Blue Sky Scenario

Photo: Kent Dayton / Harvard Chan
Photo: Kent Dayton / Harvard Chan

Jeremiah Zhe Liu, SM ’15, PhD ’20, is creating new biostatistical methods that may help reverse the devastating effects of air pollution in his native China.

On calm days, an acrid fog descends over Taiyuan, China, the hometown of Jeremiah Zhe Liu. Tucked into the northeastern corner of the nation, this industrial city rises from a landscape that is otherwise quiet and agricultural—a bit like Iowa, Zhe Liu says—except for the presence of huge coal mines and steel plants that employ many of Taiyuan’s 3.5 million inhabitants.

While these sites are economic assets for Taiyuan, the cloud of coal dust, waste gases, and smog they create is so dense that respiratory disease is an accepted fact of life. “It’s something we considered normal growing up,” says Zhe Liu, SM, ’15.  “It was weird not to have some sort of lung disorder, because the air was so bad.”

Throughout his childhood, whenever friends or family members wheezed and hacked, his mother—a public health specialist—would quietly remind him of the connection between these health disorders and the pollution surrounding their community. It wasn’t until he came to the United States in high school, however, that those lessons really hit home.

“I thought blue skies like those I saw in picture books were just a fantasy until I stepped off the plane at JFK,” he says. “The air smelled so different, and I kept wondering, why is there no coal dust on the streets? That’s when I realized how bad it was in my hometown. It made me determined to help save my countrymen from a torture they weren’t even aware of.”

Tourists from mainland China pose for photos in front of an outdoor banner picturing what the Hong Kong skyline looks like on a clear day. Air pollution has reached record levels in Hong Kong n recent years, as it has in mainland China.
Tourists from mainland China pose for photos in front of an outdoor banner picturing what the Hong Kong skyline looks like on a clear day. Air pollution has reached record levels in Hong Kong n recent years, as it has in mainland China./ Photo: © Alex Hofford


With his mother’s encouragement, Zhe Liu launched his academic career in public health by studying statistics and computer science at the University of Iowa. “She taught me that if you know the technical methods, then the substantive part is a lot easier to learn,” he says. In the process, he discovered his talent for mathematical analysis—and first learned about the Six Cities study, the groundbreaking 1993 research from the Harvard T.H. Chan School of Public Health that connected air pollution to mortality risk throughout the U.S., paving the way for stronger federal regulations on fine particulate matter. It shifted his direction as a scholar.

“Before Six Cities, air pollution research was limited to single sites. No one had tried to link multiple cities to do a national-level assessment before,” he says. “I was struck by the scale and ambition of the project,” and, he adds, the fact that it was made possible primarily by the use of innovative statistical methods.

Zhe Liu became determined to master techniques like these and put them to work conducting similar studies in his own country, which is home to many of the world’s most polluted urban areas. In 2014, only eight of China’s 74 biggest cities met the government’s own air quality standards. In the 2014 Environmental Performance Index, China ranked 176th out of 178 nations in air quality.

Having just earned his SM degree from Harvard Chan last May, Zhe Liu is now working toward his PhD here in 2020, forging advances in biostatistics to reach his goal of undertaking research similar in ambition to the Six Cities study.


Although data is easy to access in our digital world, finding information that overlaps in time, space, sampling parameters, and so on isn’t so simple. Zhe Liu is chipping away at one of the field’s toughest methodological problems, says his adviser, Brent Coull, professor of biostatistics.

Say you want to map air pollution in the Greater Boston region. You might find plenty of particle sensors installed on rooftops downtown but only a scattered few out in the surrounding suburbs. To complicate matters, not all of those sensors may have been active at the same time. They may not have been operated by the same organizations. They may not have even been able to count the same-size particle. The result: a mountain of fragmented data. How do you make sense of it all?

Many researchers get around the problem by starting fresh and laying down their own monitoring campaigns for new studies. But that’s expensive and eats up lots of resources. “If we can leverage existing data, it provides a more cost-effective way of getting at larger health questions,” Coull says. “Jeremiah is on the cutting edge of this technique.”


That air pollution devastates human health is settled science. Zhe Liu’s goal is to refine a method that will pinpoint the specific causes and sources of pollution-related disease, which could help policymakers calculate the costs and benefits—to both public health and the overall economy—of reducing those exposures.

“We know that fine particulate matter or PM 2.5”—particles 2.5 microns or less in diameter—“harms respiratory health,” he explains. “What we’re trying to do is tease out the components of PM 2.5 into different pollution signatures: which are emis- sions from highways, which from nearby factories, which from heating oil in old buildings. If you know the source of the pollution, that gives you a priority target.”

In his own lifetime, Zhe Liu saw the sources of pollution shift in his hometown. “The local economy relied on heavy industry and coal mining. But as time went on, the coal started to run out and heavy industry died—but the impact from car emissions and construction sites went up,” he says. “Pollution never dies, but the sources change.”

He doesn’t underestimate the cultural or political obstacles to translating his research into policy. “From my perspective, the only interest of the Chinese government—the only way for it to justify its existence—is to stimulate the economy at an unbelievable speed,” he says. “It will be a major challenge to implement environmental policies unless you’re able to translate public health benefits into economic benefits.”

If he were the country’s air quality czar, Zhe Liu adds, he would first restrict road traffic, in part by encouraging bicycle use. “But is that a realistic policy? Would people be willing to give up the pride and convenience associated with a new car? Would government be willing to go against the interests of the market? I’m not sure.”


As he describes his work, Zhe Liu exudes a quiet confidence, occasionally taking long pauses to find just the right word, obviously comfortable with the silences and even the imponderables. “My fear is that China has already missed its chance and that there’s no turning back,” he concedes. “For a polluted city to recover a healthy atmosphere, the financial investment is almost unthinkable.”

But if China’s authorities do finally commit to solving the pollution crisis, he says, that, too, would be almost unthinkable. For a moment, Zhe Liu imagines Taiyuan with pristine air. “It would remove the physical and mental shackles that people endure for a life- time. They would be saved from a life in which they are destined to suffer lung disease, beginning in infancy; in which children can’t play outside because parents worry about them breathing too much dirty air; in which young people can’t enjoy long walks with their love; in which adults must stay inside after work watching TV, to avoid the harmful air outside; in which so many die from pollution-related heart or lung conditions, starting in their 50s,” he says. “And the sky would be blue.”

David Levin is a Boston-based science journalist. He can be reached through his website

Watch a video or listen to a podcast of Jeremiah Zhe Liu and other Harvard Chan students.