgments,” she says.
You need some technical chops to use the data. Hasenkopf and Flasher don’t vet the data at all; they just cite their sources – all government-run sites.
“What you see is that sometimes we will have, say, negative values in our system, which are not meaningful measurements, but we don’t want to cut them out because then you can see where the instruments that probably aren’t working correctly are,” Flasher explains. “But, yeah, if you wanted to take a copy of our whole database, you would need to understand where the data was coming from and what it meant.”
Thus, data has the potential to be useful, but only when analyzed in its appropriate context.
Nick Shapiro, a fellow at the Chemical Heritage Foundation and an Open Air Fellow at Public Lab, says “probably the best use of data is for opening up questions and making new questions ask-able.” Why is Santiago de Chile’s only air pollution monitor located inside a park, right next to the city’s wealthiest neighborhood? Where does the air pollution in China come from? Who is suffering, who is benefitting, and who is responsible?
The Spanish architect Nerea Calvillo is extremely skeptical about the ability of open data to drive on-the-ground changes. That’s because of her experience launching In the Air, a project she hoped “would raise awareness and become a platform for political action against air pollution.” She scraped air quality data from Madrid’s 24 government monitors and built a visualization of it on her website.
When the project had been running for some time, the visualizations suddenly stopped working one day. Calvillo investigated and realized that Madrid’s City Council had taken the monitors from the city’s worst-polluted sites and either moved them to parks or gotten rid of them entirely.
“The argument from the City Council was that they were adapting the network to the European Union regulations,” Calvillo remembers. “Basically what they were doing was deciding which air to monitor to get a better average and then comply with the EU regulations.”
“Political parties and NGOs denounced it, putting legal claims against the City Council, and creating a scandal in the press,” Calvillo remembers. Though the stations were not moved back, the incident “contributed to making air pollution a public matter of concern.”
Hasenkopf points to the EPA as an open air quality data success story. Indeed, the EPA uses research from academics when it conducts periodic reviews of air quality standards (known as Integrated Science Assessments) and publishes revised standards.
In other words, says Hasenkopf, “open data spurs the whole cycle” of revising air pollution standards in the US.
But the open data exists because of legislation, not vice versa. No one I spoke to for this article pointed to an example of a time when open data drove new legislation regulating air pollution, only when local communities organized to drive change. Making the leap from open data to political will is the as-yet-unattained goal.
After working on In the Air for a number of years, Calvillo says she “realized that [the problem of air pollution] is much more complicated” than she originally thought. “Making visible the invisible is not enough,” she concluded. “That may raise awareness and change how human beings feel, … but political change is much more difficult.”
That is, the fact of exposing data doesn’t in itself give rise to solutions.
“More data is good, but it’s not the most important thing,” Calvillo says. “The change is going to come from political decisions about how to change our production systems, our consumption systems, our transportation systems. We know they’re bad already, so we can make decisions already.”
Do we need to know how polluted Ulaanbaatar is, what the exact health effects of its air are, to decide to try to solve the problem of cooking and heating with raw coal?
“Sometimes to take action we don’t need numbers. We just need to decide to change things,” she says. “Do we actually need more and more and more data? Do factories pollute? It is not a question of how much. They’re polluting, and that’s what matters.”