Carnegie Mellon University

Pater Adams and his PhD students

Center for Air, Climate, and Energy Solutions

Decreased lung function, heart attacks, and even death—air pollution can have serious health impacts. CEE PhD student Marguerite Marks has firsthand experience with health issues linked to air pollution. While Marks was pursuing her PhD, her new baby developed a virus-induced asthma called bronchiolitis, an ailment that can be exacerbated by air pollution.

“Not only is it hard to watch your child struggling to breathe, but, over the past two years, I’ve spent five nights in the hospital, with multiple doctors’ visits and medications. He’s fine now, but kids with severe asthma whose parents have to do that all of the time, that’s a really big toll both emotionally and economically,” says Marks, who hopes her current work with the Center for Air, Climate, and Energy Solutions (CACES) will help illuminate the effects of air pollution.

Founded in 2016 with $10 million in funding from the Environmental Protection Agency (EPA), CACES is a five-year collaborative research center at Carnegie Mellon University with a focus on the impacts of air pollution, technology, climate change, and related policies on local air quality and health. Within Carnegie Mellon alone, the students and faculty in CACES represent six departments, including CEE, and over a dozen faculty from outside institutions are involved.

CEE Professor Peter Adams is a principal investigator with CACES and has been involved with the center from its start. “The goal of CACES is to give decision makers better tools to think about energy systems, energy transitions, climate, and health implications, so that they can look at everything holistically and make smarter decisions for the benefit of society,” he explains.

Understanding Past Health Impacts


Working with Adams at CACES, Marks is modeling the historical levels of a type of particulate matter called PM2.5, so named for having a diameter of 2.5 microns or less.

While PM2.5 is known to be harmful to our health, few measurements of PM2.5 in the United States were made before 1997, when the EPA first regulated it. Without that data, researchers have struggled to fully assess the health effects of PM2.5.

Marks is working to fill in that data gap back through the 1980s. After Marks prepares and inputs the historical meteorology and emissions data she’s collected from the EPA and other sources, she develops estimates for PM2.5 as well as other gaseous pollutants over the specified time period.

“There’s a much better history of measuring gaseous pollutants, like NO2, ozone, and SO2, so we have several means to test the model’s validity for the 1980s,” explains Marks. “We can also spot check using some scattered PM measurements that were done at universities in those earlier years.”

Once Marks has confirmed the accuracy of the model’s estimates, epidemiologists in CACES will analyze her data alongside extensive national health data to improve understanding of the impact of varying PM2.5 exposure levels and compositions on the US population.

“We’re certain that it’s bad to breathe in pollution, but there are many nuances that are less well understood,” she says. “If you’re exposed when you’re a kid, is that worse long term than if you are exposed as young adult? Or is it worse if there are multipollutant mixtures like ozone in addition to PM and NO2? Does the pollution source matter? Is it worse to live near a freeway than a factory? We don’t know. Combining the detailed health data with comprehensive air-quality data will let us dig into those questions.”

Improving Physical Air-Quality Modeling

While Marks prepares her models, Adams is also working with two Engineering and Public Policy students—Shayak Sengupta and Pablo Garcia—to develop sophisticated air-quality models that will provide more localized and detailed pollution information than has ever been available.

With these models, organizations will be able to understand better the social costs and benefits, including health impacts, of policy and technology changes.

“There aren’t ready-made tools that tell you, if you emit air pollution here, here, and here, or if you switch from coal to natural gas, or if you replace dirty Pittsburgh diesel buses with modern diesel buses or something even cleaner, what does that do? How many lives does it save? Is it worth the investment? Those are the kind of questions that we want to answer,” says Adams.

Currently, the EPA and other researchers study air pollution and health impacts with tools called chemical transport models, or CTMs, typically modeling 12- or 36-km blocks of the country at a time. In CACES, Adams and his team are working to make their models accurate at the 1-km scale.

At this scale, Adams and his students expect to uncover localized, potentially harmful differences in air pollution level and composition that would otherwise go overlooked. With 1-km models, they can also see how decisions impact each individual neighborhood—including areas along bus routes or major roadways—and consider questions of environmental justice to determine whether minority groups or people with low socioeconomic status are exposed to higher air pollution levels.

To prepare these models, researchers use meteorology data and emissions estimates from dozens of manmade and natural sources, including plants, cars, restaurants, power plants, and more. Creating a model with a spatial resolution as high as 1 km requires collecting and estimating this emissions information in great detail across all times of day, days of week, and seasons.

Even at low spatial resolution, CTMs are difficult and time-consuming, and few organizations outside the EPA have the resources or expertise to run them. In particular, state organizations have few tools for deciding how to meet federal regulations—and when those regulations aren’t met, the health of state residents is in danger. That’s why Adams and his team inside CACES are also building reduced complexity models that are much faster and simpler to use.

Developed by Adams, Estimating Air-pollution Social Impact Using Regression (EASIUR) is one such reduced complexity model that can quickly estimate health damages based on the amount of specified pollutant emitted in a certain location. Closely derived from the gold-standard CTMs, EASIUR gives states, cities, and private organizations the power not only to test individual proposals, but also to compare multiple scenarios for improving air quality, giving them clear, defensible answers on the health impacts of various technologies and policies.

“Even at the EPA level, instead of evaluating one or two potential decisions, they could look at a whole suite of options and then home in on the ones that seem most promising,” adds Adams.

To ensure accuracy, the estimates from all models run by Adams and his team will be evaluated against real air-quality measurements taken by another CACES team. That team is measuring air quality across three test cities, including Pittsburgh, driving around to capture and map data at different times in different neighborhoods as well as placing sensors throughout each city.

Exploring Future Energy Scenarios

In addition to air-quality modeling, Adams is assisting with another CACES project, one focused on assessing future scenarios and policies using the new data and tools from CACES.

“This is where we take these tools on a test drive. It’s a way to show that we can pull all of this together into a meaningful analysis of future energy scenarios and transitions,” says Adams, who is advising Engineering and Public Policy PhD student Michael Roth on his work in CACES.

Roth’s goal is to model what the US energy sector and emissions will look like by 2050 under three possible scenarios: a carbon tax, an air pollution tax, and legislation regulating both of those things.

“We’ll be seeing how social welfare changes under different types of regulations. We want to know if anything can be gained from regulating air pollution emissions and greenhouse gas emissions together or if perhaps they should be treated separately,” explains Roth, who will also investigate the possible value of regional and local regulations, and the ways in which having regional air pollution policies might change the national energy system.

Collaboration at the Center of Everything

While regulations and new technology in recent decades have spurred dramatic improvements in US air quality and associated health, nearly 80,000 premature deaths nationwide are still attributed to air pollution every year, and everyone inside CACES realizes there’s a lot riding on their work.

To complete each project, constant collaboration is required. “CACES is a unique and interesting place to work because it’s so interdisciplinary. It brings together different fields, and it’s right in the name—air quality, climate, and energy—it’s this nexus of really crucial things right now,” says Marks, who knows the epidemiologists are counting on her data. In turn, the results of the epidemiology studies will help CACES researchers more accurately model the health impacts of energy decisions.

As these diverse subject matter experts build tools that will guide decisions about the future of our country and planet, everyone is counting on each other. “CACES is so far from the cliché of a lone scientist working at a lab,” says Garcia, who is on the modeling project with Adams. “This way of working is what I have always wanted science to be; a collective enterprise of knowledge building and sharing. I believe the work from CACES is going to have a huge impact on air-quality science and policy for decades.”